← All themes

Cloud Decentralization

461 statements · 2018–2026

80% of data-center capacity is unused; edge cuts cost 2–5x

Cloud Mobility with Legendary Developer Greg Osuri, Founder and CEO of Overclock Labs (The Silicon Valley Insider Show)

“80% of the capacity sitting in data centers today is unused. So by connecting this latent capacity with the users that want to access this, we can reduce the cost of compute significantly anywhere from one half to one fifth.” — Greg Osuri, 00:07:36

Context: The founding thesis of Overclock Labs — connecting edge nodes (“close proximity to the user”) achieves higher price performance using latent data-center capacity. The 1/2–1/5 cost claim recurs as Photon’s mandate at [00:15:17].


"Burgers and fries" cloud lock-in economics

Cloud Mobility with Legendary Developer Greg Osuri, Founder and CEO of Overclock Labs (The Silicon Valley Insider Show)

“Cloud provides compute, raw compute, bandwidth, memory, CPU, whatnot, and storage. On top of that, they provide value-add services. And the revenue model is basically burgers and fries. You got to sell burgers and not sell fries. But you make money on fries.” — Greg Osuri, 00:25:43

Context: One of his longest-running recurring arguments (repeated for years afterward): clouds lose money on raw compute and lock customers in via ~200 value-added services.


Photon's mandate: cut cloud cost to 1/5–1/2

Cloud Mobility with Legendary Developer Greg Osuri, Founder and CEO of Overclock Labs (The Silicon Valley Insider Show)

“We’re calling it Photon, right, for now, but Photon’s mandate is to solve the cost of cloud computing anywhere from one-fifth to one-half.” — Greg Osuri, 00:15:17

Context: Pre-announcement of the network that became Akash — “we’re going to announce this thing basically in a week or two,” first prototype March 19, 2018, launch targeted for mid-June 2018.


Society must take compute back

Cloud Mobility with Legendary Developer Greg Osuri, Founder and CEO of Overclock Labs (The Silicon Valley Insider Show)

“There’s that innovation being stifled here. And it’s so many things that are wrong with centralization of the cloud. And considering compute is a very, very important, you know, foundational element of our society, we should, as people, have responsibility to take it back.” — Greg Osuri, 00:28:01

Context: Closing his argument about the “Amazon problem” — clouds white-labeling open-source projects (Google Dataproc vs. Databricks Spark) and out-competing their creators.


Three companies will control 80% of internet traffic

Cloud Mobility with Legendary Developer Greg Osuri, Founder and CEO of Overclock Labs (The Silicon Valley Insider Show)

“In two years, majority of the internet traffic will come from cloud infrastructure providers, out of which three companies are going to have 80% market share. So, majority of the internet is going to come from three companies. That’s one problem.” — Greg Osuri, 00:24:08

Context: Asked to lay out the problems with the current cloud landscape; his core oligopoly thesis, quantified with a two-year horizon (from March 2018).


Centralized cloud is the biggest threat to the future of society

The Decentralized Future of Cloud Infrastructure (Outlyer)

“I’m just gonna go ahead and say this: I think centralized cloud is the biggest threat to the future of our society… I want to make sure you do [agree] by the end of the talk.” — Greg Osuri, 00:01:34

Context: Opening thesis of the 2018 talk; the foundational contrarian claim from which the whole Akash project is argued.


Most global compute sits idle — bring it to market

The Decentralized Future of Cloud Infrastructure (Outlyer)

“If you have capacity — data centers that are being utilized only 15 percent of the time — the majority [of] the global compute power is sitting there… Why don’t we bring that capacity to the market? This is fundamentally wrong.” — Greg Osuri, 00:06:40

Context: The underutilization thesis behind the Akash marketplace, stated in 2018 (captions garble the figures); followed by the Industrial Revolution vertical-monopoly comparison and “the future of computing is at risk when three companies control the industry.”


The "burgers and fries" cloud business model exploits open source

The Decentralized Future of Cloud Infrastructure (Outlyer)

“They follow a model called burgers and fries model: they give you compute cheap [and] charge you for other things that are not compute… Majority [of] the revenues for [Google] comes [from] value-added services… Google Dataproc, which is essentially Databricks’ Spark… Databricks does not make anything [like] what Google makes on that product. Similarly you see [it with] Redis, with Amazon ElastiCache.” — Greg Osuri, 00:03:47

Context: Argues cloud adoption destroyed the hosted-model monetization of open source — “centralized cloud is the biggest [beneficiary] of open source software [and] they don’t really contribute that much.”


The pendulum swings back: from mainframes to PCs to cloud to decentralization

The Decentralized Future of Cloud Infrastructure (Outlyer)

“In the 60s the model was centralized — the mainframe had all the power… in the 80s we saw the rise of PCs, which… distributed power… then we saw the rise of the public cloud with the evolution of mobile phones — the mobile phone was a very thin access layer to the centralized thing called a cloud. And now you’re seeing a model going back to the decentralized model with projects like ours… Filecoin, it’s storage… we took an approach of addressing the entire generic workload model.” — Greg Osuri, 00:14:02

Context: Closing historical argument that compute centralization is cyclical and 2018-era decentralized infrastructure projects are the next swing.


Three companies will control the internet

The Decentralized Future of Cloud Infrastructure (Outlyer)

“Two years from now, majority [of] the internet will come from cloud — cloud infrastructure, to be specific. About 53 percent of global data center traffic will be from three major providers, and the market is amazing — 71 billion dollars… So cloud is here to stay, cloud is here to take over, and three companies are going to control… [the] internet.” — Greg Osuri, 00:03:02

Context: A dated, quantified 2018 forecast about hyperscaler concentration, foreshadowing his later “four companies control the internet” framing.


500,000 data centers at ~85% underutilization

Greg Osuri: The Akash Network (Block 2thefuture)

“Connecting data centers around the world, underutilized server capacity in data centers around the world — and there’re about 500,000 data centers which have about 85 percent of underutilization… the reason for this underutilization is primarily because of planning for peak.” — Greg Osuri, 00:01:27

Context: Quantified supply-side thesis; illustrated with Intuit needing ~95% capacity on Tax Day and ~5% on April 16th.


80% cost reduction vs. incumbent cloud

Greg Osuri: The Akash Network (Block 2thefuture)

“Akash is a decentralized and open marketplace for cloud computing resources, and the value proposition to the customer is to 80% cost reduction than what you would pay for the cloud today.” — Greg Osuri, 00:00:45

Context: Opening pitch of the talk; the cost claim is grounded in tapping idle data center capacity.


Clouds stifle open-source innovation

Greg Osuri: The Akash Network (Block 2thefuture)

“Google effectively took what is open source and white-labeled and called their own, and a company that actually developed Spark doesn’t get to take the market… I have friends of mine whose companies were shut down — open source companies — because cloud copied this offer.” — Greg Osuri, 00:03:40

Context: His “burgers and fries” argument that clouds monetize open-source services; Databricks/Spark vs. Google Dataproc as the example.


Majority of internet traffic from three providers within two years

Greg Osuri: The Akash Network (Block 2thefuture)

“Effectively in two years majority the internet traffic will come from three providers, and they happen to be Amazon, Google, Microsoft — and this is effectively very threatening to our society because we all know what happens when you have oligarchs control the foundational elements.” — Greg Osuri, 00:02:12

Context: Timed prediction (from April 2018, so by ~2020) used to justify building a decentralized alternative.


Akash's early architecture: on-chain procurement, off-chain deployment

Panel: Decentralized Storage (Block 2thefuture)

“For us, we’re a decentralized cloud infrastructure marketplace… Akash is where you procure your infrastructure and deploy and you settle — so that’s three pieces of the puzzle: procurement, deployment, and settlement. We use blockchain for procurement as an open marketplace… whereas deployment is off chain; we can’t really use blockchain per se because of performance, security… reasons.” — Greg Osuri, 00:05:07

Context: 2018 explanation of the Akash design, three years before mainnet — the marketplace-on-chain/workloads-off-chain split that shipped.


Decentralization is key for future society

Panel: Decentralized Storage (Block 2thefuture)

“Well, security and ownership — we all know what happens when your data [is] centralized, with Facebook there’s no auditability, there’s nothing going on but your data. So I think decentralization is key here for future society.” — Unknown, 00:04:25

Context: Answer to why decentralizing storage matters, weeks after the Cambridge Analytica story broke. Speaker: possibly not Greg — captions are not diarized and several panelists answered in sequence.


The future is a combination of centralized and decentralized systems

Panel: Decentralized Storage (Block 2thefuture)

“User interfaces in decentralized systems, they’re hard because you’re writing and reading from the chain, [and] depending on what chain, it’s really slow. So I think it’s gonna be a combination of centralized systems and decentralized systems — you look at Coinbase, centralized, whereas Bitcoin is not… we can do much better, but it’s really really hard. I think that’s gonna be a big huge topic for the next few years.” — Unknown, 00:14:27

Context: On usability trade-offs. Speaker: likely Greg (flows directly into his answer about California legislation), but captions are not diarized.


Cloud decentralization is happening now — not in five to ten years

S1E2 - Greg Osuri (CryptoChange.News)

“Oh, it’s happening right now… I think five, ten years is a long, long view. Especially with the rate at which the innovation is catching up… when multiple people are thinking of this problem, that validates the problem. And one of them will win. I mean, I’m saying we might, but someone else will win.” — Greg Osuri, 00:11:10

Context: Asked whether decentralized cloud arrives “within the next five, ten years”; he credits maturing container/Kubernetes tech plus blockchain evolving into “a programmatically managed monetary policy” (spans into the 00:11:52 block).


Data is doubling every two years but the cloud isn't keeping up

S1E2 - Greg Osuri (CryptoChange.News)

“What’s happening today is the price is increasing because the amount of data we are processing is doubling every two years… by 2020, we’ll have 44 trillion gigabytes or something like that… that’s more than the stars in the known universe… But the cloud is not catching up. The infrastructure side of things are not catching up. And increasingly, the entire industry is becoming an oligopoly controlled by three major companies.” — Greg Osuri, 00:04:51

Context: Answering why cheap compute is his mission; the 44-zettabyte-by-2020 figure comes from IDC’s data-growth forecasts of the era.


Society must reclaim compute from the elites through decentralization

S1E2 - Greg Osuri (CryptoChange.News)

“Oligopolization leads to verticalization. Verticalization leads to price [gouging], innovation stifling… we have to claim the foundational element of the Internet. The society has to claim it back. And the only way that’s possible is through decentralization, taking the power back from the elites and spreading [it] across. And that’s the mission of Akash Network.” — Greg Osuri, 00:06:25

Context: Follows his US Steel / Industrial Revolution analogy and the claim that 60% of server-hardware spend now comes from three cloud companies, so servers are designed to their specs.


Burgers-and-fries: clouds out-compete open source

"Crypto Stories: A Primer on Distributed Computing with Dani Grant, Dan Desjardins, Chandler Song, and Greg Osuri" (Village Global's Venture Stories)

“The majority of the revenue that comes to the centralized cloud providers happen to be from these value-added services. It’s like burgers and fries model. Give you cheap compute, but we make money on the value-added services. And these value-added services are usually white-labeled open-source projects.” — Greg Osuri, 00:14:07

Context: Google Dataproc vs. Databricks Spark example; argues a decentralized network has better incentives to attract open-source projects than hyperscalers do.


Cloud market share creates an oligopoly

"Crypto Stories: A Primer on Distributed Computing with Dani Grant, Dan Desjardins, Chandler Song, and Greg Osuri" (Village Global's Venture Stories)

“Amazon is great. I love Amazon and Google as well. The way the system is set up right now with the centralized model, it’s not Amazon’s fault or Google’s fault per se… with the current market share that’s being gained by these three providers, effectively creates a logopoly.” — Unknown, 00:02:19

Context: Answering “why isn’t AWS good enough?” — his recurring systemic-incentives argument. Speaker: attributed to Greg by argument style and later repetition; panel audio is undiarized (“logopoly” is likely a mis-transcription of “oligopoly”).


Protocols cooperating is the only way to beat the oligopoly

"Crypto Stories: A Primer on Distributed Computing with Dani Grant, Dan Desjardins, Chandler Song, and Greg Osuri" (Village Global's Venture Stories)

“I think the problem definitely exists. And I think the future we’re going to see protocols working with other protocols, like what you guys are doing, each one focusing on different markets… I think that’s the only way we can take on a centralized system there is today, which is oligopoly.” — Unknown, 00:38:54

Context: His closing thought on the panel — decentralized compute as an ecosystem of complementary protocols rather than a single winner. Speaker: attributed to Greg by diction (“oligopoly”) in an undiarized closing round.


85% of server capacity sits unused

E04 - Decentralizing the Cloud with Greg Osuri (HackerNoon)

“85% of server capacity sitting in corporate data centers remains unused. An extreme example: Intuit, the company behind TurboTax — on a tax day they run at 97 percent utilization; come next day it drops… nine months of the year they run at almost three percent utilization… on one hand you have this massive scalability problem, on the other hand you have this massive [unused] supply.” — Greg Osuri, 00:05:49

Context: The supply-side statistic he would repeat for years; leads into “we can effectively create a system where there is no middleman [and] effectively turn every company with capacity into a cloud provider, walking the same path Amazon took” [00:06:31].


A new supercloud from disparate compute

E04 - Decentralizing the Cloud with Greg Osuri (HackerNoon)

“[I’m] working on a decentralized cloud infrastructure network that aggregates computing power from disparate sources and creates a new supercloud.” — Greg Osuri, 00:00:48

Context: October 2018 — his one-line thesis, using “supercloud” years before the term became an industry buzzword.


First decentralized cloud to run a full web application

E04 - Decentralizing the Cloud with Greg Osuri (HackerNoon)

“This is the first decentralized cloud that can run a fully functioning web application — be [it] Ruby, Python, Node, doesn’t really matter.” — Greg Osuri, 00:16:35

Context: His personal blog (gregosuri.com) was hosted on the Akash testnet at the time [00:15:52]; containers let developers keep their existing stacks.


Push a job and forget it — nobody can take you down

E04 - Decentralizing the Cloud with Greg Osuri (HackerNoon)

“For me it’s about censorship resistance and fault resilience — the idea that I can push a job or a function and forget about it; it will run no matter what, as long as I have tokens to pay for it. That is the essence of a decentralized system: there’s no single point of failure and there’s nobody that can take you down.” — Greg Osuri, 00:10:08

Context: Answering why a developer would choose a decentralized system — years before the Parler/deplatforming events he’d later cite as vindication.


Three gatekeepers are stopping the cloud from scaling

E04 - Decentralizing the Cloud with Greg Osuri (HackerNoon)

“The amount of data [and] data processing needs are doubling every two years [but] the processing power hasn’t been catching up… access to modern infrastructure is safeguarded by the cloud infrastructure providers… about twenty-one hyperscale providers, out of which three providers have a majority of the market share… they are being gatekeepers — they are not allowing the cloud to scale.” — Greg Osuri, 00:04:22

Context: His founding market analysis of the AWS/Google/Microsoft oligopoly; quote spans into the [00:05:05] block.


You can't take on Goliath alone

E04 - Decentralizing the Cloud with Greg Osuri (HackerNoon)

“The reason we’re building this is to create a network, or a cloud infrastructure, that is not owned by anyone but open to everyone… an individual standing up cannot take on Goliath — we’re talking about Amazon, Google, Microsoft. The only way we can do this is together.” — Greg Osuri, 00:28:08

Context: Closing call to action inviting protocol developers to join the effort against the cloud oligopoly.


Computing pendulum: from mainframes to cloud and back to decentralized architectures

Serverless Panel: Komal Mangtani, Greg Osuri, Guillermo Rauch, Gwen Shapira (Kong)

“If you look at evolution of computing, in the beginning you had mainframes — it was a very centralized model… then the client-server came along and clients were equally as powerful as servers, so you had a way more distributed model. Then went back to the cloud, where cloud became the centralized entity… Now we’re seeing more decentralized, distributed architectures… you’re gonna see, as we move more into the decentralized distributed cloud, serverless will play a huge role.” — Greg Osuri, 00:18:16

Context: Greg’s core recurring thesis, stated here in 2018: centralization/decentralization cycles in computing history, with the next swing toward decentralized cloud. Follows his 85%-idle-capacity supply argument.


Three companies will control 80% of cloud in two years

CEO of Overclock Labs On the World of Blockchain & Upcoming Technology In Development (CoinBundle)

“It’s a cloud tax — we all pay the cloud tax, we all pay the Amazon tax. More importantly, the direction cloud infrastructure is going today is scary, considering three companies will control 80% of market share in two years. When three companies, oligarchs, control foundational elements of the internet, it’s a very scary situation.” — Greg Osuri, 00:23:21

Context: Asked how everyday users benefit from Akash; he frames cloud concentration (then ~60% share) as a societal danger, comparing it to US Steel’s dominance in the Industrial Revolution.


Akash as a new operating system for the internet tapping 85% unused capacity

Overcoming the Challenges of Distributed Networks with Greg Osuri CEO Overclock Labs & Akash Network (Techpost)

“We believe the foundation should be scalable, frictionless and free of censorship, and we believe in the unstoppable web. What we’re doing with Akash is essentially creating a new operating system for the Internet, and how we’re doing that is by tapping into unused 85% capacity.” — Greg Osuri, 00:03:02

Context: Opening pitch; he completes the figure in the next breath — “in 8.2 million data centers” — describing Akash as a peer-to-peer network for “planetary scale applications.”


Cloud is built for hyperscale; Akash optimizes for thousands of nodes at a fraction of the cost

Overcoming the Challenges of Distributed Networks with Greg Osuri CEO Overclock Labs & Akash Network (Techpost)

“Cloud wasn’t quite cutting it for them in terms of performance because cloud is optimized for hyper scale… high-density data centers, but Akash doesn’t have that limitation. Akash will optimize a global scale for thousands of nodes… by using Akash they could get cost down, I think it’s 4x or 5x, than what they would otherwise pay on the cloud.” — Greg Osuri, 00:08:07

Context: Describing Cactus, a peer-to-peer music streaming app using relay architecture at the edge; next block adds “about 40 times more distribution than the cloud.”


60% of Ethereum nodes run on AWS — that's not decentralization

"Hashing It Out #70 - Akash Network - Greg Osuri" (Hashing It Out (The Bitcoin Podcast Network))

“60% of Ethereum nodes are running on AWS right now, which is kind of sad… it doesn’t really matter how decentralized the network is, [it] really comes down to your physical deployment. If you’re running on Amazon, I don’t consider that decentralized at all, especially when the majority of your network is running on a single cloud provider.” — Greg Osuri, 00:58:06

Context: His motivation for DISCO (decentralized infrastructure for serverless computing operations) and home nodes (quote spans into the 00:58:49 block); he adds that with physical access and no SGX, an adversary intending to shut Ethereum down could do it “in a matter of minutes” (00:59:31).


Akash gives placement sovereignty to the user — and clouds themselves plug in

"Hashing It Out #70 - Akash Network - Greg Osuri" (Hashing It Out (The Bitcoin Podcast Network))

“What Akash does really is open up the market to these massive clusters, to small little deployments all over the world and gives the sovereignty to the user to choose what they want. And of course, the multi-cloud architecture in Akash also means that cloud companies plug into Akash directly, which they’re doing right now.” — Greg Osuri, 00:19:50

Context: From the recovered mid-episode span: latency-sensitive jobs can pin local-residency clusters, then “switch over to a more cost optimal deployment” (00:20:36); he says tier-2 cloud providers were already plugging in, framing Akash as additive multi-cloud rather than a cloud replacement.


150 edge metros at sub-10ms latency via the Packet program

The Akashian Challenge Phase 1 Livestream (Akash Network)

“We have about 150 edge sites that Akash is ready to go on… when we unlock those edge sites through Packet, I believe it’s going to be 150 metro areas delivering at under 10 millisecond latency — that’s extremely attractive for high-content, high-bandwidth streaming.” — Greg Osuri, 00:50:19

Context: On the Packet (Equinix) edge program; positioning Akash for streaming and low-latency edge workloads circa 2020.


Interoperability is the future of the decentralized cloud

The Akashian Challenge Phase 1 Livestream (Akash Network)

“We believe the future is interoperability, because Akash cannot do everything you need for a full cloud deployment… we’d rather work with teams or projects that have their expertise… interoperability, I believe, will be the future and is key for scalability and adoption.” — Greg Osuri, 00:40:56

Context: On IBC integrations planned with NuCypher (key management), Solana (smart contracts), stablecoins for settlement, and storage networks (IPFS, Sia).


Only 9 of 91 testnet nodes were on AWS, vs Ethereum's 60%+

The Akashian Challenge Phase 1 Livestream (Akash Network)

“There were, I believe, nine or ten nodes that went down right around the time there was an outage in AWS — so only nine out of 91 nodes are actually on AWS. I take that with a lot of pride, because Ethereum has over 60% of the nodes on AWS, or cloud in general.” — Greg Osuri, 00:06:43

Context: Recapping the AWS outage that hit mid-testnet; used as evidence Akash’s validator base practices the decentralization it preaches.


Unstoppable, borderless workloads as a core use case

The Akashian Challenge Phase 1 Livestream (Akash Network)

“Because Akash is unstoppable — that means you cannot take down the workload unless you have the key to the workload… borderless knowledge sharing, borderless commerce, truly borderless that cannot be stopped, is a big use case for Akash.” — Greg Osuri, 00:52:25

Context: Describing Phase 3 incentives for “ownerless oracles” (e.g., DAO-safe frontends like Uniswap’s) and censorship-resistant publishing.


85% of global cloud capacity sits unused

"The Akashian Challenge Livestream: Phase 3 Week 1" (Akash Network)

“We tap into world’s unused cloud capacity — 85% of the global cloud capacity that’s sitting in data centers is not used… probably even more, we’re being conservative with our estimates.” — Greg Osuri, 00:32:55

Context: His supply-side thesis, illustrated with Intuit: TurboTax servers peak at 95-96% in tax season but run ~10% the other nine months, and servers have a ~3-year shelf life — “if you don’t use it you lose it.”


Akash: 2x-10x cheaper than the incumbent cloud

"The Akashian Challenge Livestream: Phase 3 Week 1" (Akash Network)

“Akash Network is the world’s first non-custodial cloud that can offer anywhere up to two times to ten times lower cost than the current cloud.” — Greg Osuri, 00:30:01

Context: Rebutting a questioner calling Akash a “blockchain solution to a problem that doesn’t exist”; he adds cost savings matter most for machine learning and big-data workloads.


Akash as the gateway to the decentralized cloud stack

"The Akashian Challenge Livestream: Phase 3 Week 1" (Akash Network)

“A big vision for us for Akash is to become the gateway for decentralized cloud… what we want is to enable Akash users to use key management on NuCypher and pay using Akash token… so that cross-chain interoperability is very, very important for Akash.” — Greg Osuri, 01:00:30

Context: Answering why Akash integrated Kava/USDX; multi-currency settlement so one token pays for compute, storage, DNS, and key management across networks — “just like you don’t use 100 credit cards.”


A world where the cloud runs on Superminis in people's homes

Greg Osuri - AKASH Network (At Stake)

“Building a free and open cloud is very, very important. And extending that to regions that experience oppression is very, very important… So you can imagine a world where people host their own super minis in different parts of the world and run the cloud on those super minis. That’s a fantastic outcome, right?” — Greg Osuri, 00:22:57

Context: Answering a question about censorship resistance; ties home compute to unstoppable, censorship-resistant services (messaging in oppressed countries) and to peers like Orchid and Sentinel.


Akash as the gateway to a "Cambrian explosion" of decentralized cloud

Greg Osuri - AKASH Network (At Stake)

“We envision Akash to be in that forefront as a gateway to the decentralized cloud… I think we’re just seeing a Cambrian explosion of decentralized dCloud or dWeb… of projects that are actually shipping products today… the next step for us as an industry is to figure out how to reduce friction for adoption.” — Greg Osuri, 00:33:23

Context: Vision of Akash orchestrating Handshake (DNS), NuCypher (key management), Filecoin (storage) behind one interface with unified payment — “hyper-connected and hyper-interoperable decentralized cloud.”


Akash's mission is to break the cloud oligopoly

Greg Osuri - AKASH Network (At Stake)

“Cloud infrastructure, the backbone of the internet, it’s getting increasingly centralized. It’s controlled by the monopoly of literally four companies… you pay the cloud tax, the Amazon tax, we call it, invisibly. And the mission of Akash is to really break the logopoly.” — Greg Osuri, 00:05:16

Context: Opening framing of Akash; “logopoly” is Whisper’s rendering of “oligopoly.” Continues into the peer-to-peer marketplace pitch for “resilient and unstoppable applications at a much lower cost.”


Non-custodial UX: a "Uniswap for cloud deployments"

Greg Osuri - AKASH Network (At Stake)

“That is a power of non-custodial interfaces… extending that similar experience for cloud is amazing… We don’t have a web interface yet, but you can kind of see how that will turn out to be like a Uniswap for like cloud deployments.” — Greg Osuri, 00:13:13

Context: He calls user experience the “unrealized benefit” of decentralized cloud, bigger than the cost advantage — a recurring UX-first argument.


Akash as the gateway to a decentralized cloud

"The Akashian Challenge: Closing Live Stream" (Akash Network)

“We are a big contributor to IBC, and the reason why we wanted to use IBC is because we believe in a frictionless interoperable cloud, [with] Akash really being positioned as a gateway to a decentralized cloud.” — Greg Osuri, 00:50:16

Context: Greg outlining 2021 plans (mainnet 2, economics/incentives, interoperability) — early statement of the recurring “gateway/supercloud” framing.


Akash as the open-source AWS with provider add-on services

"The Akashian Challenge: Closing Live Stream" (Akash Network)

“Amazon AWS started with S3 and maybe some compute, and today it’s an entire jungle of services — it’s a monolith, they control the whole thing. We see this as the open-source version of that, and we see it going to having that amount of diversity of services available on the network.” — Adam Bozanich, 00:57:20

Context: Answer on private Docker image support, envisioning provider-offered add-on services (log retention, fast storage). Speaker: Adam Bozanich (CTO), not Greg; Greg adds the design principle of “doing less things but doing it really well.”


Akash could host 200 Solanas; supply is not the problem

The Akashian Challenge Livestream: Phase 3 Week 2 (Akash Network)

“100%. I mean, I think we can cover at least 200 Solanas with the current capacity that we have… we don’t have a supply problem.” — Greg Osuri, 00:49:27

Context: Asked whether rented Akash capacity would suffice in 2021-22 to cover a significant part of Solana’s compute needs; Greg says demand, not supply, is the constraint.


Akash in 2031: a $300B market, everywhere like the cloud

"The Akashian Challenge: Closing Live Stream" (Akash Network)

“300 billion dollars at 100 [percent] market penetration would be nice. It’ll be like the clouds — it’ll be everywhere.” — Unknown, 00:58:45

Context: Viewer asked what Akash will look like in 2031. Speaker: likely Greg (captions not diarized; Adam follows with “we’ve been on this thing for quite a long time… it could really be a game changer”).


At maturity Akash will have thousands, if not millions, of providers

The Akashian Challenge Livestream: Phase 3 Week 2 (Akash Network)

“How do you detect this illegal content? How do you remove it? Especially considering if you’re Amazon you probably have one company, but in Akash you have thousands if not millions of providers when we reach maturity.” — Greg Osuri, 00:52:18

Context: Answering how Akash handles illegal content; the scale forecast of “millions of providers” is the notable quantified piece.


Governance-based content takedowns would mean censorship

The Akashian Challenge Livestream: Phase 3 Week 2 (Akash Network)

“The moment you say we can introduce a governance process, that means… you’re introducing censorship… we don’t have the tools and technologies to even understand what the content is.” — Greg Osuri, 00:53:00

Context: Greg’s contrarian stance that Akash will not adopt on-chain governance for content removal, shifting moderation liability to individual providers instead.


The real problem: cheaper, better, faster compute

"The Akashian Challenge: Closing Live Stream" (Akash Network)

“The real problems we’re solving is cheaper, better, faster compute… Akash, by outsourcing and really bringing on different providers and letting users set the price, you are actually reducing the cost of compute… the amount of steps you need to take to go deploy something on Amazon is exponentially greater than what you can do on Akash.” — Greg Osuri, 01:12:58

Context: Greg’s answer to “many blockchains aren’t solving real problems — what problem is Akash solving?”; adds “the product is the best due diligence.”


Usability plus censorship resistance is the winning combo

The Akashian Challenge Livestream: Phase 3 Week 2 (Akash Network)

“From one hand you have this incredible usability gains, from the other hand you have this amazing censorship resistance in a permissionless way… all this great adoption that we had is just a testament to these two incredible important features that Akash can provide.” — Greg Osuri, 00:45:54

Context: After analogizing Uniswap’s few-click swaps vs. a centralized exchange’s 20-step flow, Greg claims Akash extends that same usability leap to cloud deployment vs. AWS.


Users will self-deploy DEX UIs signed by web wallets

The Akashian Challenge Livestream: Phase 3 Week 2 (Akash Network)

“Right now it’s command line, but soon enough… you can see a world where their web interfaces are much more simple interfaces, ideally signed by web wallets, and the users themselves can deploy any DEX UI. So that’s really where we’re going.” — Greg Osuri, 00:43:48

Context: Discussing ThorChain wanting users to host their own DEX UIs on Akash; Greg frames non-custodial hosting as shifting regulatory liability from creator to user.


Akash as the first open-source cloud

Citizen Cosmos and OmniFlix Network: Conversations with Jack ft. Greg Osuri Akash Network, S1E1 (Citizen Web3)

“That gave way to really building a tool that can leverage Kubernetes and leverage the enormous amount of computing we have to build a second generation, new generation open source cloud — the first open source cloud — and that’s really what Akash is.” — Greg Osuri, 00:02:07

Context: His founding framing: cloud infrastructure was the one layer open source hadn’t yet captured.


Crypto nodes concentrated on Amazon — Thorchain bounty

Citizen Cosmos and OmniFlix Network: Conversations with Jack ft. Greg Osuri Akash Network, S1E1 (Citizen Web3)

“We got picked up by Thorchain… they were having an effort to decentralize because a lot of their nodes are running on Amazon… I reached out to them: look, we can run Thor nodes pretty well, how about doing a thing — and I gave a bounty… about four thousand dollars worth of RUNE tokens and AKT tokens just for writing a guide.” — Greg Osuri, 00:12:49

Context: Early instance of his recurring “crypto runs on centralized cloud” argument; he calls it “the first time someone else is paying or incentivizing to put their nodes on Akash” [00:14:12].


IBC removes the last obstacle to replacing centralized cloud

Citizen Cosmos and OmniFlix Network: Conversations with Jack ft. Greg Osuri Akash Network, S1E1 (Citizen Web3)

“You have to use different tokens and different signatures and all kinds of things — so imagine trying to automate a solution that spans across these networks, like impossible, right? IBC is going to solve that in an amazing way… the real power is when you’re going to solve these friction problems for decentralized cloud adoption. Once you solve that, there’s nothing in front of us to actually replace incumbents, like centralized cloud providers, in a unified way.” — Greg Osuri, 00:22:46

Context: His “interoperable cloud” vision: Akash compute + Filecoin/Arweave/Sia storage + NuCypher KMS + Handshake DNS composed via IBC into a full cloud stack.


Akash as the first open-source cloud

Conversations with Jack ft Greg Osuri - S1E1 - by OmniFlix Network & Citizen Cosmos (OmniFlix Network)

“That gave way to really building a tool that can leverage Kubernetes and leverage the enormous amount of computing we have to build a second, new generation open source cloud — the first open source cloud — and that’s really what Akash is.” — Greg Osuri, 00:02:06

Context: His founding framing: cloud infrastructure as the layer open source hadn’t yet captured.


Crypto nodes concentrated on Amazon — Thorchain bounty

Conversations with Jack ft Greg Osuri - S1E1 - by OmniFlix Network & Citizen Cosmos (OmniFlix Network)

“They were having an effort to decentralize because a lot of their nodes are running on Amazon, and Thorchain — if you know, the team is an anonymous team — they embrace the decentralized values really well, to the core. And I reached out to them: look, we can run Thor nodes pretty well, how about doing a thing — and I gave a bounty… overall it’s about four thousand dollars worth of RUNE tokens and AKT tokens just for writing a guide.” — Greg Osuri, 00:13:31

Context: Early version of his recurring “crypto runs on centralized cloud” critique; “the first time someone else is paying or incentivizing to put their nodes on Akash” [00:14:13].


IBC removes the last obstacle to replacing centralized cloud

Conversations with Jack ft Greg Osuri - S1E1 - by OmniFlix Network & Citizen Cosmos (OmniFlix Network)

“You have to use different tokens and different signatures and all kinds of things — so imagine trying to automate a solution that spans across these networks, like impossible, right? IBC is going to solve that in an amazing way… the real power is when you’re going to solve these friction problems for decentralized cloud adoption. Once you solve that, there’s nothing in front of us to actually replace incumbent, like centralized cloud providers, in a unified way.” — Greg Osuri, 00:22:39

Context: The “interoperable cloud” vision: Akash compute + Filecoin/Arweave/Sia storage + NuCypher KMS + Handshake DNS composed via IBC; quote spans into the [00:23:21] block.


85-95% of compute is not being used

Akash Network, Cloud Computing on Blockchain (MikoBits Blockchain, NFT and DeFi Show)

“We began the company as an edge computing platform… that’s when we discovered this is a massive opportunity of unused compute. When we were actually deploying this platform we noticed 85% or so of the compute is not being used — that’s our conservative estimate. Some of our estimates are actually like 95% of the compute is not being used.” — Greg Osuri, 00:34:59

Context: Origin story from 2015; the underutilization statistic is the foundational thesis behind Akash’s supply model.


Akash is 2-3x, sometimes 10x, cheaper than the big clouds

Akash Network, Cloud Computing on Blockchain (MikoBits Blockchain, NFT and DeFi Show)

“Any big value is cost. I mean, costs anywhere from two to three x lower than the Amazons and Google to Microsoft, in some cases even ten times lower. So that’s incredible value proposition.” — Greg Osuri, 00:01:27

Context: Opening description of Akash, alongside “what takes normally tens of hours on the cloud takes minutes” and censorship resistance “becoming more and more important considering what’s happening today” (Jan 2021).


An interoperable decentralized cloud with better UX than today's cloud

Akash Network, Cloud Computing on Blockchain (MikoBits Blockchain, NFT and DeFi Show)

“…interoperable world, where the user experience is going to be much better than what you’ll experience on the cloud today, and we are working with NuCypher and Filecoin to make this happen.” — Greg Osuri, 00:39:14

Context: Predicting SDL will orchestrate Graph, Handshake, NuCypher, Filecoin, Sia, and Arweave together, payable in any protocol’s token (“I think we’re going to see a very interoperable world” begins in the prior caption block).


The fastest growing cloud in the world during testnet

Akash Network, Cloud Computing on Blockchain (MikoBits Blockchain, NFT and DeFi Show)

“During the testnet phase we were the fastest growing cloud in the world, with 3,800 deployments, about 480 active developers. To give you a context, Polkadot has about 420 monthly active developers.” — Greg Osuri, 00:02:12

Context: State-of-the-network answer before Mainnet 2 (platform launch dated Feb 14-15, 2021 in the conversation).


Akash's mission: society controls the system

The Unstoppable Live Stream w/ Greg Osuri and Brian Fox (Akash Network)

“The mission of Akash is to develop a modern, high performance, sovereign open cloud… where the system is not controlling the society but the other way — the society is controlling the system. It’s the first attempt of creating an open source cloud.” — Greg Osuri, 00:09:11

Context: Framing computing as “the fabric of human society” put at risk by a closed oligopoly; the crypto token is the sustainability mechanism open source lacked.


Banned communities will redeploy on Akash

The Unstoppable Live Stream w/ Greg Osuri and Brian Fox (Akash Network)

“My most satisfying moment will be like when something gets shut down on Discord, someone should put up a Matrix or a Rocket Chat on Akash; something gets shut down on Reddit, someone should boot up a Discourse forum on Akash; a website gets taken down, someone should go start a blog on Akash… liberty is foundational for a free society.” — Greg Osuri, 00:19:55

Context: Days after the WallStreetBets censorship episode, which both he and Fox reference; asked what he’d most like to see deployed on Akash.


Decentralized software will beat centralized UX

The Unstoppable Live Stream w/ Greg Osuri and Brian Fox (Akash Network)

“I think decentralized software will be far better, far superior in experience to centralized software, just because the overhead they need is so minimal — and we’re just starting to scratch the surface.” — Greg Osuri, 00:31:19

Context: Comparing two-command Akash deploys against AWS signup/verification friction, and Uniswap swaps against centralized-exchange withdrawals; concedes token on-ramps remain the unsolved problem.


The cloud is closed: four companies control 81%

The Unstoppable Live Stream w/ Greg Osuri and Brian Fox (Akash Network)

“That cloud is increasingly becoming more closed, with 81 percent or so market share controlled by four companies.” — Greg Osuri, 00:07:46

Context: Answering what inspired him to build Akash; he likens cloud oligopoly to what Microsoft/Apple did to open software.


Cloud needs its Linux 1991 moment

The cloud, open source and flash loans with Greg Osuri (Citizen Web3)

“I mean, we had operating systems that were closed before 1991… 1991 came about with the introduction of Linux, the first free and open operating system that fundamentally changed the fabric of software.” — Greg Osuri, 00:04:57

Context: His analogy that a closed cloud is like closed operating systems pre-Linux; decentralized cloud is the open successor, with benefits “we haven’t discovered” yet.


Everybody pays the Amazon tax

The cloud, open source and flash loans with Greg Osuri (Citizen Web3)

“So you are paying for the cloud without knowing you’re paying for the cloud. There is this infamous term called the Amazon tax. Everybody pays an Amazon tax, but you just don’t realize it. So it’s a very integral part. It’s a core fabric that runs the internet and it’s closed.” — Greg Osuri, 00:04:15

Context: Answering why anyone needs a decentralized cloud; cloud framed as the operating system of modern applications.


Oligopolies stifle innovation — and cloud is becoming one

The cloud, open source and flash loans with Greg Osuri (Citizen Web3)

“The structural issues with oligopolies are oligopolies tend to stifle innovation… It’s not like people stop flying. It’s just people stop innovating in flight. It’s not like people stop making phone calls, it’s just people stop innovating in telecom… And cloud is becoming one.” — Greg Osuri, 00:06:31

Context: Airlines/telco analogy for what happens when moats concentrate; ties into his claim that open-source companies can no longer get funded because Amazon white-labels their software.


The "Amazon is a safe bet" ideology is cracking

The cloud, open source and flash loans with Greg Osuri (Citizen Web3)

“Yes, this whole like Amazon is a safe bet ideology is actually cracking because the bigger something gets, the more inefficient it becomes. That’s just a natural law.” — Greg Osuri, 00:18:37

Context: Citing Amazon outages that took down Binance, FTX, and BitMEX in the prior month; ~4,000 invisible cloud providers exist with no marketplace to reach customers.


World's first peer-to-peer cloud marketplace

The cloud, open source and flash loans with Greg Osuri (Citizen Web3)

“So in just the simplest terms, Akash is, I would claim, the world’s first peer-to-peer marketplace for cloud computing… So in a sense, Akash is our first free market-driven cloud.” — Greg Osuri, 00:00:41

Context: Opening definition of Akash; “free market-driven” meaning buyers and sellers set prices instead of the big three cloud companies.


2–3x price advantage from market-driven pricing

The Akash Mainnet 2 Livestream w/ Greg Osuri and Adam Bozanich (Akash Network)

“2x to 3x, in some cases much higher, price advantage compared to the current market.” — Greg Osuri, 00:04:14

Context: Because Akash uses a market-driven reverse-auction model where “the user sets the price instead of the companies” (00:03:31).


Adoption will climb a credibility ladder, starting with DeFi front-ends

The Akash Mainnet 2 Livestream w/ Greg Osuri and Adam Bozanich (Akash Network)

“Getting to trust a platform is climbing the credibility ladder. As we prove ourselves that this is a viable alternative, we’re going to see a lot more workloads. I think right off the bat we’re going to see a ton of decentralized finance front-ends using Akash… a lot of developer tooling… a lot of block explorers and ecosystem tooling.” — Greg Osuri, 00:35:17

Context: Asked about expected customer segments; he also predicts heavy Cosmos sentry-node usage and says start with non-mission-critical workloads like CI servers.


Akash as the gateway for decentralized cloud

The Akash Mainnet 2 Livestream w/ Greg Osuri and Adam Bozanich (Akash Network)

“It positions Akash as the gateway for decentralized cloud… if you want storage — Akash doesn’t do storage, Akash does compute — long-term storage on Filecoin or Sia coin or Arweave can be used via IBC.” — Greg Osuri, 00:09:56

Context: Vision of composable web3 infrastructure where compute, storage, and DNS interoperate across chains.


Breaking the four-company cloud oligopoly

The Akash Mainnet 2 Livestream w/ Greg Osuri and Adam Bozanich (Akash Network)

“The cloud, which is the foundational fabric of modern society… is increasingly becoming centralized, controlled by four companies. Number one thing Akash addresses is breaking this oligopoly and giving cloud back to the people.” — Greg Osuri, 00:02:48

Context: Answer to “what need does Akash’s decentralized cloud marketplace address”; his foundational recurring thesis. (Captions render “cloud” as “clout.”)


Holy grail: managed services and a serverless supercloud

The Akash Mainnet 2 Livestream w/ Greg Osuri and Adam Bozanich (Akash Network)

“The holy grail eventually is to support managed services, and we’re working towards it… we want to do so in an open source and in a permissionless manner like what we do with Akash… it’ll be great to actually have your database managed by an expert versus self-managing your database.” — Greg Osuri, 00:37:23

Context: He completes the thought as “the whole holy vision for a serverless supercloud that is fully open source and fully decentralized” (00:38:06) — an early on-record use of the supercloud framing.


The cloud's promise is a false premise — outages prove it

The Akash Mainnet 2 Livestream w/ Greg Osuri and Adam Bozanich (Akash Network)

“Cloud gives you this false premise that, hey, it’s ultimately scalable — put all your eggs in one basket and we’ll take care of everything. And what happened today, this morning: Binance went down, FTX went down, because they all trusted Amazon is the panacea… it’s just false. These things keep going down over and over and again.” — Greg Osuri, 00:04:57

Context: A major AWS outage that morning had disrupted crypto exchanges; he positions Akash as a disaster-recovery/failover site since “you can’t keep your eggs in one single basket.”


85% of the world's compute capacity sits underutilized

The Akash Mainnet 2 Livestream w/ Alessio Treglia, Jack Zampolin, Greg Osuri and Adam Bozanich (Akash Network)

“As we started deploying we discovered there’s a huge problem that happened to be underutilization of compute… around 85 percent of the compute capacity out there is underutilized.” — Greg Osuri, 00:09:15

Context: Greg’s origin story: companies provision data centers for peak load (Intuit tax season, Walmart holidays), leaving massive idle capacity — the supply side of Akash’s marketplace thesis. He also cites ~8.2 million data centers that could leverage Overclock’s software.


Akash can eventually scale to compete with AWS and Google Cloud

The Akash Mainnet 2 Livestream w/ Alessio Treglia, Jack Zampolin, Greg Osuri and Adam Bozanich (Akash Network)

“What you planned to build is something that can eventually scale to compete with AWS and Google Cloud and all of these really major cloud providers, and the only way to do that is to build a free market because these companies have figured out how to suppress competition.” — Jack Zampolin, 00:22:45

Context: Framing the Equinix deal as validation (“hundreds of billions of dollars in infrastructure spend behind it”). Speaker: Jack Zampolin, not Greg.


Akash testnet developer engagement rivaled Polkadot's mainnet

The Akash Mainnet 2 Livestream w/ Alessio Treglia, Jack Zampolin, Greg Osuri and Adam Bozanich (Akash Network)

“Polkadot currently has about 420 monthly active developers, we had about 480, and Ethereum mainnet has about 2,500 monthly active developers, so that gives you a perspective as to the engagement that we were able to attract.” — Greg Osuri, 00:19:14

Context: Greg recapping Akashian Challenge Phase 3 results (~480 active developers, ~90 unique applications deployed ~3,800 times).


Cloud providers are middlemen exploiting demand without solving supply

The Akash Mainnet 2 Livestream w/ Alessio Treglia, Jack Zampolin, Greg Osuri and Adam Bozanich (Akash Network)

“We have this increased demand for the cloud which currently is about 370 billion dollars and you have this increased supply that’s not used, and you have this middlemen called cloud providers that they take advantage of the increased demand but not really solving the supply problems.” — Greg Osuri, 00:10:40

Context: Core recurring Akash argument — the founding inefficiency thesis that led to an open, decentralized compute marketplace.


Equinix Metal will bootstrap Akash's compute capacity

The Akash Mainnet 2 Livestream w/ Alessio Treglia, Jack Zampolin, Greg Osuri and Adam Bozanich (Akash Network)

“The biggest of course is our integration with Equinix. Equinix Metal is by far the largest interconnected data center network in the world — we’re talking about over 220 data centers in about 25 countries.” — Greg Osuri, 00:02:52

Context: Company update ahead of Mainnet 2; Greg adds (00:04:16) that “as we scale and as we get adoption we will likely see more and more high quality data centers like Equinix.”


Global compute in the hands of two or three companies is dangerous

The Akash Mainnet 2 Livestream w/ Alessio Treglia, Jack Zampolin, Greg Osuri and Adam Bozanich (Akash Network)

“I think we could all agree that there’s a lot of danger in centralization of basically global compute, so it’s in the hands of a couple or three companies now. We know that they’ve colluded with one another before on other things.” — Adam Bozanich, 00:34:04

Context: Listing benefits of deploying on Mainnet 2. Speaker: Adam Bozanich (co-founder/CTO), not Greg.


Mainnet 2 is the first non-custodial, self-sovereign, permissionless cloud

The Akash Mainnet 2 Livestream w/ Alessio Treglia, Jack Zampolin, Greg Osuri and Adam Bozanich (Akash Network)

“It’s the first time you’re seeing a non-custodial, self-sovereign and a permissionless cloud that’s fully open source… there’s nothing like this that ever exists out there, and really excited to see the possibilities that we haven’t imagined yet.” — Greg Osuri, 00:35:33

Context: Greg summing up mainnet 2 benefits — no logins, no email, no personally identifiable information required to deploy.


Oligopolies stop innovating — the cloud is next

Greg Osuri, CEO of Akash Network, Decentralized Cloud | Blockchain Interviews (Crypto Coin Show)

“Oligopolies tend to create products that are misaligned sometimes to the customer base, and they tend to create competitive moats that makes it prohibitive for new players to come and partake in what should be a competitive marketplace… cloud, if you think about it, is the fabric that runs the society.” — Greg Osuri, 00:02:09

Context: Comparing the big-four cloud to airlines and telecom — industries that “just stopped innovating” after consolidation.


Oligopolies stop innovating — the cloud is next

Greg Osuri, CEO of Akash Network, Decentralized Cloud | Blockchain Interviews (InvestmentPitch Media)

“Oligopolies tend to create products that are misaligned sometimes to the customer base, and they tend to create competitive moats that makes it prohibitive for new players to come and partake in what should be a competitive marketplace… cloud, if you think about it, is the fabric that runs the society.” — Greg Osuri, 00:02:09

Context: Comparing the big-four cloud to airlines and telecom — industries that “just stopped innovating” after consolidation.


Ownerless web services owned by the world

Greg Osuri, CEO of Akash Network, Decentralized Cloud | Blockchain Interviews (Crypto Coin Show)

“I had this vision of ownerless web services, where people can actually own web services — not a single entity, not a single human being, but the world can actually own a web service that is managed by a DAO and has a threshold signature for deploying and controlling — all that can be realized now with the decentralized cloud.” — Greg Osuri, 00:12:48

Context: On IBC composability arriving; paired with his promise of a “Uniswap-style deployment experience” with no emails, passwords, or credit cards.


Ownerless web services owned by the world

Greg Osuri, CEO of Akash Network, Decentralized Cloud | Blockchain Interviews (InvestmentPitch Media)

“I had this vision of ownerless web services, where people can actually own web services — not a single entity, not a single human being, but the world can actually own a web service that is managed by a DAO and has a threshold signature for deploying and controlling — all that can be realized now with the decentralized cloud.” — Greg Osuri, 00:12:48

Context: On IBC composability arriving; paired with his promise of a “Uniswap-style deployment experience” with no emails, passwords, or credit cards.


Akash as Airbnb for compute across 8.2 million data centers

The Akash Mainnet 2 Livestream w/ Sunny Aggarwal, Greg Osuri and Adam Bozanich (Akash Network)

“Akash is positioned to have this hyper customizable cloud, and that potentially can be reached, installed in 8.2 million data centers in the world. So the way you want to picture Akash is really: if Akash is Airbnb [for] compute, Amazon is like Hilton.” — Greg Osuri, 00:22:17

Context: Explaining audited attributes — an unbounded, free-market attribute system versus the “homogeneous” experience of big chains and the bounded customization of centralized clouds.


Akash is the gateway to the decentralized cloud

The Akash Mainnet 2 Livestream w/ Sunny Aggarwal, Greg Osuri and Adam Bozanich (Akash Network)

“Akash is the only network that does compute… when you have to deploy something, you got to deploy compute and then you attach all the services to compute… so that puts us in a very amazing position of being this gateway for the decentralized cloud.” — Greg Osuri, 00:56:25

Context: On integrating with Handshake/ENS (DNS) and Sia/Filecoin/Arweave (storage); he cautions “do not believe the hype is adoption” and predicts a product-market-fit shakeout of decentralized networks “in the next few years” ([00:57:06]).


In 10 years Akash will be fully self-sustaining; 1% of cloud = $3B

The Akash Mainnet 2 Livestream w/ Sunny Aggarwal, Greg Osuri and Adam Bozanich (Akash Network)

“In 10 years I would imagine it’s going to be a fully self-sovereign… self-sustaining system, and if you think about the cloud, it’s what, 300 billion dollars, right? So even a reasonable one percent market share is three billion dollars of… economic activity coming into the network.” — Greg Osuri, 01:12:02

Context: Answering how the network stays secure after 2030 as inflation decays: staking rewards get replaced by take income from network activity (“very hypothetically forward-looking numbers”).


Predicted 2x savings vs. Amazon, saw 8x

The Akash Mainnet 2 Livestream w/ Sunny Aggarwal, Greg Osuri and Adam Bozanich (Akash Network)

“Some of the users have gained or noticed 8x cost savings in comparison to Amazon Fargate… realistically our prediction was like 2x, but seeing 8x was… a really cool thing.” — Greg Osuri, 00:04:27

Context: First-week Mainnet 2 metrics; he adds savings vary with supply/demand and “I’m pretty sure next week is going to be a double or more” of the 53 active developers.


85% of data-center capacity is underutilized

AKASH NETWORK (AKT) - Decentralized clouds, Staking, Cosmos & more! - Interview with CEO Greg Osuri (Michaël van de Poppe)

“The customers running their own data centers are sitting on humongous underutilized resources — 85% or so of data center capacity that’s deployed is underutilized, and we’re talking about 8.2 million data centers.” — Greg Osuri, 00:04:24

Context: The founding insight behind Akash, discovered while deploying Kubernetes-based solutions to enterprise customers at Overclock Labs.


Akash is the first open source cloud

AKASH NETWORK (AKT) - Decentralized clouds, Staking, Cosmos & more! - Interview with CEO Greg Osuri (Michaël van de Poppe)

“The only way to do that is to remove the sense of control and centralization, to really go with an open model both for the source code… and provide a platform that’s permissionless, self-sovereign and non-custodial — and that’s what Akash is. Akash is, in a sense, the first open source cloud.” — Greg Osuri, 00:13:00

Context: Summing up the three problems Akash targets — cost, developer productivity, and censorship (he cites deplatforming risk even for “small misalignment with the company’s policy”).


Goal: cloud power affordable to 10-year-olds within four years

AKASH NETWORK (AKT) - Decentralized clouds, Staking, Cosmos & more! - Interview with CEO Greg Osuri (Michaël van de Poppe)

“I learned how to code when I was 12… and I had to borrow computing. I want to make sure the next generation is not going to find it hard… I want to make it affordable to a degree that 10-year-olds can get cloud power if they want. That’s — in four years I’d like to see that.” — Greg Osuri, 00:51:42

Context: His answer to the host’s “life goal” question, rooted in growing up on a farm in India without a computer; the quote begins in the [00:50:58] block.


Hyperscalers should offload compute onto Akash

AKASH NETWORK (AKT) - Decentralized clouds, Staking, Cosmos & more! - Interview with CEO Greg Osuri (Michaël van de Poppe)

“You don’t see them as competition, you see them as complement… it’s actually better for these centralized cloud providers — Amazons and Googles and Microsofts — to use Akash to offload compute. In fact… we’ve seen some amazing interest in a lot of the big guys.” — Greg Osuri, 00:35:54

Context: Akash as “the cloud of clouds” (Akash = Sanskrit for sky, “where the clouds are formed”); Equinix — ~220 data centers in ~25 countries, supplier to the hyperscalers — had just become a provider.


85-95% of data center compute sits unused

Interchain.fm Ep. 15 Akash: Taking on AWS with Decentralized Cloud Computing (Cosmos)

“Around eighty-five percent or so of compute that sits in this data centers remain unused, in some cases even ninety-five percent… the reason for that is planning for peak.” — Greg Osuri, 00:02:10

Context: The founding observation from Overclock Labs’ Kubernetes deployments that led to the Akash marketplace idea in 2016.


Akash is 8x cheaper than Amazon today

Interchain.fm Ep. 15 Akash: Taking on AWS with Decentralized Cloud Computing (Cosmos)

“Some of the users are getting eight times lower costs than Amazon… a half gigabyte box… cost[s] about 16 [dollars] a month; on Akash it costs about one dollar eighty cents… it’s cheaper, it’s faster, and it’s censorship resistant, and no one can stop you from using Akash.” — Greg Osuri, 00:08:30

Context: Explaining why blockchain projects (e.g., anonymity-preferring ThorChain) were early adopters — cost plus permissionlessness.


"If it runs on Amazon, it runs on Akash"

"Base Layer Episode 209: Greg Osuri, Founder of Akash Network on Decentralizing Compute" (Base Layer)

“One of the things about Akash is it’s a generic cloud platform. So any machine learning application can run essentially. It’s not like a specialized computing language or anything like that. If it runs on Amazon, it runs on Akash.” — Greg Osuri, 00:29:56

Context: Distinguishing Akash from special-purpose compute networks; a recurring compatibility claim, followed by the closing “Uniswap-style deployment experience for the cloud” line (00:31:24).


ML companies pay ~20% of their margins to Amazon

"Base Layer Episode 209: Greg Osuri, Founder of Akash Network on Decentralizing Compute" (Base Layer)

“Machine learning companies today — Andreessen Horowitz came out the report recently, which is very fascinating… they said about 20% of their margins are actually paid out to Amazon. And this is a very cutthroat business, right? It’s almost like making these businesses unsustainable to a degree.” — Greg Osuri, 00:11:46

Context: Part of his “Amazon tax” argument (a16z’s “The New Business of AI” post); the setup quote about the “burgers and fries” lock-in model is in the 00:10:58 block.


The interoperability "holy grail": pay for any decentralized service with any token

"Base Layer Episode 209: Greg Osuri, Founder of Akash Network on Decentralizing Compute" (Base Layer)

“Ideally, in an ideal world, we want a smooth sort of like experience for a developer where you just come, say what you want to do, and the system should handle that. And that’s the holy grail of where we are going with interoperability.” — Greg Osuri, 00:23:43

Context: Describing composing Akash with NuCypher, Sia/Filecoin, Handshake, Sentinel, Orchid; the follow-on (00:24:27) says IBC will let developers pay with “an Akash token… a handshake token… a graph token.”


Adoption grew 40% week-over-week; only adoption matters

The Unstoppable Livestream: exiledsurfer and Greg Osuri (Akash Network)

“Everything we do contributes to these two numbers, because adoption turns out it’s the only thing that matters. And if you look at adoption, we grew forty percent week [over] week from last week in terms of active deployments.” — Greg Osuri, 00:04:15

Context: Greg’s daily metrics: active deployments and ease of deploying. He cites “52 active deployments with a little over 4,000 applications in total” (the “52” is likely a caption artifact, possibly “520”).


Four companies control 81% of the cloud

The Unstoppable Livestream: exiledsurfer and Greg Osuri (Akash Network)

“What, four companies control 81 [percent] of the cloud? … if you look at any industry, [oligopolization] is an issue that most people don’t really talk about — like telecom, right? Telcos: 21 companies literally control almost hundred percent.” — Greg Osuri, 00:26:57

Context: Greg agreeing with exiledsurfer’s critique of oligarchic control of media and finance, extending it to cloud and telecom concentration.


If AWS does what Akash does, we won — open systems always win

The Unstoppable Livestream: exiledsurfer and Greg Osuri (Akash Network)

“If AWS does what Akash does, we won, right? That’s the point. Akash is not a company, it’s a network. It’s a network of network[s] where the owners operate and use the network.” — Greg Osuri, 00:54:37

Context: Answering “what stops AWS/Azure/GCP from copying Akash?” He backs it with history at 00:55:22: “Open systems historically speaking have always exhibited to have the highest forms of scalability… internet succeeded where AOL failed because it’s a permissionless and open system” — growth is “gradually, then suddenly.”


No-code, one-click resilient deployment is coming very soon

The Unstoppable Livestream: exiledsurfer and Greg Osuri (Akash Network)

“Our goal is to reduce time to deploy from 90 minutes to two minutes… you should be able to pick a recipe, like ‘oh, I want a WordPress with fully redundant storage spread across 10 different data centers,’ click, and this is how much I’m willing to pay. Absolutely, that’s going to happen very soon.” — Greg Osuri, 00:46:04

Context: Answering whether non-coders will be able to deploy on Akash; Greg frames Akash as a low-level primitive layer that platforms will build no-code experiences on top of.


The May 2021 crash proved decentralization is the future

The Unstoppable Livestream: exiledsurfer and Greg Osuri (Akash Network)

“This week proved that centralized infrastructure failed where decentralized infrastructure didn’t… if anything that has to be proven this week, [it] was a testament to why decentralization is the future.” — Unknown, 00:14:07

Context: After exiledsurfer described Sovryn staying online during “wrecking ball Wednesday” while Coinbase and Binance went down. Speaker: likely Greg responding, but attribution not fully certain (undiarized captions).


Time-to-deploy will drop from ~90 minutes to two minutes

The Unstoppable Livestream: exiledsurfer and Greg Osuri (Akash Network)

“It takes about 45 minutes to hour and a half to really understand how the system works and how to deploy. I like to see that come down to two minutes. I know it’s a hard target but I think we can do it.” — Greg Osuri, 00:07:49

Context: Greg on Akash’s biggest product goal: “phenomenal user interfaces,” web and CLI, and a product easier than incumbent cloud.


$300B cloud market heading to $400B; Akash as the only open-source cloud

Почему дорого запускать стартап в централизованных облаках? Интервью с Greg Osuri, CEO Akash Network (S3XYBTC)

“Мы часть рынка, оцененного в 300 миллиардов долларов, который по прогнозам вскоре станет 400 миллиардным рынком. Никто не знает, насколько сильно он вырастет в отдаленном будущем. У нас есть этот гигантский рынок, и мы хотим позиционировать себя как единственное облако с открытым исходным кодом.” (“We are part of a market valued at $300 billion, forecast to soon become a $400 billion market. No one knows how much it will grow in the distant future. We have this giant market, and we want to position ourselves as the only open-source cloud.”)” — Greg Osuri, 00:25:53

Context: After describing Amazon/Google/Microsoft monetizing open-source projects (Docker cited as a company once valued near $1B that couldn’t monetize); open-source cloud framed as the fix.


8 million data centers vs. a choice of four companies

Почему дорого запускать стартап в централизованных облаках? Интервью с Greg Osuri, CEO Akash Network (S3XYBTC)

“Вы получаете доступ к огромному количеству предложений от примерно 8 миллионов дата-центров в мире. Вы не ограничены выбором из четырех компаний.” (“You get access to a huge number of offers from roughly 8 million data centers in the world. You are not limited to choosing among four companies.”)” — Greg Osuri, 00:08:19

Context: The “Airbnb for cloud” analogy — choice and exotic configurations (edge placement near cell towers, low-latency transit providers) versus the Hilton-like uniformity of AWS.


No one can take an app off the platform — not even me

Почему дорого запускать стартап в централизованных облаках? Интервью с Greg Osuri, CEO Akash Network (S3XYBTC)

“Akash предоставляет такие гарантии. Никто не может убрать приложение с платформы, даже я.” (“Akash provides such guarantees. No one can remove an application from the platform — not even me.”)” — Greg Osuri, 00:05:10

Context: Contrasting with Amazon/Microsoft deplatforming projects over ideology (the Parler takedown era); his definition of permissionless.


Startups bleed 20% to Amazon — mostly on GPU bills

Почему дорого запускать стартап в централизованных облаках? Интервью с Greg Osuri, CEO Akash Network (S3XYBTC)

“Сейчас очень дорого запускать и компанию в облаке… Эксперты говорят, что 20% прибыли идет Amazon. Это безумие. Если вы стартап и 20% ваших денег идет в Amazon на оплату счетов за графические адаптеры, я не думаю, что вы долго сможете оплачивать такие счета. Это серьезно вредит стартапам.” (“Right now it’s very expensive to launch a company in the cloud… Experts say 20% goes to Amazon. That’s insane. If you’re a startup and 20% of your money goes to Amazon paying GPU bills, I don’t think you can keep paying those bills for long. It seriously hurts startups.”)” — Greg Osuri, 00:30:05

Context: The interview’s title thesis — centralized cloud cost as an existential drag on startups, especially AI ones.


Akash as the gateway to the decentralized web

We're live with Greg Osuri from Akash Network (AKT) (Bittrex Global)

“Our vision is to position Akash as a gateway for decentralized cloud… Akash will be positioning as a gateway to the decentralized web, and in order for us to achieve that status we have to solve interoperability.” — Greg Osuri, 00:43:40

Context: Explaining heavy investment in Cosmos IBC, Osmosis liquidity, Gravity DEX, and an Ethereum bridge as prerequisites for that gateway role.


Every layer of the cloud stack is decentralizing

We're live with Greg Osuri from Akash Network (AKT) (Bittrex Global)

“Every part of the cloud stack is getting decentralized, and we put all of them together into one umbrella called d-cloud, and Akash we’re positioning as the gateway for the d-cloud.” — Greg Osuri, 00:17:26

Context: Explaining the term “d-cloud” (which he says Akash coined): compute (Akash), storage (Sia/Filecoin), DNS (Handshake), key management (NuCypher).


From 90-minute to one-minute deployments

We're live with Greg Osuri from Akash Network (AKT) (Bittrex Global)

“Today it takes about 90 minutes for someone that has no idea what Akash [is] coming to the website and actually deploying an application… I want to bring that to one minute — click, click, click.” — Greg Osuri, 00:38:45

Context: His second core KPI (time-to-deploy) alongside active leases; achieving it is what he says product-market fit looks like.


Nobody in d-web has product-market fit yet

We're live with Greg Osuri from Akash Network (AKT) (Bittrex Global)

“D-web or d-cloud is not there yet. We are seeing a lot of amazing signals, but I can’t really tell you that we have achieved product market fit — if anybody says that in d-web, they’re lying… people using your product doesn’t mean you have product market fit.” — Greg Osuri, 00:40:08

Context: Unusually candid contrarian assessment of the 2021 decentralized-web sector, contrasting it with DeFi and NFTs which he says do have PMF. (Caption garbles “they’re lying” as “they’re like right.”)


Tapping 8.2 million underutilized data centers

We're live with Greg Osuri from Akash Network (AKT) (Bittrex Global)

“The premise of Akash Network is to tap into 8.2 million data centers that remain underutilized — 85% or so capacity in those data centers being underutilized.” — Greg Osuri, 00:12:34

Context: His foundational supply-side thesis for why a decentralized compute marketplace can exist; recurring figure across 2020-2021 appearances.


No one can take Akash down — including Greg

The Biz Call: 20210428 GregOsuri AkashNetwork (robfrankel)

“It’s open, so it’s unstoppable, it’s permissionless — no one can stop you from using Akash, including myself. I cannot take it down… there’s a big reason of having the sovereignty — having this guarantees business continuity, so we are seeing a lot of companies using Akash as a disaster recovery or a failover kind of thing.” — Greg Osuri, 00:05:38

Context: Why customers choose Akash beyond 8x cost savings, amid 2021 deplatforming concerns.


The exit strategy is acquiring the trillion-dollar market

The Biz Call: 20210428 GregOsuri AkashNetwork (robfrankel)

“We have like a trillion dollar market right now — the exit strategy is to acquire the trillion dollar market. Eventually, Akash is an open system, so anyone can use it… there is no reason for me to not think that we would [reach] 100 [percent] market penetration.” — Greg Osuri, 00:09:10

Context: Answer to “what’s the exit strategy or long-term goal?”


Akash is Airbnb, Amazon is Hilton — from Equinix racks to devices in the home

Future of Web 3 and the Battle for Data Sovereignty with Greg Osuri, CEO of Akash Networks (Orchid Labs)

“If Akash is Airbnb, Amazon is Hilton. Akash is a very versatile system, so much so that you can have both very mission critical infrastructure that’s running on Equinix, which is the largest interconnected data center network in the world, to devices that can live in the home, with varying degrees of reliability.” — Greg Osuri, 00:12:48

Context: His signature marketplace analogy; he goes on to cite 8.2M data centers, sub-10ms cell-tower colocation, underwater data centers, and future “data centers in space.”


Four companies control 81% of a trillion-dollar-by-2025 market — the "Amazon tax"

Future of Web 3 and the Battle for Data Sovereignty with Greg Osuri, CEO of Akash Networks (Orchid Labs)

“Right now about 81 percent of market share is controlled by four companies, and it is about to be a trillion dollar market by 2025. It is a massive market… cloud… is a fabric for today’s society… everyone pays an Amazon tax, an invisible tax.” — Greg Osuri, 00:09:12

Context: Argument for why matching idle supply with demand is “a dumb problem to correct” — with cost reduction as the natural economic result.


Users see up to 8x cost savings over Amazon

Future of Web 3 and the Battle for Data Sovereignty with Greg Osuri, CEO of Akash Networks (Orchid Labs)

“The results are obviously very beneficial for users. So some users experience anywhere from 8x cost [savings] switching from Amazon to Akash… and more importantly they can do their savings in a fully controlled… way without having any permission whatsoever.” — Greg Osuri, 00:01:27

Context: Describing the reverse-auction marketplace that “lets the user set the price” — “truly setting a market driven pricing model for the cloud.”


85% of data center capacity is unused; Akash is one-third Amazon's price

Akash Networks (And How Handshake Works Together) (SkyInclude)

“We make a better cloud by tapping into unused capacity, and turns out about 85 or so [percent of] capacity in 8.4 million data centers is unused… we provide a much similar capability but much better user experience at a price point that is one-third… what Amazon would charge… because we tap into servers that are not used, and as we get more supply it’s only going to get cheaper.” — Greg Osuri, 00:02:07

Context: His recurring supply-side argument against the four companies controlling cloud; quote spans into the [00:02:51] block for the one-third pricing claim.


Centralized platforms are being weaponized; dweb is disaster recovery

Akash Networks (And How Handshake Works Together) (SkyInclude)

“Even the lower-level services such as Amazon Web Services deplatforming Parler… the point is these companies have the power, and that very weakness is being weaponized by state actors… in a case like that you want to have a disaster recovery platform that has much lower chances of failure… if you look at decentralized web, I think it is a lot more crucial than we originally thought, and 2020 proved that to be the case.” — Greg Osuri, 00:20:32

Context: Prompted by the Fastly outage that morning and deplatforming discussion; he argues protocol changes get more scrutiny so “system-wide failures don’t happen that very often in decentralized networks” [00:19:50].


The holy trinity for an unstoppable web

Akash Networks (And How Handshake Works Together) (SkyInclude)

“You have three protocols that are decentralized… working with each other, interoperating… I call it the holy trinity — Skynet, Akash and Handshake — because that is the big foundation for an unstoppable web.” — Greg Osuri, 00:05:39

Context: On users organically combining Akash (compute), Sia/Skynet (storage), Handshake (naming) into full decentralized deployments.


Akash costs one-third of centralized cloud

DeFi Summit - Conference Day 1 - June 15th (Tuesday) (Dcentral Con)

“When it comes to Akash, the cost advantage of using Akash versus a centralized cloud is one third… sometimes here is one fourth, like when compared to like Google or whatnot. And the reason for that is because of openness. Open marketplaces generally tend to optimize for the forces of demand supply for pricing versus someone, an individual set the price.” — Greg Osuri, 06:15:44

Context: Third of his three advantages of decentralized systems (openness, censorship resistance, cost); quantified 3-4x cost claim vs. hyperscalers, attributed to market-based pricing.


Decentralized cloud is the future (and more secure)

DeFi Summit - Conference Day 1 - June 15th (Tuesday) (Dcentral Con)

“We’re fully transparent. We’re fully open. We have public scrutiny on everything we put out there. That makes it a lot, lot more secure than your traditional cloud. And hands down, I think decentralized cloud is the future.” — Greg Osuri, 06:20:46

Context: Concluding his security answer: Akash enforces resilient architectures (keys spread across five boxes, remote signers) that traditional cloud does not.


Decentralized infrastructure is ready; d-web is here

DeFi Summit - Conference Day 1 - June 15th (Tuesday) (Dcentral Con)

“Decentralized infrastructure is ready. D-web is here. Adoption is growing. And I hope people will now be in a situation that they can take decentralized web more seriously.” — Greg Osuri, 06:34:27

Context: Closing remarks, right after quoting Rosa Luxemburg — “people that don’t move don’t notice their chains” [06:33:46] — likening centralized infrastructure to unnoticed chains.


DeFi's stack isn't decentralized — "just own it"

DeFi Summit - Conference Day 1 - June 15th (Tuesday) (Dcentral Con)

“Graph is essentially indexing nodes running on servers. Where are these servers running? Amazon, right?… Where is that DNS running? Probably Amazon or probably Cloudflare. Right? You’re not decentralized. You’re not. Just own it.” — Greg Osuri, 06:26:34

Context: His “cognitive dissonance we have in crypto, especially DeFi, is adorable” riff [06:25:17], walking through Uniswap’s architecture layer by layer (Ethereum contracts aside, the indexer, web UI, and DNS are centralized).


DeFi should run on Handshake, Akash, Sia/Skynet

DeFi Summit - Conference Day 1 - June 15th (Tuesday) (Dcentral Con)

“I’m still bedazzled as to why Uniswap is not on Handshake. Handshake is a truly decentralized DNS network… I’m still bedazzled why Graph is not running nodes on Akash… I’m still bedazzled like why the UI is not running on Sia or Skynet or Akash… unless, I would say, like three fourth of your infrastructure is decentralized, you’re not DeFi.” — Greg Osuri, 06:27:50

Context: The tools to decentralize the full web stack already exist in 2021; sets a “three-fourths of your infrastructure” bar for calling something DeFi.


Ethereum full node: ~$25k on Infura vs one-third on DeCloud

DeFi Summit - Conference Day 1 - June 15th (Tuesday) (Dcentral Con)

“I believe, to run a full node for an Ethereum node costs about $25,000 on Infura… it costs about $25,000 to run. And decentralized clouds will probably — right now, it’s one third. I mean, so if you like money, you probably will like decentralized cloud.” — Greg Osuri, 06:30:12

Context: Third point (cost) of why DeFi must move to decentralized cloud; he flags the figure himself with “don’t quote me on this.”


Open finance needs open cloud; regulatory clock is ticking

DeFi Summit - Conference Day 1 - June 15th (Tuesday) (Dcentral Con)

“DeFi needs to be open. Open finance needs open cloud. I’m going to say it again: open finance needs open cloud. Being on Amazon, being on cloud, you’re not decentralized. Prove me wrong. Number two, there’s a huge regulatory risk that you are basically avoiding, and time is not going to be on your side very soon.” — Greg Osuri, 06:30:45

Context: His summary of the “how can cloud power DeFi” answer — openness, regulatory risk, and cost make decentralized cloud “a no brainer” for DeFi projects.


Platform power will inevitably be abused

DeFi Summit - Conference Day 1 - June 15th (Tuesday) (Dcentral Con)

“I’m not worried about when they use for good, but I’m really worried about when they use for bad. And if you tell me nothing bad happens, I’ll tell you you’re full of shit. So something will happen, something bad will happen. And that’s when — when a centralized platform can censor a sitting US president, that tells you the amount of power these platforms have.” — Greg Osuri, 06:14:17

Context: Pros/cons question from Mona; Greg’s censorship-resistance argument, adding that state actors are already “weaponizing de-platforming” against organizations they disagree with.


Usability gap will close via abstractions

DeFi Summit - Conference Day 1 - June 15th (Tuesday) (Dcentral Con)

“Will that change tomorrow? Yes, it will, with more abstractions that we build on the decentralized infrastructure, and people are building those abstractions. But as it stands today, it’s hard to use. It’s for advanced usage.” — Greg Osuri, 06:17:13

Context: Listing the cons of decentralized cloud circa 2021 — deploying on Akash still required knowing “resiliency through redundancy”; he predicts abstraction layers fix this.


Akash as the gateway to the decentralized web

Live AMA featuring Greg Osuri from the Akash Network! (Bittrex Global)

“Our vision is to position Akash as a gateway for decentralized cloud… Akash will be positioning as a gateway to the decentralized web, and in order for us to achieve that status we have to solve interoperability.” — Greg Osuri, 00:38:23

Context: Justifying heavy investment in Cosmos IBC, Osmosis liquidity, Gravity DEX, and an Ethereum bridge.


Every layer of the cloud stack is decentralizing

Live AMA featuring Greg Osuri from the Akash Network! (Bittrex Global)

“Every part of the cloud stack is getting decentralized, and we put all of them together into one umbrella called d-cloud, and Akash we’re positioning as the gateway for the d-cloud.” — Greg Osuri, 00:12:11

Context: Explaining the “d-cloud” term Akash coined: compute (Akash), storage (Sia/Filecoin), DNS (Handshake), key management (NuCypher).


From 90-minute to one-minute deployments

Live AMA featuring Greg Osuri from the Akash Network! (Bittrex Global)

“We care about that a lot. I want to bring that to one minute — one minute, click, click, click… automate this entire bootstrapping or booting process from 90 minutes to one minute.” — Greg Osuri, 00:34:08

Context: His second core KPI (time-to-deploy, then ~90 minutes for a newcomer) alongside active leases; hitting it is what he says product-market fit looks like.


Nobody in d-web has product-market fit yet

Live AMA featuring Greg Osuri from the Akash Network! (Bittrex Global)

“D-web or d-cloud is not there yet. We are seeing a lot of amazing signals but I can’t really tell you that we have achieved product market fit — if anybody says that in d-web, they’re lying… people using your product doesn’t mean you have product market fit.” — Greg Osuri, 00:34:50

Context: Candid contrarian assessment of the 2021 decentralized-web sector versus DeFi/NFTs, which he says do have PMF. (Auto-caption garbles “lying” as “live right.”)


Tapping 8.2 million underutilized data centers

Live AMA featuring Greg Osuri from the Akash Network! (Bittrex Global)

“The premise of Akash Network is to tap into 8.2 million data centers that remain underutilized — 85 percent… capacity in those data centers have been underutilized.” — Greg Osuri, 00:07:14

Context: His foundational supply-side thesis for the decentralized compute marketplace; recurring figure across 2020-2021 appearances.


~70% of blockchain nodes run on Amazon — with their keys

Enter Decentralized Cloud & Unstoppable Web with Akash Network & Deeper - DeFi Conference Day 1 (Dcentral Con)

“I see a lot of nodes being deployed on Amazon — I think arguably about 70% of the nodes running are running on Amazon today, and they all have their keys in Amazon itself. If the server on the Amazon gets compromised, you’re screwed… no one is going to save your keys, not Jeff Bezos — in fact, they probably have access to your keys and you don’t even know about it. It’s that bad. Current cloud is so insecure that people don’t talk about it unless something really bad happens.” — Greg Osuri, 00:12:45

Context: Arguing d-cloud is more secure because Akash enforces “resiliency through redundancy” (keys spread across five providers, remote signers); concludes “hands down, decentralized cloud is the future” ([00:14:11]).


"Centralized is yesterday... the future is decentralized — it's an evolution"

Enter Decentralized Cloud & Unstoppable Web with Akash Network & Deeper - DeFi Conference Day 1 (Dcentral Con)

“For me, I don’t see… centralized is yesterday, and then the future is decentralized. It’s an evolution… that means all the things that we have in terms of efficiency and whatnot will come in decentralized networks itself, and that is just a factor of time.” — Greg Osuri, 00:05:44

Context: Responding to Eric Ma’s pros/cons of centralization; Greg concedes efficiency but frames every centralized advantage as temporary.


Akash will offer more managed services than Amazon's ~250

AKASH NETWORK - DECENTRALISED CLOUD - DEEP DIVE - WITH THE CEO - 2 OF 2 (Crypto Rich)

“Akash is going to come up with managed services, but we’re going to do that in a decentralized way like we’ve done with compute. It’s a very hard problem to solve… it’s going to take a while, but when that comes — Amazon has about 250 managed services, but Akash will have a lot more than that because of the decentralized permissionless marketplace.” — Greg Osuri, 00:12:42

Context: Explaining that managed services (where AWS “makes most of the money”) are the long-term roadmap beyond bare-bones container hosting.


Decentralization is the only way to break the cloud oligopoly

AKASH NETWORK - DECENTRALISED CLOUD - DEEP DIVE - WITH THE CEO - 1 OF 2 (Crypto Rich)

“Four companies control 80 percent of the market share and there’s really no way to break this oligopoly except to take a route which those four companies can’t even fathom taking, which is decentralization and open source.” — Greg Osuri, 00:04:18

Context: Explaining why Akash takes a decentralized approach against AWS, Google, Microsoft, and Alibaba’s combined dominance.


Open infrastructure is the template for the future

AKASH NETWORK - DECENTRALISED CLOUD - DEEP DIVE - WITH THE CEO - 1 OF 2 (Crypto Rich)

“Once you rebuild your traditional infrastructure to be more open, you’re essentially creating the template for the future that a viable open systems is possible, and you don’t have to build on the traditional cloud. I think that’s going to be the biggest paradigm shift for Akash and rest of the world.” — Greg Osuri, 00:19:10

Context: Describing his favorite demand cohort — innovators rebuilding docker registries, Cosmos validators, and WordPress to run on decentralized rails.


Power should be with the people, not platform gatekeepers

AKASH NETWORK - DECENTRALISED CLOUD - DEEP DIVE - WITH THE CEO - 1 OF 2 (Crypto Rich)

“The problem is that whoever heads the institution has all the power, and I don’t like that. I rather want the power to be with the people. So we built Akash Network to keep the power to the people.” — Greg Osuri, 00:20:35

Context: Discussing deplatforming (Parler taken down by AWS/Cloudflare) as evidence that centralized clouds can silence groups regardless of politics.


Cloud is becoming a trillion-dollar oligopoly

"Akash Network Deep Dive: The Unstoppable Cloud, Powered by Cosmos!" (Cryptocito)

“You have this cloud basically getting consolidated into four companies — like Amazon, Google, Microsoft and Alibaba — all creating effectively an oligopoly… and that industry is growing at such a rapid pace I believe it’s going to be a trillion-dollar industry very very soon.” — Greg Osuri, 00:05:44

Context: Origin-story framing of the supply-demand mismatch Akash connects; he later argues oligopolies “don’t innovate — it’s all about value extraction” using airlines and cell carriers as examples (00:19:57), and cites companies paying ~20% of revenues to Amazon (00:21:22).


Decentralization will win on UX, not ideology

"Akash Network Deep Dive: The Unstoppable Cloud, Powered by Cosmos!" (Cryptocito)

“Right now it’s not decentralized fully, but will it be decentralized? Yes. And what is the trigger? It’s not going to be ideology — we know that, people don’t care about values as much as you and I think they should… I think it’ll come down to user experience and collaborative features on top.” — Greg Osuri, 00:30:41

Context: Prediction about how the full web stack migrates to decentralized infrastructure; he concedes Akash’s frontend hosting UX then lagged Netlify/Vercel.


Everyone pays an invisible Amazon tax

Everyone Pays an Amazon Tax with Greg Osuri, Founder of Akash Network (Orchid Labs)

“When you see how important it is, the kind of role it plays in our daily lives without even realizing — everyone pays an Amazon tax, an invisible tax… I think it’s a dumb problem, to correct supply with demand, and the result is obviously the cost benefit, because what happens when supply meets demand? You’re going to see a cost reduction — that’s just the nature of economics.” — Greg Osuri, 00:01:26

Context: The clip’s title thesis: cloud oligopoly rent flows through every online service people use.


Solve unused capacity and you don't need AWS

"Akash Network Deep Dive: The Unstoppable Cloud, Powered by Cosmos!" (Cryptocito)

“We’re attacking the problem at the core. The core of the problem which created Amazon Web Services was unused capacity. If we solve that problem, we don’t need Amazon Web Services… it’s not going to be a single day or a single year journey.” — Greg Osuri, 00:16:22

Context: His thesis that Akash inverts AWS’s own origin story (monetizing Amazon’s idle peak-load capacity); earlier he reports data centers he deployed to ran at ~15% utilization, “85% of the time they’re not used” (00:04:59).


The world is ready for a decentralized Amazon

"Akash Network Deep Dive: The Unstoppable Cloud, Powered by Cosmos!" (Cryptocito)

“We doubled the number of total workloads deployed in Akash in a month… we still have 100 [percent] growth month over month at month five… it’s very clear that the world is ready for a decentralized version of Amazon.” — Greg Osuri, 00:17:05

Context: Early mainnet growth stats, contrasted with a “great” benchmark of 22% month-two growth.


Trillion-dollar cloud market by 2025, 81% held by four companies

Everyone Pays an Amazon Tax with Greg Osuri, Founder of Akash Network (Orchid Labs)

“Right now about 81 percent of market share is controlled by four companies, and it is about to be a trillion dollar market by 2025. It is a massive market, and if you think about cloud, everything today goes on the cloud — it is a fabric for today’s society.” — Greg Osuri, 00:00:43

Context: Quantified market forecast paired with his concentration statistic; foundation of the rent-seeking argument.


85-95% of data-center compute sits unused

Akash Network Is The Airbnb Of Cloud (TFiR)

“While the demand is there, we have enormous supply that’s being unlocked in data centers, because turns out 85 to 95 percent of the compute in data centers remain unused… you normally plan for peak and you always want to keep peak capacity.” — Greg Osuri, 00:00:46

Context: The founding thesis: Akash unlocks locked-up capacity into “the first open source, peer-to-peer, permissionless, self-sovereign cloud” priced by the market instead of institutions.


A mature Akash catalog will out-scale Amazon's 250 services

Akash Network Is The Airbnb Of Cloud (TFiR)

“Amazon right now has about 250 services, but when Akash is fully mature and, you know, decentralized — where anybody can add any service that they want and there’s a way to verify these — I don’t see how we can[‘t] have a much more extensive catalog than any other cloud providers.” — Greg Osuri, 00:19:55

Context: After describing plans for decentralized managed services (e.g., MySQL run by low-latency providers in-datacenter); the bracketed correction reflects an apparent caption garble. Growth at the time: ~10 new active apps/day, ~150 deploys/day.


Cloud is a public utility that must be taken back from oligopolies

Akash Network Is The Airbnb Of Cloud (TFiR)

“If we look at cloud as a public infrastructure, a public utility, we don’t have anywhere close the level of scrutiny the other public utilities like power or water has… when we recognize cloud is important, it’s important to take back the cloud from oligopolies, for lack of a better word.” — Greg Osuri, 00:14:13

Context: Citing Cloudflare/Amazon outages caused by unscrutinized misconfigurations; decentralization brings the public oversight a mission-critical utility deserves. Quote spans into 00:14:58. An early energy-utility analogy for compute.


Data gravity is a pricing problem, not a technological one

Akash Network Is The Airbnb Of Cloud (TFiR)

“The data gravity problem is not a technological problem, it is a pricing problem… the root of that problem is the egress costs… they want your data to stay in the cloud.” — Greg Osuri, 00:08:38

Context: On vendor lock-in; says Akash is designing a bandwidth marketplace with Equinix (its largest provider at the time) to price data transfer equitably (00:09:19).


Deplatforming is being weaponized by state actors

Akash Network Is The Airbnb Of Cloud (TFiR)

“What’s crazy is deplatforming is getting weaponized by state actors against legitimate companies… so people are using Akash as a disaster recovery mechanism.” — Greg Osuri, 00:22:03

Context: Second user cohort after the Parler takedown era — “the fact that there is a platform that has so much power is disturbing” (00:21:22) — with Akash as the censorship-resistant backup to primary AWS deployments.


Saving money on cloud is a fallacy — repatriation is real

Akash Network Is The Airbnb Of Cloud (TFiR)

“The whole thing about you saving money on a cloud is a fallacy… it’s a trend now, there’s a repatriation of data centers: any high-growth Silicon Valley company like Dropbox or Snapchat, they’re actually moving back to data centers for a reason.” — Greg Osuri, 00:06:28

Context: Responding to the host’s point that you need to be a mathematician to compute cloud pricing; Akash’s answer is monetizing underutilized capacity with “sovereignty on both your performance as well as your cost.”


Akash workloads doubling every month

Unstoppable Cloud and Building a Decentralized Web with Greg Osuri (The Index Podcast)

“It’s growing really fast, and it’s still doubling every month if you think about it, the amount of workloads, active applications being deployed. And I’m so proud that we actually now have a fully, a good viable alternative to Google, Microsoft, and Amazon that is fully open.” — Greg Osuri, 00:07:08

Context: Growth snapshot from August 2021 — ~150 active applications, 100–150 new deploys/day, price ~1/3 of AWS (stated at [00:01:33]).


Censorship by the few is censorship on the many

Unstoppable Cloud and Building a Decentralized Web with Greg Osuri (The Index Podcast)

“Censorship by the few is censorship on the many, essentially, right? So, censorship in general is a terrible thing. I don’t want anybody to have the ability to silence somebody.” — Greg Osuri, 00:32:37

Context: Asked how Akash prevents abuse of censorship resistance; he distinguishes censorship from moderation — providers may refuse to serve, but “as long as there’s one provider that’s willing to serve you, you’ll be served” ([00:33:59]).


Decentralized systems build superior UX

Unstoppable Cloud and Building a Decentralized Web with Greg Osuri (The Index Podcast)

“Decentralized systems, when they embrace decentralization, end up building far superior user experiences than a centralized system. For example, Uniswap… It takes 30, 40 steps by the time you can swap an asset [on Coinbase]. Versus you come to Uniswap, you click, click, click, and you swap.” — Greg Osuri, 00:18:54

Context: Contrarian UX thesis — permissionless state systems allow radically simpler interfaces; he extends it to “similar experiences for something like Cloud, where you can deploy permissionlessly by using your Keplr wallet” at [00:20:20].


Four companies control the cloud — and that is scary

Unstoppable Cloud and Building a Decentralized Web with Greg Osuri (The Index Podcast)

“You actually end up spending a lot more money on the cloud than running a data center. And these cloud companies now have 81% market share. So if you imagine cloud is a fabric of society… Majority of this cloud is actually controlled by four companies. And that very thought is scary.” — Greg Osuri, 00:05:30

Context: The Akash origin story — enormous idle supply meets a cloud market “about to be a trillion dollar market,” with AWS alone valued over $500B.


Private deplatforming power is intolerable

Unstoppable Cloud and Building a Decentralized Web with Greg Osuri (The Index Podcast)

“The fact that there are platforms that can silence you that are not government-controlled, but private-controlled… considering that these Amazons and Googles and Microsoft can shut you down if they don’t like you… That much power in few is not something that we can sleep with.” — Greg Osuri, 00:09:24

Context: Describing his second user cohort — anti-deplatforming deployments; he pairs it with the full decentralized stack (Handshake DNS + Akash + Sia Skynet) at [00:10:09].


Akash frees the cloud like the internet was freed from AOL

Akash Network’s (CRYPTO:AKT) ($AKT) interview with Mr. Greg Osuri CEO. (New to The Street TV)

“Permissionless systems are generally known to scale at a global scale — internet being a big example. When internet was broken from the jails of AOL, we really saw the penetration that it has today, and similarly Akash is doing that to cloud computing.” — Greg Osuri, 00:07:20

Context: After noting ~a billion people lack credit cards to access hyperscaler clouds, including a kid in Africa whose first cloud experience was Akash.


Every layer of the cloud stack is decentralizing

Akash Network’s (CRYPTO:AKT) ($AKT) interview with Mr. Greg Osuri CEO. (New to The Street TV)

“Every stack of the cloud today is getting decentralized, all the way from bandwidth allocation — you have networks like Sentinel or Helium or Orchid… discovery layers such as Handshake or ENS domains that are decentralizing DNS… compute layer, Akash is the player there… storage layer, we have Filecoin and Skynet and Arweave… every element of the stack is decentralizing, and Akash, being the compute layer, is the glue that brings all these decentralized networks in a form that’s easily available to the developer.” — Greg Osuri, 00:12:26

Context: His web3 infrastructure thesis: web3 fixes web1’s inefficiency and removes web2’s rent-seeking middlemen.


Users pay an invisible Amazon tax that web3 will lower

Akash Network’s (CRYPTO:AKT) ($AKT) interview with Mr. Greg Osuri CEO. (New to The Street TV)

“Everyday users are actually paying invisible Amazon tax, if you know it or not, and that will be lowered without even knowing what happened. So web3 is an infrastructure layer that’s optimizing and building a more equitable cloud.” — Greg Osuri, 00:13:52

Context: Predicting web3 infrastructure savings will invisibly pass through to consumers of services like Netflix.


What Bitcoin did to money, Akash is doing to cloud computing

Akash Network’s (CRYPTO:AKT) ($AKT) interview with Mr. Greg Osuri CEO. (New to The Street TV)

“Decentralization generally means a more equitable — it is the great equalizer… if you look at what’s happening with Bitcoin or Ethereum… similarly, what Bitcoin did to money, I think Akash is doing that to cloud computing.” — Greg Osuri, 00:10:12

Context: His recurring one-line thesis, offered while explaining decentralization as repurposing existing infrastructure vs. expansion.


A paradigm shift from centralized to decentralized architectures

How do data scientist get into blockchain? Greg Osuri - The Data Scientist Show #007 (The Data Scientist Show)

“You have self-driving vehicles — they’ll start using Akash. You have all these amazing things to look forward for. Overall I think we are in a big paradigm shift from centralized architectures to decentralized architectures from a technological landscape, and I’m just super excited to be contributing and leading these efforts.” — Greg Osuri, 01:17:37

Context: Closing remarks; his recurring macro thesis that decentralized architectures displace centralized ones, with autonomous vehicles as future Akash users.


Akash today is like AWS in 2008 — adoption takes time

How do data scientist get into blockchain? Greg Osuri - The Data Scientist Show #007 (The Data Scientist Show)

“Using Amazon back then was like no way in hell — if you use Amazon you get fired. Now if you do not use Amazon you get fired, I suppose. You’re seeing similar patterns with Akash. Not too many people know about Akash, but it’s one of those things that takes time for technologies to really show their power.” — Greg Osuri, 00:22:50

Context: Recalling that Kaiser Permanente banned S3 when he built their first internal cloud in 2008; frames Akash as following the same skepticism-to-default adoption arc as AWS.


Censorship resistance: Parler would run fine on Akash

Blockchain & Cloud Computing - Akash Network W/ Greg Osuri (Arcane Bear)

“Akash Network is censorship resistant… no one can take down the workload as long as there’s a provider willing to serve the workload… Parler would run very efficiently on Akash, because as long as it’s half the country that’s agreeing with them.” — Greg Osuri, 00:25:33

Context: On the Parler/AWS deplatforming; he adds that permissionless communities self-moderate genuinely abhorrent content and providers can see what they host.


Deplatforming is weaponized — Akash as disaster recovery

Blockchain & Cloud Computing - Akash Network W/ Greg Osuri (Arcane Bear)

“Deplatforming is being weaponized by state actors, against corporate transfers, against individuals that disagree with them, and we need a solution… you can run your primary site on Amazon… in case you get deplatformed you can switch over to Akash in an instant, instantly, without losing any of your customers.” — Greg Osuri, 00:31:59

Context: Prompted by Anthony Pompliano’s YouTube takedown the day before; disaster-recovery/failover pitched as a real adoption vector.


Four companies control 80% of a market headed to $1 trillion

Blockchain & Cloud Computing - Akash Network W/ Greg Osuri (Arcane Bear)

“Four companies essentially control 80 percent of the marketplace, and in the process they became rent seekers… an industry today is about 390 billion dollars and which is projected to be a trillion dollars very soon — majority of that is controlled by four companies.” — Greg Osuri, 00:07:06

Context: His web 2.0 critique: over-capitalization made clouds “the center fabric of our society” run by rent-seeking oligopolists (“MAGA”: Microsoft, Alibaba, Google, Amazon).


80% of cloud controlled by four companies

Akash Network CEO on competing with Amazon AWS, decentralized cloud computing, and free speech (Yahoo Finance)

“Internet was created as a decentralized network to fight against centralization… but over time it got increasingly centralized, and now the landscape looks like 80 percent or so of the cloud computing, which is a backbone for the internet, is controlled by four companies — Amazon being the biggest one.” — Greg Osuri, 00:01:29

Context: His recurring cloud-centralization thesis, answering how sites get de-platformed.


Early growth: 500 apps, 30k deployments, doubling monthly

Akash Network CEO on competing with Amazon AWS, decentralized cloud computing, and free speech (Yahoo Finance)

“We launched the platform about six months ago and so far we have about 500 applications live… over 30,000 deployments that happened over the last six months… until last month we were actually doubling in terms of workloads.” — Greg Osuri, 00:03:36

Context: Quantified adoption snapshot ~6 months after the Akash marketplace launch.


Open source makes Akash as secure as or more secure than AWS

Akash Network CEO on competing with Amazon AWS, decentralized cloud computing, and free speech (Yahoo Finance)

“It’s probably more secure than your traditional Amazon because Akash being open source, all our source code is actually open for a public audit, so it offers similar levels if not better levels of security guarantees Amazon’s or the big cloud offers.” — Greg Osuri, 00:03:36

Context: Rebutting the host’s security concern about choosing Akash over AWS.


Provider self-moderation instead of central takedowns

Akash Network CEO on competing with Amazon AWS, decentralized cloud computing, and free speech (Yahoo Finance)

“The way Akash network operates is self-moderation. That means providers on the Akash network choose who they want to serve… as long as there’s a provider that’s willing to serve Parler, Parler can run on Akash. But if there are no providers… that website won’t be served on Akash platform.” — Greg Osuri, 00:05:45

Context: His standard answer to how a censorship-resistant network handles harmful content.


80% of cloud controlled by four companies

Akash Network CEO, Greg Osuri on Yahoo Finance (Akash Network)

“Internet was created as a decentralized network to fight against centralization… but over time it got increasingly centralized, and now the landscape looks like 80 or so of the cloud computing, which is a backbone for the internet, is controlled by four companies — Amazon being the biggest one.” — Greg Osuri, 00:02:10

Context: Answering how sites get de-platformed; his recurring centralization thesis.


Early growth: 500 apps, 30k deployments, doubling monthly

Akash Network CEO, Greg Osuri on Yahoo Finance (Akash Network)

“We launched the platform about six months ago and so far we have about 500 applications live… over 30,000 deployments that happened over the last six months… until last month we were actually doubling in terms of workloads, applications coming online to Akash network.” — Greg Osuri, 00:04:21

Context: Quantified adoption snapshot ~6 months after mainnet marketplace launch (mid-2021 baseline).


Open source makes Akash as secure as or more secure than AWS

Akash Network CEO, Greg Osuri on Yahoo Finance (Akash Network)

“It’s probably more secure than your traditional Amazon because Akash being open source, all our source code is actually open for a public audit, so it offers similar levels if not better levels of security guarantees Amazon’s or the big cloud offers.” — Greg Osuri, 00:04:21

Context: Rebutting host’s security concern about using Akash over AWS.


Provider self-moderation instead of central takedowns

Akash Network CEO, Greg Osuri on Yahoo Finance (Akash Network)

“The way Akash network operates is self-moderation. That means providers on the Akash network choose who they want to serve… as long as there’s a provider that’s willing to serve Parler, Parler can run on Akash. But if there are no providers… that website won’t be served on Akash.” — Greg Osuri, 00:06:30

Context: His standard answer to “what if something really bad runs on Akash and can’t be taken down.”


Open systems always beat corporate land-grabs (Microsoft/AOL analogy)

Priv8 Virtual Privacy Summit 2021: Featuring Glenn Greenwald, Dr. Steven Waterhouse, and more! (Orchid Labs)

“This is not the first time a centralized company tried to put claim to an open system. Microsoft in the 90s tried to put claim to internet, AOL tried to put claim by actually developing tools for the internet, but ultimately an open system won. So the composability of open system will ultimately be the winner.” — Unknown, 04:56:12

Context: On Meta/Microsoft claiming the metaverse; his recurring open-systems-win thesis. Quote begins mid-block after the moderator (“I’m actually very optimistic…”); undiarized captions, but the argument and phrasing match Greg’s earlier remarks on the panel ([04:54:48]: “I think Mark introduced the metaverse — best thing that happened to metaverse in a weird way”).


The cloud is controlled by four players — few controlling the many

Priv8 Virtual Privacy Summit 2021: Featuring Glenn Greenwald, Dr. Steven Waterhouse, and more! (Orchid Labs)

“Today most the cloud computing market is really controlled by four players… the effects of which go beyond just cost… it really comes down to… what happens when few really control the many.” — Greg Osuri, 04:15:31

Context: Panel self-introduction; he says he built Akash to “democratize cloud computing” and transfer power from entrenched infrastructure players to something “more democratized and owned by the user” ([04:14:45]).


Four players control the cloud — what happens when few control the many

Priv8 Virtual Summit: Greg Osuri, Tor Bair, Ari Paul's Take on the Decentralized Future (Orchid Labs)

“Today most of the cloud computing market is really controlled by four players… the effects of which go beyond just cost. It really comes down to the whole topic of this conversation: what happens when few really control the many.” — Greg Osuri, 00:02:11

Context: His self-introduction — Akash as transferring power from “oligopolized infrastructure players” to users; quote begins in the [00:01:29] block.


Open systems will beat centralized claims on the metaverse

Priv8 Virtual Summit: Greg Osuri, Tor Bair, Ari Paul's Take on the Decentralized Future (Orchid Labs)

“I’m actually optimistic. It’s not the first time a centralized company tried to put claim to an open system — Microsoft in the 90s tried to put claim to the internet, AOL tried to put claim by actually developing tools for the internet, but ultimately an open system won. So the composability of an open system will ultimately be the winner.” — Greg Osuri, 00:43:19

Context: Responding to worries about Meta/Microsoft capturing the metaverse; a 2021 instance of his recurring open-beats-closed prediction.


Enterprises will adopt decentralized cloud in five to six years

DeFi Decoded - Greg Osuri, CEO of Akash on why DeFi needs a Decentralized Internet to succeed (Ninepoint Partners)

“So in five to six years’ time, you’re going to see some of these enterprises adopt… we’re seeing champions now in different companies, we’re seeing enterprise adoption, but we may not necessarily focus on it.” — Greg Osuri, 00:35:27

Context: Direct answer to Tapscott’s request for a prediction on whether giants like Walmart or Goldman Sachs will use decentralized cloud; Greg invokes the Innovator’s Dilemma and insists developer-first adoption (like early AWS) must come first.


Over 85% of data-center capacity is unused; any data center can become a cloud provider

DeFi Decoded - Greg Osuri, CEO of Akash on why DeFi needs a Decentralized Internet to succeed (Ninepoint Partners)

“As we’re deploying Kubernetes into different data centers, we notice enormous amounts of capacity in these data centers is not used — so when north of 85% of capacity is not used — and we built the technology to enable any data center to become a cloud provider.” — Greg Osuri, 00:07:02

Context: Greg’s founding story for Akash: the underutilization insight from deploying Kubernetes at Overclock Labs (2015-2017). He also claims “there are 8.4 million data centers in the world” and Amazon’s edge is only ease of access.


The future is sovereign chains in a multi-chain world

DeFi Decoded - Greg Osuri, CEO of Akash on why DeFi needs a Decentralized Internet to succeed (Ninepoint Partners)

“I believe in a multi-chain future, absolutely, there’s no doubt… like we saw with Web2 too — you build on a shared platform like Firebase, but the time when you have to really scale up, you go sovereign. I think the future is sovereignty.” — Greg Osuri, 00:22:37

Context: Explaining why Akash built a sovereign Cosmos/IBC chain instead of on Ethereum or Solana: control over scalability and economics (~1/4-cent gas vs. $70 Ethereum deployments).


Web2 concentrates power; Web3 distributes it

DeFi Decoded - Greg Osuri, CEO of Akash on why DeFi needs a Decentralized Internet to succeed (Ninepoint Partners)

“Web3 is taking a much more distributed, much more diverse attitude than Web2, because Web2 is all about concentration of power; Web3 is about distribution of power. And we’re seeing that very clearly with dWeb, and I’m really excited for that future.” — Greg Osuri, 00:28:19

Context: Summing up his tour of the composable dWeb stack (Akash compute + Filecoin/Sia/Arweave storage + Handshake DNS + Livepeer), with IBC eventually letting any token pay for any service.


Akash should become invisible infrastructure

Akash Office Hours : End of Year Product Update (Akash Network)

“[We want to] get to that framework-level integration so it becomes part of the workflow — so that Akash is just invisible in the background. You don’t even know you’re using Akash, but you’re using Akash. And there’s the beauty of permissionless systems: there’s no signup [or other] nonsense that you need to use.” — Greg Osuri, 00:40:46

Context: The usability roadmap: fix the CLI first, then one-click deploys from frameworks like Starport, Ruby on Rails, and Node.js.


Beating Amazon requires 300 services — via a decentralized marketplace by end of 2022

Akash Office Hours : End of Year Product Update (Akash Network)

“Two things need to happen for Akash to be over Amazon. First… Amazon has 300 backend services that form this phenomenal service fabric… Akash only has one service, which is container hosting, so we need to come to 300 services in order to have parity with Amazon. And we’re not going to do that as a single organization… we’re going to enable a services marketplace by end of 2022.” — Greg Osuri, 00:27:49

Context: Answering the recurring “when Akash over Amazon” question; the second requirement is usability designed for modern developers as “Amazon is increasingly getting distant away from a developer.” An early statement of the services-marketplace vision he still pitched in 2025.


2022 is the year the world shows what it can do with Akash

Updates From the Lab 1/5/22 W/Akash (Osmosis)

“2021 was a preview for Akash — we showed the world what Akash can do. I think 2022 is when the world shows what they can do with Akash. I’m so excited, and this is the year which defines us, really.” — Greg Osuri, 00:40:22

Context: Concluding the Kubernetes analogy — taking web2 lessons and “using the catalyst of decentralization as a hyper growth engine.”


85% of data-center capacity is unused; Equinix is the largest supplier

Updates From the Lab 1/5/22 W/Akash (Osmosis)

“Equinix happened to be the biggest supplier… they offloaded, I believe, like ten percent of their underutilized supply… it’s about 85 [percent] or so [of] computing capacity in data centers [that] are not used. For them, if they don’t offload the capacity, it’s useless… essentially Akash becomes like the spot market for the globe.” — Greg Osuri, 00:53:12

Context: Answering a Regen Network member’s question about who the ~79 providers are and why Akash can undercut AWS. Spot-market line at 00:54:40.


Akash costs one-third to one-fifth of Amazon

Updates From the Lab 1/5/22 W/Akash (Osmosis)

“The outcome is a significantly lower price, right? So Akash right now offers anywhere from one third to one fifth the cost, the price point Amazon offers.” — Greg Osuri, 00:19:05

Context: His intro description of Akash as a permissionless reverse-auction marketplace for unused cloud-grade compute cycles.


Services marketplace by end of 2022 — "going on the offensive"

Updates From the Lab 1/5/22 W/Akash (Osmosis)

“One of the things we have on the roadmap is our services marketplace by end of 2022 — that’s really when we start going on the offensive… Amazon has about 300 services, right, and Akash is building a decentralized marketplace [of] services, and we assume a lot of these services are going to have their own tokens.” — Greg Osuri, 00:44:42

Context: Databases, load balancers, caching servers as tokenized services; he frames Osmosis as “a critical payment rail for the web3 infrastructure” enabling swaps between service tokens. Second sentence lands at 00:46:08.


The largest migration of web2 to web3 is underway

Updates From the Lab 1/5/22 W/Akash (Osmosis)

“If you think about what’s happening today, we’re seeing the largest migration of web2 to web3, hands down.” — Greg Osuri, 00:38:13

Context: Citing a prominent open-source developer choosing to build a fully open-source, user-owned system natively on Akash instead of a closed-source web2 product.


~85% of data-center compute sits unused

Akash (40%+ staking APY) is taking over the $Bn Cloud Computing Market by AWS? Greg Osuri EXCLUSIVE (Digital Asset Yield Summit)

“We notice most compute in data centers, around 85 or so [percent] of compute in data centers, is not used. So that’s when the idea for a marketplace came about.” — Greg Osuri, 00:24:16

Context: Recurring founding claim behind Akash’s supply thesis; he repeats it later citing “about 8.2 million data centers in the world” with ~85% unused ([01:29:59]).


Cloud is headed to a trillion-dollar market, devices beyond that

Akash (40%+ staking APY) is taking over the $Bn Cloud Computing Market by AWS? Greg Osuri EXCLUSIVE (Digital Asset Yield Summit)

“Cloud business [is] about 700, you know, a billion dollars, right? It’s going to be a trillion-dollar market. But beyond that, the devices and the compute market — it’s just humongous.” — Greg Osuri, 01:07:05

Context: Explaining why AKT’s fixed supply should derive real value from usage demand rather than staking rewards; calls the $300M market cap “a blip” (explicitly notes this is not financial advice).


Current cloud is overpriced; capabilities, not cost, win long-term

Akash (40%+ staking APY) is taking over the $Bn Cloud Computing Market by AWS? Greg Osuri EXCLUSIVE (Digital Asset Yield Summit)

“We think current cloud is overpriced… on the long run it’s going to be the capabilities that’s going to differentiate Akash, not cost… Amazon has 60% margins. That’s crazy… Snowflake is a public company — 50% of their revenue, that profit, goes to Amazon.” — Greg Osuri, 01:39:55

Context: Answering how Akash stays cheaper as it scales: open price competition finds the “right price,” and cloud margins must compress “no matter what.”


Decentralized managed-services marketplace by end of 2022

Akash (40%+ staking APY) is taking over the $Bn Cloud Computing Market by AWS? Greg Osuri EXCLUSIVE (Digital Asset Yield Summit)

“We’re hoping to see something in by end of 2022. So 2023 I think is going to be… if we have a functioning system like this, it’s a game changer. It changes everything.” — Greg Osuri, 00:39:19

Context: A marketplace where open-source creators (Percona for MySQL, Redis Labs) provide managed services with reputation systems and royalties for creators; being designed with bash creator Brian Fox.


"We will be better than Amazon"

Akash (40%+ staking APY) is taking over the $Bn Cloud Computing Market by AWS? Greg Osuri EXCLUSIVE (Digital Asset Yield Summit)

“We can’t be just yelling away ‘better than Amazon.’ We will be better than Amazon as long as we keep building in the ethos that it’s a market, it’s open marketplace for anyone to participate.” — Greg Osuri, 00:33:34

Context: Answering what it takes to win share from AWS/GCP; first step is capability-matching Amazon’s ~300 managed services, then composing them via SDL plus permissionlessness.


Browsers blocking URLs is a nearer threat than regulators

Splitting ATOMs: How decentralized is Cosmos? ⚛️ (Cryptocito)

“I think we should be more worried about browsers blocking URLs. See what’s happening with Skynet, for example, Sia — Chrome blocks 80 percent of Sia URLs; most of Europe you can’t even access Sia URLs… in Gmail if you put a Sia link it will go away, Twitter will block it. I think you’re gonna see a lot of that come by.” — Greg Osuri, 00:46:25

Context: Responding to Zaki’s point that regulators will target frontends; Greg says Akash is moving away from reused TLDs toward unique addresses to resist this (00:47:08).


Decentralized apps must not run on Amazon

Splitting ATOMs: How decentralized is Cosmos? ⚛️ (Cryptocito)

“I truly believe all the applications — I mean especially decentralized applications — should run on a decentralized cloud. Just don’t run on Amazon… if a lot of nodes are running on Amazon, Amazon shutting you down could affect the security of the network, but rather if you have nodes running elsewhere, you’re okay.” — Greg Osuri, 00:42:47

Context: On validator infrastructure’s web2 dependency; adds Akash is “like 90% cheaper than Azure and like 80% cheaper than Amazon — so why wouldn’t you” (00:43:31).


Sovereign frontends launched from your wallet

Splitting ATOMs: How decentralized is Cosmos? ⚛️ (Cryptocito)

“There’s no reason why you shouldn’t launch an Osmosis UI directly from Keplr that’s fully dedicated to an RPC server that’s fully dedicated to you — maybe even a validator node if you want that level of sovereignty, a full-stack DeFi layer… you should be able to launch that directly from your Keplr wallet… the user ultimately is responsible for the UI and the company doesn’t operate [it] anyways, and Akash provides you the runtime for your container.” — Unknown, 00:47:51

Context: Introduced by the co-host as “some of the cool ideas that you [Greg] thought about”; a personal-infrastructure vision of users self-hosting their own frontends/nodes on Akash. Speaker: the setup line may be the co-host paraphrasing Greg’s idea; the elaboration is Greg’s.


85% of data-center compute sits unused

Akash Weekly: Chia (Akash Network)

“As we were deploying the solution to different customers that own data centers, we realized most compute in these data centers was not used — 85 [percent] or so — and that’s when the idea to create a spot market for compute came about.” — Greg Osuri, 00:04:24

Context: Origin story from his multi-cloud Kubernetes work at Overclock Labs (2015-2017); the founding statistic behind Akash’s marketplace thesis.


A decade-long fight for a trillion-dollar market

"Akashonomics: AKT Release Schedule" (Akash Network)

“Taking on the cloud, a trillion-dollar market, is not going to happen in a matter of a couple of years. It is a long endeavor… we’re going to be here for the next seven more years without even thinking about it.” — Greg Osuri, 00:42:43

Context: Responding to team-dumping FUD; he concedes Akash then lacked the UX and capabilities to win Amazon customers head-on but has “the core, which is decentralized, and the pricing.”


Akash as the de facto standard for web3 infrastructure

"Akashonomics: AKT Release Schedule" (Akash Network)

“Akash is becoming the de facto standard to run web3 infrastructure, and that’s happening with I think four out of the top 100 blockchains now, and that’s going to continue to happen.” — Greg Osuri, 00:59:04

Context: Citing Juno’s website, Osmosis RPC/API servers, and Stargaze validators running on Akash; also predicts Akash becomes “the data services provider for DAOs” since multisig-controlled deployments remove key-man risk.


Akash vs AWS is Airbnb vs Hilton

Akash Weekly: Introducing new VP of Engineering, Chanda Dharap (Akash Network)

“Akash has what, 600 active deployments, compared to a 500 billion dollar business which is Amazon Web Services — we’re not a threat yet… it’s going to be more of an Airbnb versus Hilton. Hilton did not deem Airbnb as a threat until they surpassed Hiltons and all the hotel chains’ market cap… similarly Amazons and Googles and Microsofts can’t offer what Akash can offer — their model is not set up that way.” — Greg Osuri, 00:41:35

Context: Answering how AWS/Azure view Akash; argues open compute marketplaces are structurally impossible for incumbents to replicate.


Complement big cloud now, exceed it eventually

Akash Weekly: Introducing new VP of Engineering, Chanda Dharap (Akash Network)

“Most likely we’re going to not evolve as that threat, but we’re going to evolve as a complement to the big cloud… eventually Akash will support everything Amazon supports and beyond that. Let the market decide.” — Greg Osuri, 00:43:01

Context: Follows the Airbnb analogy; also notes Google hired ~140 people for a blockchain team but “you can’t buy a community” — community is the competitive moat. Second sentence is at [00:43:43].


Everybody gets an RPC node — launched from your wallet

Akash Weekly: Chia (Akash Network)

“How much does Infura charge for running an Ethereum RPC node versus Akash charging? I think it’s exponentially lower. And ideally… wallets like MetaMask actually give you an ability to launch RPC nodes that’s dedicated for you, and that you know for sure is secure enough, on Akash directly from the wallet — that’s what I like to see… everybody gets an RPC node.” — Greg Osuri, 00:30:42

Context: Asked what he most wants from testnet participants; a composability vision for personal infrastructure — every wallet user running their own dedicated node.


Node services (Pocket/Infura/Alchemy) will run on Akash

Akash Weekly: Chia (Akash Network)

“A good pathway from here would be someone like Pocket Network, who specializes in running nodes, or even Infura or even Alchemy — Akash will become another provider… and if these platforms give the users a choice of running over centralized infrastructure versus decentralized, I think it will be a fantastic outcome.” — Greg Osuri, 00:15:48

Context: After reporting ~135 Avalanche and ~117 Terra nodes launched during Testnet 3 — framing node infrastructure as proof Akash can run production blockchain workloads.


One of the largest testnets in web3

"Akash Weekly: Testnet 3, Week 1 Challenges" (Akash Network)

“So far we’re able to attract over 8,000 participants. This is by far the biggest testnet for Akash, for sure, but I think it has to be one of the largest testnets in the web3 space itself.” — Greg Osuri, 00:07:04

Context: Contrasting Akash’s incentivized-testing discipline with Cosmos-ecosystem projects that “launch and fail… or worse, actually launch and have exploits.”


Only Akash can enable compute futures, because its data is open

Akash Weekly: Introducing new VP of Engineering, Chanda Dharap (Akash Network)

“We always talked about pricing, an open model enabling or unlocking futures for cloud… Akash is the only cloud network or platform that can do futures. You cannot do futures on Amazon because the data is not open. You need the data to be open, so we have an advantage.” — Greg Osuri, 00:16:31

Context: Greg’s answer on what he’ll focus on now that Chanda handles execution: revisiting Akash economics (designed ~2020) and pricing models impossible on centralized clouds.


Persistent storage enables a decentralized CDN on Akash

"Akash Weekly: Testnet 3, Week 1 Challenges" (Akash Network)

“Akash now, effectively by enabling persistent storage, you can actually build a very modern CDN on top of Akash — ideally decentralized, which will be really cool.” — Greg Osuri, 00:20:40

Context: Responding to OmniFlix asking about CDN use cases; he tempers it later — Cloudflare has ~100 points of presence, “we don’t have 100 different points yet, we will in the future.”


Self-hosted RPC as the answer to censorship

Akash Weekly: Chia (Akash Network)

“Now you have all kinds of censorship happening at the RPC level… RPC servers have to obey the law of the land wherever the servers are hosted… giving ecosystems’ users [the ability] to host their own RPC servers, of course in compliance with whatever country they’re operating from, is a great use case for Akash.” — Greg Osuri, 00:32:07

Context: On the MetaMask-is-censoring-you narrative — he argues MetaMask is just a client and the fix is users scheduling their own RPC servers in a country of their choice via Akash.


Storage networks, Solana validators, and full nodes will run on Akash

"Akash Weekly: Testnet 3, Week 1 Challenges" (Akash Network)

“You’ll be able to run a lot of the storage networks… running on top of Akash, and you’re going to see other validators like Solana, which have huge storage requirements, running on Akash. You’re going to see full nodes in the entire Cosmos ecosystem.” — Greg Osuri, 00:17:05

Context: On what persistent storage unlocks; he calls it the feature that “completes a big gap” with traditional cloud. Later predicts Lumen and Chia miners will “flood the market with cheap, high-performance storage” ([00:39:54]).


300-500 providers beats Amazon on geo-presence

Akash Weekly Spaces - June 15th (Akash Network)

“Once providers are profitable, that is something that can be used to attract other providers… that’ll massively grow the provider base. Once we have 300 to 500 providers, that is significantly higher geo presence than let’s say an Amazon.” — Greg Osuri, 00:32:08

Context: AMA answer on attracting big web2 companies: PMF and provider profitability first (~30 providers at the time), then distribution, “cost being 90% lower,” and “world’s first open cloud” as the USP.


Faster: low-latency delivery from 500,000 nodes worldwide

Akash Weekly Spaces - June 15th (Akash Network)

“Our selling points, our USPs, are better, cheaper, faster. How is it faster? Low latency delivery using 500,000 nodes across the world. That’s a very unique property for Akash. But before we can run, we gotta walk — we gotta show providers this is profitable.” — Greg Osuri, 00:34:15

Context: Long-term vision of a massively distributed edge-scale provider network as Akash’s structural advantage over hyperscaler regions.


AWS can cancel web3 with a stroke of a pen

Akash Weekly Spaces - June 22nd (Akash Network)

“Obviously AWS can cancel web three with a strike of a pen, right? So one man controls it, essentially.” — Greg Osuri, 00:32:40

Context: Responding to a community member’s “if all web3 runs on an AWS server, is it really web3?”; he adds Akash now requires its own validators to run on Akash and that ~52% of Osmosis nodes sit on traditional cloud.


The signal is clear: Akash is the future of web infrastructure

Akash Weekly Spaces - June 22nd (Akash Network)

“In a bear market, all the noise gets canceled, the signal gets amplified. And signal is very clear right now that [Akash] is going to be the future of web infrastructure.” — Greg Osuri, 00:34:02

Context: Closing his updates; he cites product-market-fit signals — users requesting features, real usage — and says the network is entering “real product management” stage.


85% of data center capacity unused; Akash 95% cheaper

Decentralized Cloud - DCENTRAL Austin 2022 (Dcentral Con)

“The idea behind [Akash] is to unlock unused capacity — about 85 or so [percent] of global data center capacity is unused. So by enabling those providers to offer their compute in an open, competitive way, we can solve efficiency problems, [and] the outcome is the cost — a cost about 95 percent lower than your traditional cloud. Like last night I was able to deploy an Osmosis validator for seven bucks a month.” — Greg Osuri, 00:02:52

Context: Panel introduction; adds “when the bear market hits and profitability for node operations is in question, AKT is your best friend.” Quote spans into the [00:03:36] block.


85% of data center capacity unused; Akash 95% cheaper

I Moderated A Panel at DCENTRAL Austin 2022 on Decentralized Cloud! (MineYour.₿iz Classic)

“The idea behind [Akash] is to unlock unused capacity — about 85 or so [percent] of global data center capacity is unused. So by enabling those providers to offer their compute in an open, competitive way, we can solve efficiency problems, [and] the outcome is the cost — a cost about 95 percent lower than your traditional cloud. Like last night I was able to deploy an Osmosis validator for seven bucks a month.” — Greg Osuri, 00:02:52

Context: His panel introduction; adds “when the bear market hits and profitability for node operations is in question, AKT is your best friend.” Quote spans into the [00:03:36] block.


Amazon's redundancy ends at the data center

Decentralized Cloud - DCENTRAL Austin 2022 (Dcentral Con)

“On-chain computing is hyper redundant — that’s why it’s expensive, that also is why it’s secure. Off-chain compute, the redundancy depends on the scheduling algorithm. Amazon is not redundant… if an Amazon data center goes down, it goes down — it’s redundant within the scope of the data center… the challenge is to design something that allows for cross-site, or maximum possible, redundancy while allowing for performance.” — Greg Osuri, 00:05:01

Context: Answering “why not just deploy on AWS”; he adds he deployed a validator “within five minutes from 3,000 feet using clicks” [00:05:44].


Amazon's redundancy ends at the data center

I Moderated A Panel at DCENTRAL Austin 2022 on Decentralized Cloud! (MineYour.₿iz Classic)

“On-chain computing is hyper redundant — that’s why it’s expensive, that also is why it’s secure. Off-chain compute, the redundancy depends on the scheduling algorithm. Amazon is not redundant… if an Amazon data center goes down, it goes down — it’s redundant within the scope of the data center… the challenge is to design something that allows for cross-site, or maximum possible, redundancy while allowing for performance.” — Greg Osuri, 00:05:01

Context: Answering why users shouldn’t just deploy on AWS; he adds he deployed a validator “within five minutes from 3,000 feet using clicks — we’re getting there very very quickly” [00:05:44].


Amazon is crypto's middleman — 40% of Ethereum nodes

Akash Weekly Spaces - June 29th (Akash Network)

“Over 40 [percent] of Ethereum nodes currently run on Amazon alone, and over 54 [percent] of the nodes are running on the big cloud… that exposes the vulnerability… is Ethereum really as decentralized as it claims to be when one company has over forty percent of the network processing power?” — Greg Osuri, 00:07:55

Context: Citing ethernodes.org real-time data; he follows up that “the whole notion of not having a middleman — but actually we’re having middlemen in crypto, and that middleman happens to be Amazon.”


Web3's dependence on Web2 must end

Akash Weekly Spaces - June 29th (Akash Network)

“The whole purpose of building decentralized cloud infrastructure is to build a reliable infrastructure… the fact that a lot of dapps and nodes and our entire infrastructure depends on Web2 is not acceptable in my book, and the Cloudflare outage clearly showed that our decentralized infrastructure is very, very reliant on Web2 — and that needs to change now.” — Greg Osuri, 00:06:29

Context: Reacting to the June 2022 Cloudflare outage that took down much of the crypto ecosystem; he adds “we have a lot of work to do and we have to do it quicker.”


300-500 providers would out-distribute Amazon

Ask Akash: Why not focus on Web 2 adoption for Akash? (Akash Network)

“Once providers are profitable, that is something that can be used to attract other providers, because there’s actual money to be made here… that will enable provider incentives, that’ll massively grow [the] provider [base]. Once we have 300 to 500 providers, that is significantly higher geo-presence than, let’s say, an Amazon.” — Greg Osuri, 00:03:34

Context: July 2022 growth thesis: provider profitability compounds into distribution that hyperscalers can’t match.


60% of Ethereum nodes run on AWS — the foundation must decentralize

WCEF 2022: Infrastructure for the New Internet - Web3 (World Crypto Economic Forum (WCEF))

“Somewhere around sixty percent of Eth nodes are running on Amazon Web Services. So if Amazon tries to pull the plug, it’s gone… The foundation needs to be decentralized, because the companies like Amazon and Google, they have no choice except to comply.” — Greg Osuri, 00:21:32

Context: On New York’s just-passed mining ban (“banning SHA-256… a horrible law, I’m pretty sure it’s going to reverse in a few years”) and hostile regulatory action as a systemic threat.


Airbnb for cloud compute — 95% of data center capacity sits unused

WCEF 2022: Infrastructure for the New Internet - Web3 (World Crypto Economic Forum (WCEF))

“Akash was founded about four years ago with an idea of creating an Airbnb for cloud compute. What that means is there’s enormous amount of cloud-grade compute sitting in data centers across the world that is underutilized — about 95% of which is not used — and on the other hand we have this crazy demand for cloud with prices only going up.” — Greg Osuri, 00:07:52

Context: His panel introduction; adds Akash is “one of the fastest growing clouds in the world” at 25M total gigabyte-hours, 3.5M in the last month.


The rise of DWS — decentralized web services

WCEF 2022: Infrastructure for the New Internet - Web3 (World Crypto Economic Forum (WCEF))

“We’re seeing the rise of DWS, or decentralized web services, I call it… Akash definitely is the small global spot market that never existed before. So I think each of the platforms are finding the niches and specializing to those niches, and I think we’re going to continue seeing that solidification, product market fits.” — Greg Osuri, 00:10:01

Context: Trend question; he groups Akash, Livepeer, Arweave, Graph, Storj, Filecoin, and Sia under the DWS umbrella, each specializing (e.g., Arweave in permanent storage).


The USP: world's first open cloud, 90% cheaper, distribution Amazon can't give

Ask Akash: Why not focus on Web 2 adoption for Akash? (Akash Network)

“You have heavy distribution from a provider base and you have amazing user experience from a user base, along with product market fit, along with cost being 90 [percent] lower. Now that’s the unique selling proposition that Amazon can’t give — and all this in an open source [way]. Our USP is world’s first open cloud… our USP is better, cheaper, faster. How is it faster? Low latency delivery using 500,000 nodes across the world.” — Greg Osuri, 00:04:17

Context: Answering how Akash could eventually attract large Web2 companies; the 500,000-node figure is aspirational scale, not then-current (Akash had ~30 providers at the time).


Akash will be the fastest network to deliver the cloud — a thousand regions vs. Amazon's 16-17

Akash Weekly - July 6th 2022 (Akash Network)

“Over time that’s going to go to a point that Akash will be the fastest network to deliver the cloud on. Imagine having a thousand endpoints, a thousand, like, regions, compared to Amazon’s 16 or 17 regions… the old clouds of the world are limited to data centers because they have to go build their own data centers. Akash — we don’t build a data center, people with data centers come to us. It’s a fundamentally different model.” — Greg Osuri, 00:33:35

Context: From ~39 providers at the time, he forecasts provider count accelerating into massive node distribution enabling high-performance, low-latency compute.


Mainstream web2 is a trillion-dollar opportunity for Akash

Akash Weekly - July 6th 2022 (Akash Network)

“With that high performance and low cost data processing we can build the next layer… the services layer, and then you’ll be able to attract the mainstream web2 folks, and that’s a trillion dollar market opportunity. But we’re going to be diligent, we’re going to be phased, we’re going to be disciplined.” — Greg Osuri, 00:34:19

Context: Final phase of the adoption plan; he notes Overclock has runway (“next four years or so”) to execute through the bear market.


Mass adoption will come from abstractions where users don't know they're using Akash

Akash Weekly - July 6th 2022 (Akash Network)

“I think the real mass adoption is going to come from higher level abstractions where the users don’t even know they’re using Akash — like DAO DAO, for example… dial users can get messaging and forums and all the tools DAOs need in a single button, paid using Juno tokens.” — Greg Osuri, 00:49:55

Context: AMA answer on how Akash addresses cloud complexity for non-technical users; invisible infrastructure as the path to mass adoption.


5,000 providers in 5,000 locations — low-latency compute impossible on the cloud

Akash Weekly - July 13th 2022 (Akash Network)

“We have about [4]50 providers. So you can imagine a world where we are gonna have 5,000 providers — that means 5,000 different locations, right? Not very far away… high performance computing that is globally distributed can be delivered… within a very low network latency. And this is a very unique offering, only off Akash… this is impossible on the cloud.” — Greg Osuri, 00:52:04

Context: Edge/geo-distribution thesis: provider growth turns Akash into a globally distributed HPC fabric centralized clouds can’t match. Caption says “50 providers” here but Greg reported ~450 active providers earlier at 00:29:14.


Akash marketplace utilization ~30% vs Filecoin's 0.7%

Akash Weekly - July 13th 2022 (Akash Network)

“We saw about close to 29, 30% of [provider] liquidity just for our memory, which is fairly high… if you relate to other players in the market — Filecoin has storage, their liquidity is about 0.7%. So to give you comparison, a 30% usage rate is a very [good] indication of a very healthy market.” — Greg Osuri, 00:30:40

Context: Weekly metrics update framing marketplace health as liquidity/fill rate (“if I bring 10 apples… I want to walk away with zero apples”); follows all-time highs of 1,000 AKT daily spend and 10 TB active memory.


A thousand cloud-grade data centers at 90% cheaper than Amazon

Akash Weekly - July 20th (Akash Network)

“A thousand cloud grade data centers all available for you to use at 90 [percent] cheaper than Amazon — and that’s something that we’re building towards.” — Greg Osuri, 00:16:38

Context: Completing the “innovation beyond imagination” riff; the “low latency, high performance applications… the web2 world has never heard of.”


More regions than Amazon already; thousands of regions coming

Akash Weekly - July 20th (Akash Network)

“Close to 50 providers across the world — that’s 50 regions, way more than Amazon. Amazon I think has only 20 or so regions. So we are a lot more distributed now and we’re going to be a lot more distributed. Imagine when we have these thousands of regions, imagine the kind of applications that you can build. I call that innovation beyond imagination.” — Greg Osuri, 00:15:56

Context: Prompted by Praetor’s new global provider map; his recurring low-latency/edge thesis about massive geographic distribution beating hyperscalers.


Akash as the fastest cloud, a thousand regions vs Amazon's 17

The Akash Adoption Plan (Akash Network)

“Over time that’s going to go to a point that Akash would be the fastest network to deliver the cloud. Imagine having a thousand endpoints, a thousand regions, compared to Amazon 16 or 17… Amazon and the old clouds of the world are limited to data centers because remember they have to go build their own data centers. Akash, we don’t build a data center — people with data centers come to us. It’s a fundamentally different model.” — Greg Osuri, 00:04:13

Context: Quantified contrast with hyperscalers; at the time Akash had 39 providers, which he predicted would accelerate into “enormous node distribution.”


Compute must move to where the user is

The Akash Adoption Plan (Akash Network)

“How do you serve that data in a high performance way, in a low latency way? You do that by reducing the network latency between the end user and the server, and that means you need to go where the user is. That means the servers need to be in the same location of the user.” — Greg Osuri, 00:03:31

Context: Phase three of the plan — the performance problem after solving cost; an early version of his recurring edge/local-compute thesis.


Trillion-dollar mainstream opportunity

The Akash Adoption Plan (Akash Network)

“With that high performance and low cost data processing you can build the next layer… the services layer, and then you’ll be able to attract the mainstream web2 folks — and that’s a trillion dollar market opportunity. But we’re going to be diligent, we’re going to be phased, we’re going to be disciplined.” — Greg Osuri, 00:04:55

Context: Closing the phased roadmap; sizing the endgame market once cost and performance problems are solved.


DAOs will deploy cloud workloads directly via smart contracts

Akash Weekly - July 27th 2022 (Akash Network)

“By enabling interchain accounts, an application on Juno can deploy directly on Akash using a Juno account and not an Akash account… now you’re going to see DAOs, members of DAOs, are directly deploying onto Akash using a smart contract from Juno, which is nothing like we ever saw in Ethereum. These are the kind of use cases that are going to showcase Cosmos and the power of Cosmos to the rest of the ecosystem.” — Greg Osuri, 00:14:21

Context: Announcing that Cosmos interchain-accounts support is code-complete and in review on GitHub; Greg predicts smart-contract-driven cloud deployment as a showcase Cosmos use case.


Using Amazon is fundamentally wrong for web3

Akask Weekly - Aug 3rd 2022 (Akash Network)

“If you were to use cloud infrastructure, I’d rather prefer putting your infrastructure in a bare metal server sitting in a secure data center that you have complete control over, and not Amazon. Using Amazon for a web3 project is just fundamentally wrong — over-concentration of power… Akash is not really a cloud, it’s a market.” — Greg Osuri, 00:40:05

Context: Panel question on web3 teams preaching decentralization while running on web2; at [00:42:17] he sharpens it: “web3 projects should not use Amazon, period,” while granting startups’ need for iteration speed.


Web3 on AWS is a bomb waiting to go off

Akask Weekly - Aug 3rd 2022 (Akash Network)

“Amazon Web Services is the number one host for crypto today. It’s fully closed source, it’s fully opaque — we have no idea what’s happening underneath the hood… there’s so many things that could go wrong in a closed source system, and web3 relying on the systems is not okay. We’re just waiting for a bomb to happen, and I don’t want to know what happens when sixty percent of the Ethereum network is attacked because they were vulnerable for a zero day on Amazon.” — Greg Osuri, 00:07:10

Context: Extending the Solana wallet supply-chain hack lesson to cloud infrastructure; he adds Amazon went down 113 times in 2021 per his open-source outage tracker.


Akash is the foundational decentralized layer to rebuild web3's centralized dependencies

Akash Weekly - August 10th 2022 (Akash Network)

“Using that as a foundation, as a heavily distributed, decentralized system as a foundation — imagine the kind of tools you can build on top of it, all the way from source code hosting to node hosting to shared storage solutions on top of Akash. I think Akash gives you the foundational layer for folks to build additional abstractions on top that is founded on a decentralized base.” — Greg Osuri, 00:41:41

Context: Responding to Todd’s points that web3 is dangerously reliant on GitHub and Cloudflare (Tornado devs’ repos vanished) and that Cosmos validators concentrate in Hetzner data centers.


Akash is the first viable open cloud vs. the oligopoly

Akash Weekly - August 17th 2022 (Akash Network)

“Akash Network is the first viable open cloud. So cloud computing today, the infrastructure side of things, is essentially controlled by few closed companies — Amazon, Microsoft, Alibaba and Google being the majority stakeholders… Akash is the first viable attempt to decentralizing this oligopolized closed system called cloud computing.” — Greg Osuri, 00:42:02

Context: AMA answer on what problem Akash solves; he adds it is a non-custodial system “97% cheaper than the cloud” (in the [00:42:46] block).


Phase four: thousands of nodes for low-latency distributed compute

Akash Weekly - August 17th 2022 (Akash Network)

“Phase four is to really offer low latency, heavily distributed computation, and in order for that to happen we need to have thousands of nodes spread across the world. Right now we have about 50 providers spread across the world. By the time we reach phase four we will essentially offer a product that’s unlike any other product in web2 or web3.” — Greg Osuri, 00:48:29

Context: The endgame of the roadmap — a globally distributed edge-like compute network, quantified against the ~50 providers live in August 2022.


Sovereign cloud services are the future

Akash Weekly - August 17th 2022 (Akash Network)

“Now we’re seeing this VPN service on Akash, 54 cents a month, that you have full control over, that you have full sovereignty over… that gives me all kinds of new ideas, right, like what kind of services are possible in the future on Akash where you can guarantee sovereignty for the users.” — Greg Osuri, 00:09:15

Context: Reacting to a community member running a personal VPN on Akash cheaper than any commercial VPN; he frames sovereignty and privacy as the defining values of the new cloud.


Autonomous, ownerless deployments no one can take down

Akash Weekly - August 24th 2022 (Akash Network)

“You can imagine a world where a DAO, or a smart contract, actually completely owns the deployment without human intervention — so that removes the risk of the human being vulnerable… now we can have programs owning other programs and hosting those programs… I’m really getting excited about this idea that we can fully have autonomous deployments, ownerless deployments — essentially no single human being owning a deployment, fully running autonomously — and they cannot be taken down, because you can’t take down a smart contract, that means you can’t take down a deployment either.” — Greg Osuri, 00:03:35

Context: Sparked by conversations with the Juno team about fully on-chain deployments using interchain accounts and Handshake DNS discovery; he calls it “a new era of decentralized cloud that the world has never seen” [00:04:18]. Prefigures his later AI-agent/autonomous-infrastructure themes.


No Akash providers on Amazon — it's 97% cheaper here

Akash Weekly - August 24th 2022 (Akash Network)

“I’ve looked at preliminary data — I haven’t seen anyone running on Amazon at all, because it does not make sense. Like, why would you host on Amazon and provide the same compute for 97 percent cheaper? It makes no sense… the real data indicates there are no Amazon providers.” — Greg Osuri, 00:31:03

Context: Rebutting a viral graphic claiming Akash providers ran on AWS; he explains the data captured Praetor test providers (Akash lacked a testnet), and describes the verification system he built over a weekend — deploying to all ~50 providers every 10 minutes with reverse-IP lookups — with a full report promised in one to two weeks.


The front-end is the censorable layer — so Akash has none

Akash Analysis - How Akash is censorship resistant (Akash Network)

“These days the censorable part of the decentralized stack is not the actual protocol but the front end itself. So Akash network not having a front-end, or Overclock Labs which develops Akash network not developing a default front-end, puts us in a very interesting position because now our compliance requirements are much lower than most other protocols… the protocol itself doesn’t choose favorites or doesn’t really enforce censorship.” — Unknown, 00:00:00

Context: August 2022, right after the Tornado Cash sanctions; he contrasts Pocket Network’s front-end ban and says “I’m kind of proud of our decision of not developing a front-end in the first place.” Speaker: uncredited clip; likely Greg.


No Akash providers actually run on Amazon

Akash Analysis - Are Akash providers hosted on AWS? (Akash Network)

“I’ve looked at preliminary data… I haven’t seen anyone running on Amazon at all, because it does not make sense — like why would you host on Amazon and provide the same compute for 97 [percent] cheaper? It makes no sense… the real data indicates there are no Amazon providers.” — Unknown, 00:02:50

Context: Debunking a viral graphic built from test-provider IPs; he promised a public verification report from a system deploying canary workloads to all ~50 providers. Speaker: uncredited clip; likely Greg (describes personally coding the verification system).


Big clouds will ban crypto next

Akash Weekly - August 31st 2022 (Akash Network)

“This is just the first step and next step obviously is going to be the Amazons and Googles of the world.” — Greg Osuri, 00:02:55

Context: Predicting the Hetzner crypto ban will be followed by similar bans from hyperscalers, pushing crypto infrastructure to decentralized clouds.



Enterprise convergence within a couple of years

Akash Weekly - August 31st 2022 (Akash Network)

“I think in a couple of years you’re going to see this amazing convergence happen and I’m super bullish, and that’s the vision that we’re working towards.” — Greg Osuri, 00:21:49

Context: On Akash becoming enterprise-ready (private networking, private registries) so companies like Walmart can be both providers and users within a couple of years.


Enterprise repatriation creates underutilized supply for Akash

Akash Weekly - August 31st 2022 (Akash Network)

“Our thesis of, well, you’re going to have companies move away from the cloud and the data centers are going to get repatronized and there’s going to be a mass underutilization… we’re going to see a lot of these players like Walmart come away.” — Greg Osuri, 00:21:04

Context: Argues repatriated enterprise data centers (e.g., Walmart, whose utilization outside shopping season is under 5-10%) will offload spare capacity onto Akash. “Repatronized” is the auto-caption’s rendering of “repatriated.”


Most Akash validators were running on Hetzner

Ask Akash - How Decentralized are Akashs Validator Nodes? (Akash Network)

“There was a report I saw recently that indicated most of the validator nodes — not providers, validator nodes — are actually running on Hetzner. So I was a little surprised by this fact, because I would imagine if you’re an Akash validator you would not be running on a centralized [provider].” — Unknown, 00:00:00

Context: Candid admission that even a decentralized network’s validators cluster on centralized hosts; prompted by Hetzner’s crypto ban. Speaker: uncredited Ask Akash answer; likely Greg.


The infrastructure of the future is decentralized

Akash Weekly - August 31st 2022 (Akash Network)

“Your infrastructure, when hosted on a centralized provider, is not safe and there are no guarantees in this game… infrastructure of the future is decentralized and Akash is leading that charge.” — Greg Osuri, 00:02:13

Context: Reacting to Hetzner’s ban on crypto workloads, which he says will “set forth a series of bans by centralized cloud providers.”


Validators on Akash would be a risky "meta deployment"

Ask Akash - How Decentralized are Akashs Validator Nodes? (Akash Network)

“Akash validators running on Akash means a meta deployment — now that’ll be something crazy… I wouldn’t encourage the actual full nodes to be running on Akash itself, because when things go wrong the validators are supposed to fix the chain, and they’re fixing a chain that they’re hosted on — it’ll get into all kinds of unknown problems.” — Unknown, 00:01:28

Context: Nuanced self-limit on dogfooding; RPC and ancillary infrastructure on Akash is fine. Speaker: uncredited Ask Akash answer; likely Greg.


Enterprise convergence within a couple of years

Akash Analysis - The Return of the Datacenter (Akash Network)

“Akash is slowly getting to be enterprise ready, so sooner or later we’re going to be ready for a Walmart type — both as a user… right now we’re okay as a provider, but also as a user. I think in a couple of years you’re going to see this amazing convergence happen, and I’m super bullish — that’s the vision that we’re working towards.” — Unknown, 00:02:10

Context: Timed prediction (from September 2022) that enterprises will both supply and consume Akash compute. Speaker: uncredited clip; likely Greg.


Repatriated data centers will offload capacity onto Akash

Akash Analysis - The Return of the Datacenter (Akash Network)

“When they build their own data centers they’re going to have underutilized supply… most traffic for Walmart comes during a shopping season, be it Black Friday or Christmas… most of the time the utilization is going to sit less than five percent or ten percent. So when that happens, their only avenue to offload their capacity is Akash network.” — Unknown, 00:00:43

Context: Extends the repatriation trend into an Akash supply-side prediction. Speaker: uncredited clip; likely Greg.


White-paper thesis vindicated: cloud is expensive long-term

Akash Analysis - The Return of the Datacenter (Akash Network)

“The cloud industry, which is about 500 billion dollars industry, is now starting to see cracks. Companies like Walmart are coming back from the cloud because they realize cloud over time is very very expensive — and this is something we wrote in our white paper… cloud only looks great on the short term but on the long term it’s actually very expensive, and people are now starting to realize that.” — Unknown, 00:00:00

Context: Claims Akash’s 2018 white paper predicted cloud repatriation. Speaker: uncredited clip; likely Greg.


Fully decentralized stack is the only way out of the censorship regime

Akash Weekly - September 7th 2022 (Akash Network)

“You can’t go censor a smart contract, you can’t stop a smart contract. So the only way out of this, like, heavy censorship regime is a fully decentralized stack, and we’re all working very hard to get there.” — Greg Osuri, 00:15:45

Context: Framed against Hetzner’s crypto ban and 2022 censorship concerns; why Kava’s Prop 97 incentivizing Akash-hosted DeFi matters.


On-chain manifests will enable unstoppable, DAO-owned services

Akash Weekly - September 7th 2022 (Akash Network)

“A DAO on Juno can now own deployments on Akash… most deployments can become immutable… once the deployment happens, it happens forever, right, on chain. And if you can push the manifest on chain you can have this amazing, like, unstoppable services, and since it’s all on chain it’s fully verifiable.” — Greg Osuri, 00:14:18

Context: Describing a planned Akash feature for purely public, on-chain manifest deployments completable by smart contracts.


The future of DeFi is full supply-chain decentralization

Akash Weekly - September 7th 2022 (Akash Network)

“I think the future of DeFi is going to be a full supply chain decentralization. A full supply chain decentralization begins with the git repos, the merge mechanisms, the identity establishment by the contributors — all the way from the contributions to the deployment, fully on chain, fully verifiable. And that’s the kind of future Akash is already getting ready for.” — Greg Osuri, 00:15:01

Context: Extending the on-chain deployment idea to code provenance via decentralized GitHub alternatives (Radicle/Utopia), so a CoinGecko could verify a DAO-approved frontend URL.


Hetzner is the first of a series of cloud crypto bans

Akash Analysis - What the Hetzner ban means for Akash (Akash Network)

“This announcement is a major announcement which I believe will set forth a series of bans by centralized cloud providers, and this is exactly the reason why Akash exists — Akash exists for a non-censorable, sovereign, always-on cloud.” — Unknown, 00:00:00

Context: September 2022 Hetzner crypto ban; he notes Hetzner ran most Cosmos node infra and 17% of Ethereum’s cloud nodes. Speaker: uncredited clip; likely Greg.


The Amazons and Googles will ban crypto next

Akash Analysis - What the Hetzner ban means for Akash (Akash Network)

“This is just the first step, and next step obviously is going to be the Amazons and Googles of the world… the only reliable, scalable, sovereign, censorship-resistant infrastructure of the future is decentralized, and Akash is leading that charge.” — Unknown, 00:01:26

Context: Escalation prediction extending the Hetzner ban to hyperscalers; paired with the claim that “there are no guarantees” hosting on centralized providers. Speaker: uncredited clip; likely Greg.


Akash 95% cheaper; 1,000-1,500 providers expected soon

Ask Akash - How will $AKT benefit from enabling Fiat Payments (Akash Network)

“Akash is still like 95 percent cheaper than your traditional cloud, with current distribution of like 50 providers or so — it’s very very distributed across the globe — and with provider incentives that we’re working on, we’re going to expect at least 1,000 to 1,500 providers very soon.” — Unknown, 00:01:26

Context: Quantified cost claim plus a network-growth forecast tied to upcoming provider incentives (September 2022). Speaker: uncredited Ask Akash answer; likely Greg.


A fully verifiable software supply chain, from source code to deployment

Akash Weekly - September 14th 2022 (Akash Network)

“You can imagine a scenario where a build on Radicle or Utopia gets deployed — a verified build, right? Verify that the developer is legit, they write the code, that gets [an] on-cha[in]… deployment that is verifiable… so now you can essentially create the entire verifiable supply chain and have the guarantees that what you’re running in terms of code is actually legit.” — Greg Osuri, 00:50:10

Context: Responding to DNS/front-end attacks (Uniswap-style) and wallet key-leak incidents; envisions browsers warning users or letting them self-host verified frontends on Akash.


Decentralization will take center stage

Akash Weekly - September 14th 2022 (Akash Network)

“The most important thing from a project standpoint is for us to keep building, keep shipping features, keep improving the system in a manner that it opens up a lot more use cases… and before you know it, the decentralization is going to be in the center stage.” — Greg Osuri, 00:03:37

Context: Describing web3 network effects “that are bound to go up” after the DoD win.


Half of every dollar paid for cloud services goes to "Uncle Jeff"

Akash Weekly - September 14th 2022 (Akash Network)

“50 [percent] on average [of] all the costs — the margins — actually goes to the big cloud… to the cloud providers. It’s incredible. So half of every dollar we pay to cloud services, all of us, goes to Uncle Jeff… the rent seeking model that was created, the invisible rent seeker.” — Greg Osuri, 00:13:44

Context: Citing a16z’s “cost of cloud” post; services like Notion, Asana, and Netflix pass ~50% of costs through to hyperscalers. Decentralization, he argues, also “reduce[s] the cost of the globe… in these high inflation times.”


Open technologies are the future — governments must escape centralized choke points

Akash Weekly - September 14th 2022 (Akash Network)

“We anticipated this level of involvement long time ago. We’ve been arguing the fact that open technologies are the future because they reduce reliance… think about how much money and time investment goes into Amazon’s really centralized systems that have a lot of control and more often than not become choke points for combat operations or anything strategically important. It’s important for our government to operate without reliance on closed technologies.” — Greg Osuri, 00:02:54

Context: On the U.S. DoD website deployed to Akash via a community member. Greg adds later (00:14:27): “if Akash is good enough for a U.S. Department [of] Defense website, I think Akash is good enough for a lot of folks.”


The centralized cloud could decide to ban decentralization

Akash Weekly - September 14th 2022 (Akash Network)

“Then we have this bigger problem, which is the centralization of the cloud that… can decide one day to ban decentralization, right? And it’s [at] a very critical juncture right now.” — Greg Osuri, 00:14:27

Context: Framing censorship risk to crypto’s base layer — the argument for sovereign infrastructure, made days after Hetzner’s crypto ban.


The centralized cloud is a kill switch

Akash Weekly - September 14th 2022 (Akash Network)

“As we know, crypto right now is on a major assault by censorship as well as outright banning… I’m going to go on talking about why decentralization is important — [Hetzner] in the cloud basically is a kill switch.” — Greg Osuri, 00:12:59

Context: Previewing his Messari spaces talk on base-layer decentralization, after Hetzner banned crypto workloads (captions garble “Hetzner”); Akash offered free hosting to affected projects.


DoD website as the credibility proof point

Akash Analysis - Why decentralization is important (Akash Network)

“Akash is delivering, right? I mean if Akash is good enough for a US Department of Defense website, I think Akash is good enough for a lot of folks… these proof points are proving daily the fact that we have a viable technology that is secure, that is ready, that’s growing.” — Unknown, 00:01:27

Context: Referencing the Marine Corps unit website deployed to Akash after the 2022 DoD hackathon; adds that centralized cloud “can decide one day to ban decentralization.” Speaker: uncredited clip; likely Greg.


Half of every cloud-service dollar goes to the big clouds

Akash Analysis - Why decentralization is important (Akash Network)

“50 [percent] on average of the costs, all the margins, actually goes to the big cloud providers. It’s incredible — half of every dollar we pay to cloud services, all of us, goes to Uncle Jeff… the invisible rent seeker that we all don’t realize.” — Unknown, 00:00:46

Context: Citing a16z’s “cost of cloud” post; his recurring “Amazon tax” argument for why decentralizing cloud lowers costs economy-wide. Speaker: uncredited clip; likely Greg.


Ecosystem partners will replace the core team as the growth engine

Akash Weekly - September 21st 2022 (Akash Network)

“Platform clients that, you know, will eventually replace the Akash core team as the primary contributors to growth — and it’s already starting to happen. So Spheron is focusing on bringing on ecosystems like Polygon, countries like Africa, onto the Akash ecosystem… this level of organic involvement is exactly what we want to see, and we’re going to continue seeing that over the next few months and few years.” — Greg Osuri, 00:30:07

Context: His model for decentralized network growth: the core team builds features while partners (Spheron, Cloudmos, Praetor) expand reach organically.


From 20% to 80% of the Cosmos ecosystem in 12 months

Akash Weekly - September 21st 2022 (Akash Network)

“Over the next 12 months we want to be 80%, right? And the plan really is to focus on our immediate ecosystem, which is Cosmos, so that we can nail the user experience and we can nail the retention, and slowly expand to other ecosystems.” — Greg Osuri, 00:31:32

Context: He’d just claimed “safely we can say about 20% of Cosmos ecosystem uses Akash right now” with 50+ ecosystem projects (00:30:50); quantified adoption target.


Akash could reach 100% cloud penetration

Akash Weekly - September 28th 2022 (Akash Network)

“It’s very likely that in the future Akash would have 100% cloud penetration, which is a trillion dollar market… by the time you know it, there’s a real good possibility Akash could have 100% market share.” — Greg Osuri, 00:41:36

Context: Continuation of the AMA answer; he hedges it as “not financial advice” and “hyper risk,” tying AKT’s 389M fixed supply to securing that market. Second sentence is at [00:43:04].


Cloud margins are a permanent invisible tax

Akash Weekly - September 28th 2022 (Akash Network)

“Over 50 percent, on average, of all the money you pay to online services — like Netflix or the service fees to Uber Eats — 50% of that margin really goes to a cloud provider.” — Greg Osuri, 00:07:12

Context: Citing Anil Murty’s Messari Mainnet talk; Greg follows at [00:07:57] with “you end up having a permanent invisible tax for everything you do online being paid to one of these large providers.”


Hyperscalers will join a multi-cloud standard

Akash Weekly - September 28th 2022 (Akash Network)

“Akash has always been… to bring people to the table, create a multi-cloud standard. I think that’s very likely the future, where you’re going to have the big clouds and the small clouds and everybody at the same table, leveling the playing field and giving the consumer the ultimate benefit, which is true cost of the cloud.” — Greg Osuri, 00:10:06

Context: Responding to a Coinage Media Twitter Spaces question about what happens if Amazon, Google, or Microsoft join Akash.


The future is bringing back your own data center, offset by Akash

Akash Weekly - September 28th 2022 (Akash Network)

“Having their own data center, even though it may be expensive in the short term, actually is good in the long term… bringing back your data center is really where the future belongs. Akash can help you offset some of the costs… instead of needing 100 servers a company can have 80 or 70 servers and offload that unused capacity on the Akash Network, earn tokens, and use their tokens back when they want to scale.” — Greg Osuri, 00:08:40

Context: Argues companies like Intuit or Walmart are realizing cloud is no longer the best cost optimizer; predicts repatriation plus marketplace monetization of idle capacity.


Trillion-dollar cloud market; Akash 94% cheaper, bills halved

Akash Weekly - September 28th 2022 (Akash Network)

“They represent hundreds of billions of dollars, on track to become a trillion dollar market in the next five years or so… Akash is the first attempt at creating an open cloud where the cost is 94% cheaper. What that means is our Netflix bills are going to be almost half… our cloud bills are going to be almost half.” — Greg Osuri, 00:40:51

Context: AMA answer to an investor question on what makes Akash unique; quantified cost and market-size forecast.


20ms latency for the 95th percentile of internet users

Building The Infrastructure Of Web3 | Greg Osuri Of Akash Network | The FH Show - EP16 (Felix O. Hartmann)

“With a network of a thousand regions, the goal is to reduce latency to 20 milliseconds — network latency to 20 milliseconds for 95th percentile active internet population.” — Greg Osuri, 00:29:04

Context: Answering whether Akash could serve metaverse/cloud-gaming workloads; contrasts with ~300ms US coast-to-coast latency and says a thousand regions “unlock a whole range of applications we haven’t yet seen.”


A hyperscaler will be an Akash provider within three years

Building The Infrastructure Of Web3 | Greg Osuri Of Akash Network | The FH Show - EP16 (Felix O. Hartmann)

“In three years I would like to see a Microsoft or a Google or Amazon [on] Akash… I think it’s gonna happen… then we’ll go talking petabyte, not terabyte.” — Greg Osuri, 00:54:44

Context: Asked what the next three years look like; also cites grassroots DoD adoption (“military-grade”) and predicts “quite a lot of enterprise penetration in the next three years.”


Akash will be the most distributed network by 2024

Building The Infrastructure Of Web3 | Greg Osuri Of Akash Network | The FH Show - EP16 (Felix O. Hartmann)

“We’re targeting about a thousand providers for the next 12 months or so… we want really heavily distributed, like the original vision of the edge computing. I think we’re going to see that. So by 2024 I think Akash [is] the most distributed network.” — Greg Osuri, 00:18:31

Context: Provider acquisition plan: ~50 providers at the time (pre-incentives), targeting 1,000 providers and 60+ protocol clients within 12 months.


Edge data centers are disappearing — a warning

Building The Infrastructure Of Web3 | Greg Osuri Of Akash Network | The FH Show - EP16 (Felix O. Hartmann)

“Back then there were 8.2 million data centers; when we began [it] was at 7.1 million data centers… that’s very scary if you don’t do something about it… the edge data centers owned by companies, individuals are going away. If you don’t do something about getting them back, we will not be able to be at the purview of the cloud.” — Greg Osuri, 00:07:09

Context: Argues hyperscaler consolidation is destroying independent edge capacity; Akash’s mission framed as reversing this. (Quote spans into the 00:07:51 block.)


Everyone eventually outgrows shared infrastructure

Building The Infrastructure Of Web3 | Greg Osuri Of Akash Network | The FH Show - EP16 (Felix O. Hartmann)

“There will come a point where you cannot use shared infrastructure anymore… Dropbox started off on the cloud but they came to a point where it’s very very expensive… people are going to need their own compute, their own sovereignty — be it compute, be it software, whatever layer you’re running on — eventually for scaling, and there’s really no other way.” — Greg Osuri, 00:47:40

Context: The WordPress-to-own-servers analogy applied to blockchains (appchains vs. shared state) and to compute sovereignty generally; earlier he states “my thesis is once you bootstrap you will need sovereignty” (00:46:13).


The 50% "Uncle Jeff tax" on online services

Building The Infrastructure Of Web3 | Greg Osuri Of Akash Network | The FH Show - EP16 (Felix O. Hartmann)

“a16z came up with an amazing report recently called the Trillion Dollar Paradox, where they said 50 [percent] of every penny you spend on online services like Asana or Netflix and whatnot goes to Amazons or Googles of the world… half of every dollar you spend goes to the Uncle Jeff tax.” — Greg Osuri, 00:08:33

Context: Cited to argue cloud margins are unsustainable and companies are repatriating from cloud back to data centers; “three companies really control eighty percent of the market share [of] the cloud.”


Half of Ethereum's hosted nodes run on AWS

Akash Weekly - October 5th 2022 (Akash Network)

“The base layer is where you run the actual nodes… if you’re Ethereum, you’re essentially running on AWS — 50% of hosted nodes on Ethereum runs on AWS… so that layer itself is another level of censorship.” — Greg Osuri, 00:05:51

Context: His three-layer censorship model of the web3 stack (access point / consensus layer / base layer); Akash targets the front-end and base layers.


The only viable open source cloud

Akash Weekly - October 5th 2022 (Akash Network)

“We’re bringing that to the market with the first sort of open source cloud to do that — in fact, we are the only open source cloud that’s viable, so I guess we’ll be the first in anything we do.” — Greg Osuri, 00:02:59

Context: On shipping IP leases in a decentralized way for Mainnet 4.


A thousand providers, then "superiority" over the public cloud

Akash Weekly - October 12th 2022 (Akash Network)

“We expect at least a good thousand providers… maybe we will expect another 50 protocol clients, and then the phase three is really superiority. That’s when we’ll start seeing much better use cases that are not possible in the cloud.” — Greg Osuri, 00:07:59

Context: His three-phase master plan: security, parity with public cloud (current), then use cases centralized clouds cannot serve.


Sub-20ms latency everywhere unlocks new application classes

Akash Weekly - October 12th 2022 (Akash Network)

“What do you do when you have a thousand providers in a thousand regions that can deliver a latency signature of less than 20 milliseconds anywhere in the world? The kind of applications, the kind of games that get built — that’s the kind of stuff… we focus on stories, we focus on adoption.” — Greg Osuri, 01:02:22

Context: Answering a revenue-guidance question; he argues success is qualitative for now and edge-density latency is the real unlock.


Adoption will be network-based, not corporate-based

Akash Weekly - October 19th 2022 (Akash Network)

“People that share the values of decentralization with the projects that implement them are inherently vested to spread the philosophy. So I think adoption is going to be not a corporate-based adoption, it’s going to be network-based adoption, and that’s why I’m super bullish… it’s just spreading these value systems at internet scale.” — Greg Osuri, 00:45:02

Context: Explaining how Akash and Secret reached the Marine Corps without enterprise sales teams; he adds “short term it looks painful but… for the longest horizon possible I’m super bullish.”


Open source replaces the government's billion-dollar cloud

Akash Weekly - October 19th 2022 (Akash Network)

“A government… spends billions of dollars for Amazon to have a sovereign deployment platform, which can be replaced by open source system. So open source is clearly winning here, and that’s the story that is going to matter when it comes to larger enterprises and institutions.” — Greg Osuri, 00:40:04

Context: On the DoD deploying to its own hardware via Akash instead of contracting hyperscalers; he notes getting Kubernetes doesn’t give you deployability, Akash does.


1,000 providers = the most distributed cloud in the world

Akash Weekly - October 28th 2022 (Akash Network)

“By the time you’re done with parity, you can expect at least a thousand different providers… that means a thousand different regions. Akash will be the most distributed cloud in the world — fully open, fully permissionless… possibilities like innovation beyond our imagination… you can do low latency, for example, you could do cloud gaming — a lot of the things that you couldn’t do right now.” — Greg Osuri, 00:53:20

Context: Three-phase strategy answer (security → parity → superiority); he also claims “about 25% of Cosmos uses Akash in one form or the other” (00:54:44).


Akash is the first commercial implementation of a supercloud

Akash Weekly - October 28th 2022 (Akash Network)

“Data centers are great to get sovereignty, but they’re not good to scale… [a layer that] gives you a homogeneous way to [run the cloud] on a heterogeneous resource set — that is a supercloud, and Akash is the first commercial implementation of a supercloud… The cloud evolved — how you think about cloud in a modern sense is not how you thought about cloud in the web2 sense.” — Greg Osuri, 00:57:37

Context: Closing the AMA on long-term strategy; credits Cornell’s 2016 supercloud concept and cites Hey.com/Dropbox-style cloud exits as the demand driver.


"Akash lets you be sovereign at scale" — web2 workloads will follow DHH off the cloud

Akash Weekly - October 28th 2022 (Akash Network)

“Akash lets you be sovereign at scale… Akash lets you have your own data center while being able to tap into a vast resource network of data centers… in a permissionless manner, priced in an open setting… DHH’s blog post is an amazing testimony and a validation of our thesis… and that’s going to be the trend, my friends — we’re going to be able to move a lot of these web2 workloads… onto Akash as soon as we get to maturity.” — Greg Osuri, 00:18:03

Context: Reacting to DHH (Ruby on Rails creator) announcing 37signals’ cloud exit; Greg says cloud is “cost prohibitive” past bootstrap stage and cloud giants “are actually hurting the design of the web” — repatriation validates Akash’s thesis.


From cloud parity to superiority — the first open IP-address marketplace

Akash Weekly - October 28th 2022 (Akash Network)

“The big feature is really the IP address marketplace. This is the first ever open marketplace for IP addresses there is in the world… we always talked about getting to parity with the current cloud, but now we’re starting to get into superiority.” — Greg Osuri, 00:01:27

Context: Mainnet 4 preview (two weeks out); IP leases enable fixed-port workloads like Solana and Bitcoin nodes.


Data centers are dying while hyperscalers grow

The Problem w/Blockchains on Big Cloud (Akash Network)

“Before, in 2018, you had about 8.4 million data centers; now you have about 7.1 million data centers… and the hyperscalers are getting a lot bigger, because there are very, very little options to get cloud-like capability on data centers… you can either have a data center or you can have a cloud — you can’t use your data center capacity as a cloud provider. That’s the big challenge.” — Greg Osuri, 00:04:19

Context: His structural argument for Akash: independent capacity is stranded without cloud-grade deployment tooling, so demand consolidates to Amazon despite 7.1M data centers existing. Quote spans into the [00:05:03] block.


The Amazon tax: 50 cents of every online-service dollar

The Problem w/Blockchains on Big Cloud (Akash Network)

“Turns out 50 [cents] of every dollar you and I spend on online services — like Asana, like Netflix, like Notion or whatever — goes to Amazon or Google. Fifty percent… it’s an invisible tax that we all have, and guess who pays the bill — we end up paying, because the cost and savings come down to the user.” — Greg Osuri, 00:04:19

Context: Citing the a16z “cost of cloud, a trillion dollar paradox” article as evidence that cloud economics are getting worse and expensive decentralization is prohibitive.


A thousand providers will make Akash the most distributed cloud in the world

Ask Akash - What is Akash's strategy going forward? (Akash Network)

“By the time you’re done with parity, you can expect at least a thousand different providers… that means [a] thousand different regions. Akash should be the most distributed cloud in the world, and that’s fully open, fully permissionless… that gets into possibilities like innovation beyond our imagination — what can you do with thousands of regions across the world? You can do low latency… you could do cloud gaming.” — Greg Osuri, 00:03:41

Context: The “superiority” phase of his three-phase strategy (security → parity → superiority), stated November 2022.


Akash is the first commercial implementation of the supercloud

Ask Akash - What is Akash's strategy going forward? (Akash Network)

“Data centers are great to get sovereignty but they’re not good to scale. That’s why Akash [has] the idea of a supercloud… where you can have a heterogeneous resource set and give a homogeneous layer that addresses all the functionality of the cloud, from deployment, scheduling, fault tolerance, observability to automation… Akash is the first commercial implementation of a supercloud.” — Greg Osuri, 00:07:59

Context: Cites the Cornell-coined “supercloud” term and Akash’s 2020 white paper; also claims ~25% of the Cosmos ecosystem used Akash in some form at the time.


Cloud must be a public utility, and blockchain is the only structure for that

Akash Weekly - November 16th 2022 (Akash Network)

“Enterprise cloud works to a certain degree but it does not work the way it should work as a public utility. The cloud needs to be a public utility. The only data structure to enable a public utility is a blockchain.” — Greg Osuri, 00:54:23

Context: Answering “why does a computing marketplace need to be web3?” — his recurring cloud-as-utility thesis.


Decentralization is the only way out

Akash Weekly - November 16th 2022 (Akash Network)

“If any lesson that you had to learn from the last few weeks, the centralization led to the [contagion] and decentralization is the only way out. So that’s why cloud needs to be decentralized.” — Greg Osuri, 00:55:07

Context: Closing the AMA answer, tying the FTX/Luna collapses to the case for decentralized cloud.


Three companies control 60-70% of compute

Akash Weekly - November 16th 2022 (Akash Network)

“There are three companies that control over… I think over 60 to 70 percent of the compute and maybe like eight companies that control ninety percent of the compute… it’s getting to a point where few companies control our future.” — Greg Osuri, 00:51:25

Context: Concentration stat used to argue centralized cloud is a societal risk (stability, cost, control).


Console will drive measurably higher usage

Founders Reflect on Why Akash was Created (Akash Network)

“We’re going to be very aggressive in terms of developer marketing with Console… you’re going to start noticing much much higher user activity on Akash, and this user activity is just not madness — this is measured methodically and this is quantified.” — Greg Osuri, 00:03:39

Context: Explaining why Console’s analytics/user-journey instrumentation matters for adoption.


Cloud parity is a multi-year journey

Ask Akash: Thoughts on usage metrics over the last year (Akash Network)

“All these features are required for us to get parity [with] the cloud and then comfortably go to a big major Fortune 10 company or Fortune 50 company and be like, hey, move all your workloads [to] Akash, because now you have all the capabilities that you need to run. That’s not the next few months — it’s going to be next few years.” — Greg Osuri, 00:12:55

Context: Tempering a questioner’s optimism about timelines; stable payments, observability, automation, identity, and access control were still on the roadmap, with GPUs next after Mainnet 4.


Adoption will come through verticalization on top of the protocol

Dcentral 2022 - DeCloud - Decentralized Cloud & Storage (Akash Network)

“You can build other stuff on top of Akash, and most of the adoption, I would think, would come from verticalization… Akash is a general purpose cloud… it’s like taking an F1 car to buy groceries… I would imagine you would build additional verticals on top of it that optimize for single use cases… there are about seven different companies working on building different protocol clients on top of Akash.” — Greg Osuri, 00:18:00

Context: Standards discussion — Akash builds on Kubernetes (“standing on shoulders of giants”) rather than from scratch; forecast that vertical clients drive mainstream usage.


Akash: the world's first decentralized supercloud, targeting the true price of compute

Dcentral 2022 - DeCloud - Decentralized Cloud & Storage (Akash Network)

“We created Akash Network. Akash is the world’s first decentralized super cloud… a marketplace where one can acquire unused capacity… The goal really is to decentralize cloud deployment… avoid vendor lock-in and be able to leverage the most optimal pricing, and do so in a censorship-resistant and open setting.” — Greg Osuri, 00:02:11

Context: Panel intro; he adds Akash serves ~25-30% of the Cosmos ecosystem, is moving into ML with much cheaper GPUs, and aims “to get to true price of computing” and solve growing censorship attacks.


Non-technical users shouldn't ever see SDL

Ask Akash - How does a one use the SDL builder (Akash Network)

“As a non-technical person, I think you shouldn’t even care about SDL. Hopefully we will get to that point sometime in the future, but right now Akash is meant for developers, and SDL really makes a developer’s life simple.” — Greg Osuri, 00:02:11

Context: Product-direction statement: abstraction layers should eventually hide deployment config entirely from mainstream users.


Akash has a more secure design than Amazon

Akash Analysis Should Web2 projects consider running on decentralized infrastructure (Akash Network)

“Fortunately we don’t have any — haven’t had any security event, and I want to keep it that way. I think Akash itself has a much more secure design than Amazon — being open, being hyper focused on selecting the best infrastructures, and also not automating a lot of things people expect but rather giving the flexibility to the developer to choose the security architecture they please.” — Greg Osuri, 00:02:08

Context: Answering security concerns about decentralized infrastructure; cites remote signers for validators as an example of forced-secure trade-offs.


Single-command deploys better than Amazon, Vercel, Netlify

Akash Analysis Should Web2 projects consider running on decentralized infrastructure (Akash Network)

“If you look at some of the repos we have, it’s a single command deploys, right — from your vanilla repos to getting highly productive and into production within a single command on Akash, which is significantly better than Amazon… in fact even better than like Vercels or the Netlifys.” — Greg Osuri, 00:04:58

Context: Conceding AWS wins on iteration speed today, he claims Akash tooling under development will exceed mainstream PaaS UX and “shows you the power of non-custodial systems.”


Web3 on Amazon is fundamentally wrong

Akash Analysis Should Web2 projects consider running on decentralized infrastructure (Akash Network)

“If you were to use cloud infrastructure, I’d rather prefer putting your infrastructure in a bare metal server sitting in a secure data center that you have complete control over, and not Amazon. I just think using Amazon for a web three project is just fundamentally wrong. Over-concentration of power.” — Greg Osuri, 00:01:24

Context: Responding to a validator team defending web2 infrastructure as a stepping stone; he later repeats “web3 projects should not use Amazon… can use anything else.”


Sovereign services cheaper than commercial equivalents

Akash Analysis - VPN use case (Akash Network)

“We’re seeing this VPN service on Akash — 54 cents a month — that you have full control over, that you have full sovereignty over, that you know for sure is private… that gives me all kinds of new ideas: what kind of services are possible in the future on Akash where you can guarantee sovereignty for the users?” — Unknown, 00:00:00

Context: Community-run VPN at $0.54/month vs. ExpressVPN’s ~$4/month on shared infrastructure; forward-looking about a class of user-sovereign services. Speaker: uncredited clip; likely Greg.


A war between privacy and surveillance; centralized databases always leak

Akash Weekly - November 30th 2022 (Akash Network)

“There is a war that’s more apparent now between the advocates of privacy and the advocates of surveillance… The fact that there is a database somewhere sitting that has your IP addresses and your transaction history is fundamentally dangerous, because this database is never secure… Last thing I want is my information with my public IP address, with all the transaction history associated, by name… on the dark web — and we know that’s going to happen.” — Greg Osuri, 00:02:50

Context: Reacting to ConsenSys’s privacy policy allowing IP/transaction logging; his argument that protocols must be built so they “cannot collect information” rather than promise not to.


Akash can never ban you — and that migration is the trend

Akash Weekly - November 30th 2022 (Akash Network)

“One thing about Akash is Akash can never ban you. There’s nothing that I can do or anybody can do to ban you, and that is very powerful construct… I think that’s going to be the trend moving forward.” — Greg Osuri, 00:11:31

Context: Prompted by Hetzner’s blanket ban on web3 workloads and an NFT project (Moonwise) migrating to Akash; companies are “not evil but inherently risky” because decisions hinge on a few executives.


Akash has never gone down

Akash Weekly - November 30th 2022 (Akash Network)

“Guess what — Akash never went down, hasn’t gone down. Only time it was taken down was by its validators during upgrade processes, and that’s expected… when you upgrade it doesn’t take the infrastructure down.” — Greg Osuri, 00:47:32

Context: Answering why validators should leave centralized clouds; he cites ~80 Amazon incidents in a year and frames the choice as “hyper-risky censorship at 80% higher cost” versus Akash.


In 2-3 years, Amazon, Google, and Microsoft using Akash

Akash Weekly - November 30th 2022 (Akash Network)

“If you imagine Akash in two years or three years, I would like to see an Amazon, Google, Microsoft using Akash. I mean, the whole point of web3 is to be inclusive… web3 allows competitors to collaborate… Imagine a Google offering GPUs on Akash — that way they can keep their customer, at the same time offer a competitive GPU pricing. It’s good for Google, it’s good for Amazon. Web3 is good for business.” — Greg Osuri, 01:05:24

Context: Asked how web3 evolves in 3-5 years; notes hyperscalers already stood up web3 teams. Also an early flag of GPU spot markets as Akash’s unique value (“Akash offers a global spot market — nobody in web2 offers a global open spot market”).


In five years we won't be talking about web3 at all

Akash Weekly - November 30th 2022 (Akash Network)

“I think in five years we shouldn’t be talking about web3, because web3 will be so integral to our daily lives that we won’t be having this discussion — it’ll be just something else.” — Greg Osuri, 01:09:04

Context: His consolidation thesis: protocols with “clean, fundamental, simple value” (Arweave’s permanent storage, Akash’s spot market) win and fade into invisible infrastructure; most NFTs won’t survive.


Akash is the first viable open cloud, 97% cheaper

Ask Akash - What problem does Akash solve? Can it scale? (Akash Network)

“Akash is the first viable attempt to decentralizing this oligopolized closed system called cloud computing… the outcome is an amazing non-custodial system where you have complete control of your deployments that’s 97 [percent] cheaper than the cloud.” — Unknown, 00:00:42

Context: Framing answer on what problem Akash solves; pairs with “cloud is the fabric of human society — everyone’s data is stored on the cloud and we have no idea how that’s handled.” Speaker: uncredited Ask Akash answer; likely Greg.


Over 85% of the world's computers are unused

Ask Akash - What problem does Akash solve? Can it scale? (Akash Network)

“The problem we’re solving is improving efficiency of global computation, because our thesis is this enormous amount of unused compute out there — conservatively, over 85 percent of world’s computers [are] not used.” — Unknown, 00:01:26

Context: His recurring quantified underutilization thesis. Speaker: uncredited Ask Akash answer; likely Greg.


The invisible Amazon tax on modern society

Most UNDERRATED Cosmos GEM? Akash Network Deep Dive with Greg Osuri! (Cryptocito)

“Half of every dollar we pay to online services like Netflix and Asana goes to a cloud provider, either Amazon, Google or Microsoft. 80% of that market is controlled by four companies… we pay an invisible Amazon tax. When Amazon goes down, half the internet goes down — that’s how powerful it is.” — Greg Osuri, 00:17:14

Context: He continues: “these companies are not elected, they’re not regulated, they’re not governed by the people, they’re completely controlled by their figureheads” — the civic argument for an open cloud.


Users care about access, not decentralization — GPUs are the wedge

Most UNDERRATED Cosmos GEM? Akash Network Deep Dive with Greg Osuri! (Cryptocito)

“People don’t care about decentralization, people don’t care about control as much as you and I like to think. But what they care is access… even though Akash is like 80% cheaper, it’s not that hard to get a regular computer on Amazon — there’s abundance of it. But I think GPUs are going to be one of those killer use cases for Akash.” — Greg Osuri, 00:21:33

Context: Candid strategy statement: Akash wins where demand outstrips centralized supply (GPUs), not on ideology or price for commodity compute.


Markets reject choke points — the industry will self-correct

Messari Happy Hour Ep11 (Messari)

“You had companies like AOL that actually went ahead and created these choke points, but the market came around and rejected those choke points and it went back to decentralization. Now we are seeing the opposite, where you have heavy concentration of compute power by a few providers… the best solution is not government intervention, it’s really the industry self-correcting.” — Greg Osuri, 00:10:14

Context: Internet history as a cycle of centralization and correction; frames web3 infrastructure as the current correction.


Own your data center, sell the spare capacity

Messari Happy Hour Ep11 (Messari)

“If you want something permanent like a blockchain… you have complete control over cost and complete control over privacy. What Akash comes into play and does is really enables you to have your own data center — that gives you sovereignty and control — and sell any unused capacity you have in the data center to people that want to use it, or tap into a larger pool when you want to scale.” — Greg Osuri, 00:23:54

Context: After contrasting colo economics (capex, 1-3 month acquisition) with cloud premium; the recurring Akash thesis of monetizing owned infrastructure.


The Amazon tax: 50% of online-service spend goes to cloud

Messari Happy Hour Ep11 (Messari)

“50% of every dollar you and I spend on online services like Asana, like Netflix, like Notion goes to Amazon or Google. The Amazon tax — it’s an invisible tax that we all have, and guess who pays the bill.” — Greg Osuri, 00:20:16

Context: Citing a16z’s “The Cost of Cloud, a Trillion Dollar Paradox”; he adds at [00:21:00] that data centers shrank from ~8.4M in 2018 to ~7.1M while hyperscalers grew, because data centers lack cloud-like tooling.


85% of data-center compute is unused

Interview with Greg Osuri of Akash Network (Web3 Working Group)

“As we’re deploying our software to a lot of these data centers, we discovered most of the compute is not used — like 85% of the time a data center is not used.” — Greg Osuri, 00:12:57

Context: The pre-blockchain (2015) discovery that seeded the Akash marketplace idea.


Akash is a "supercloud"

Interview with Greg Osuri of Akash Network (Web3 Working Group)

“The actual term for what Akash does is it’s called a supercloud. Akash is [a] supercloud in the sense that it aggregates various heterogeneous resources to give a homogeneous interface, to make it easy to deploy, easier to manage, easier to secure… across different servers, and it promises sovereignty but at scale.” — Greg Osuri, 00:16:33

Context: His canonical “supercloud” framing — own your data center for baseline, tap the network for burst capacity.


Decentralization lets competitors cooperate

Interview with Greg Osuri of Akash Network (Web3 Working Group)

“There’s no reason [why], if Twitter is doing AI, for them [not] to be able to sell their underutilized capacity to Facebook… and do so in a way that is decentralized and you don’t require trust. So I think decentralization is an amazing way for competitors to cooperate.” — Greg Osuri, 00:36:05

Context: On capital-efficient AI: trustless marketplaces let rival companies monetize idle GPUs to each other.


Half of every dollar for online services goes to three clouds

Interview with Greg Osuri of Akash Network (Web3 Working Group)

“A16z did an amazing blog post analyzing their companies… as well as the public companies… on average, half of every cent, every dollar you spend goes to Amazon, Google and Microsoft on online services… Twitter… burning like 1.5 million dollars a day on the cloud… Asana spends 60% of their margins on Amazon.” — Greg Osuri, 00:07:55

Context: The cloud-tax argument (a16z “Trillion Dollar Paradox”); later adds AWS ~40% overall margins, ~90% on EC2 instances (00:19:27).


Decentralization means removing concentration of risk

Most Successful Crypto Miner Failure, Akash Super Mini - Greg Osuri Interview CEO of Overclock Labs (Action Crypto)

“Decentralization is about removing concentration of risk. Decentralization is about removing attack vectors that compromise the network.” — Greg Osuri, 00:54:30

Context: Recurring thesis, offered while arguing hosting some nodes on AWS is fine — the problem is 70% concentration, not AWS itself; Google and Amazon should eventually be Akash providers.


Marketplaces should be decentralized, companies need not be

Most Successful Crypto Miner Failure, Akash Super Mini - Greg Osuri Interview CEO of Overclock Labs (Action Crypto)

“Any marketplace should be decentralized. Any value exchange I believe should be decentralized, because marketplaces optimize for perfect competition.” — Greg Osuri, 00:15:06

Context: Asked where the line between centralization and decentralization should be drawn; he adds “the companies I don’t believe should be all decentralized because without those companies you don’t have the markets” ([00:16:30]).


Decentralization means removing systemic failure risk

Akash Network Live with Greg Osuri: Akash's progress in 2023, open community development, and more (Akash Network)

“It’s really about removing the risk of a system failure — removing systemic failure risk… Today you have centralized systems like Amazons and Googles of the world: if the company collapses, the product [dies], no matter how important it is to you or to the society. If the company dies, [the] product [dies].” — Greg Osuri, 00:12:15

Context: His onion model of protocol attack vectors — consensus, governance economics, company influence (citing Terra), and lab dependence — which motivated the radically-open reorganization; the bracketed words are garbled in the auto-captions (00:12:58).


The internet I want: sovereign, private, decentralized, efficient

Greg Osuri, founder of Akash, answers the question "What Kind of Internet Do You Want?" #shorts (Web3 Working Group)

“The internet that I want needs to be sovereign, needs to be private, needs to be decentralized, and needs to be efficient. That’s the internet that I want… it’s a foundation.” — Greg Osuri, 00:00:00

Context: His stock closing answer to the series’ signature question; a compact statement of the values behind Akash.


A thousand providers within 12 months; more distributed than Amazon

Mission: DeFi EP 92 - Greg Osuri - Akash is taking on the giants in hosted processing (Mission: DeFi)

“It is the largest market — we have about 50 providers right now… and that’s growing very quickly, and with incentives we expect at least a thousand provider footprint in the next 12 months… it will be the most distributed network. I think from a regional distribution, Akash is a lot more distributed than Amazon’s… the way we’re trending, it’s only a matter of time.” — Greg Osuri, 00:13:31

Context: Right after the Airbnb-vs-Hilton analogy for Akash vs AWS; a quantified 12-month forecast made March 2023. He later notes low-end pricing “85% cheaper than Amazon” with high-end options 2-3x more (00:17:05).


A viable open-source cloud already exists

Where Crypto and AI Meet | Featuring Akash, Bittensor, Gensyn & DCG (April 26, 2023) (The Bittensor Hub)

“When I talk to developers, people don’t realize we actually have a viable open source cloud… I think that’s a big deal for me and a big deal for, I think, the world.” — Greg Osuri, 00:45:14

Context: Closing “what excites you most” question; he adds you can “run an Alpaca on a fully decentralized network and pay using a token without giving your email address.”


Cloud is the fabric of society and must be publicly governed

AI SUPERCLOUD with Greg Osuri of Akash Network (The Interop)

“If you consider cloud to be the fabric that holds a society, we need this fabric to be open and we need this fabric to be governed in the public and resources allocated in the public, rather than a single corporation doing it in a completely opaque manner.” — Greg Osuri, 00:14:00

Context: Closing his three-part argument (cloud margins, open-source stifling, antitrust/censorship) for why Akash needs to exist.


Hyperscalers will route users through Akash to competitors' GPUs

AI SUPERCLOUD with Greg Osuri of Akash Network (The Interop)

“In the future you won’t be surprised to see an Amazon user using a GPU on Google, paying using their Amazon account, and actually not leaving Amazon… now Amazon has a way to retain the user… and actually work with a competitor — and Akash is the bridge.” — Greg Osuri, 01:20:52

Context: Answering an audience question about conversations with Google/Amazon; says web3 teams at both view Akash as “a critical piece of the future of the cloud” (he explicitly denies any Microsoft talks). “Akash is the bridge” lands at 01:21:35.


In a year or two Akash becomes the secondary market for cloud-grade compute

AI SUPERCLOUD with Greg Osuri of Akash Network (The Interop)

“I think in a year or two timeframe you’re going to see quite a lot of [hyperscaler] participation… and as that progresses it’s not hard to see how Akash becomes a second market for cloud-grade compute.” — Greg Osuri, 00:31:16

Context: He argues the supercloud needs participation from “the big guys” or it stays a fringe network; says cloud-provider employees privately acknowledge the problem.


Post-parity: 10,000 providers in every city, workloads that follow the user

AI SUPERCLOUD with Greg Osuri of Akash Network (The Interop)

“The first major milestone for Akash Network is parity with the cloud… post parity, imagine having 10,000 providers in every city of the world — GPU, CPU, whatever resource — imagine being able to provision a workload that follows your user no matter where they go.” — Greg Osuri, 01:09:28

Context: Answer to the long-term-vision question; roadmap is cloud parity first, then hyper-distributed compute.


Centralized systems will absolutely fail at this scale

Mission: DeFi EP 98 - Can Akash be a major player in AI with their GPU market? Founder Greg Osuri (Mission: DeFi)

“That’s the kind of future that we’re looking at, and it’s only possible for a decentralized system, because centralized systems will absolutely fail. There’s no way, at the scale which we talked about, a centralized system can function… it has to be decentralized from the get-go — the core has to be decentralized.” — Greg Osuri, 00:40:28

Context: Concluding the home/grid compute vision; his argument for why the sharing layer must be decentralized from day one rather than retrofitted.


Cloud is modern society's fabric, 80% controlled by four companies

Mission: DeFi EP 98 - Can Akash be a major player in AI with their GPU market? Founder Greg Osuri (Mission: DeFi)

“The whole point of Akash is to remove the heavy concentration of power… compute cloud today is a fabric of modern society — our data, your data, everything runs [on it] and there’s no going back… the concentrated compute power: Amazon, Google, Microsoft control over 80% of the market today.” — Greg Osuri, 00:29:45

Context: Answering whether Akash’s role is to decentralize access to GPU power; same riff opens the episode as the cold-open clip. He adds startups can’t get 10 H100s at $3/hr on AWS or GCP without ~$3M commitments.


Akash is the Airbnb layer for cloud capacity

AI is using chips faster than Nvidia can make them. How crypto is helping | Akash and Render Token (Coinage)

“Because Airbnb is a network that connects unused space and offers a price point that [is more] attractive, so Akash is that sort of layer that is very similar to Airbnb for developers that host cloud data.” — Greg Osuri, 00:02:55

Context: Answering the segment’s “Airbnb for computing power” framing of data centers as hotels.


Hyperscalers will publicly work with Akash; Akash as a global compute standard

ADRIAN STEWART Akash $AKT Interview Greg Osuri #AI #GPU #AISupercloud ElonTrades Cosmos ATOM (Louisiana Swamp Rat)

“Big hyperscalers are looking at [Akash] and finding ways to, like, figure out great ways to work with [Akash] publicly… I think it’s going to be [Akash] just as a global standing of how people use compute… it’s going to be like AWS support[ing it], for example.” — Greg Osuri, 00:37:59

Context: Cites AWS’s head of blockchain talking about Akash; predicts hyperscalers reselling Akash capacity (“it’d be funny if Google users use Amazon”). Captions heavily garbled here — bracketed words restore the clear referent.


99% of Akash code will come from outside the issuer

The GPU crisis that AI needs solved now. W/ Greg Osuri, CEO Overclock Labs Akash Network (Parks and Decentralization)

“90% of contributors come from outside Overclock Labs — that means outside the issuer… in a year or two years time maybe 99% will come from outside Overclock Labs, and that’s the goal we are working towards.” — Greg Osuri, 00:30:11

Context: His “radical decentralization” answer to SEC risk: ~280 developers across 10-15 organizations, only ~20 employed by Overclock; paired with his gold-vs-Apple-stock value-accrual test for what makes a security.


Cloud is the new telco

The GPU crisis that AI needs solved now. W/ Greg Osuri, CEO Overclock Labs Akash Network (Parks and Decentralization)

“Government has to step in and break them down — like break AT&T down — without which we wouldn’t have the internet. Think about it. We are similar in cloud infrastructure: cloud is the new telco and it’s heavily gated.” — Greg Osuri, 00:10:51

Context: Comparing the AWS/Google/Microsoft/Alibaba oligopoly to the pre-breakup telephone monopoly; notes the US government was investigating AWS for anti-competitive practices.


The 50% invisible cloud tax

The GPU crisis that AI needs solved now. W/ Greg Osuri, CEO Overclock Labs Akash Network (Parks and Decentralization)

“Cloud is so integral and so invisible in our lives that you and I, we pay a cloud tax like it or not… if you use Netflix, if you use Dropbox, if you use Gmail… 50% of what you pay goes to an invisible tax collector called a cloud.” — Greg Osuri, 00:10:08

Context: Recurring quantified claim about consumer exposure to cloud oligopoly pricing.


8.4 million data centers, mostly idle

RNDR, Supercloud, 1.5T AI Market Cap – Interview with Greg Osuri about Akash (Design DAO)

“Most data centers — there are about 8.4 million data centers in the world — most data centers are not used to the point they’re supposed to be used: 85 percent of time they’re unused. If we can leverage the supercloud and connect these data centers and actually get a price point that’s significantly lower than what you would otherwise pay to these hyperscalers… that’s what we’re seeing with Akash.” — Greg Osuri, 00:16:28

Context: His recurring utilization argument; he adds it’s “no longer just cheap cloud, it’s about access” to GPUs that cloud providers selectively allocate.


The invisible cloud tax

RNDR, Supercloud, 1.5T AI Market Cap – Interview with Greg Osuri about Akash (Design DAO)

“Andreessen Horowitz wrote a very thorough, thoughtful research piece where they estimated, backed by data — they pointed out over 50 percent of the margins that these online services get goes to a cloud [provider]. If you use Dropbox or Asana or Netflix, 50% of… the profits are going to go into the cloud providers. So there’s an invisible tax we pay, like it or not.” — Greg Osuri, 00:14:23

Context: Explaining cloud lock-in (BigQuery, Aurora) as the motivation for a homogeneous supercloud interface over heterogeneous providers.


We called the supercloud five years early

RNDR, Supercloud, 1.5T AI Market Cap – Interview with Greg Osuri about Akash (Design DAO)

“Akash Network is a first decentralized supercloud. The idea of a supercloud… we introduced in our white paper in 2018. Now… if you Google supercloud you’ll see Forbes, you’ll see all these mainstream media writing about supercloud — really they discovered like last year, but we were pushing the idea for supercloud for five years.” — Greg Osuri, 00:11:30

Context: Builds on Cornell’s 2015 supercloud paper; he repeats at [00:57:00] “we know supercloud is going to be a thing, that’s why we wrote the paper five years ago… we’re still very early.”


First-ever open market for GPUs

Akash Mainnet 6 Livestream (Akash Network)

“The biggest and the most exciting feature is the GPU marketplace. This is, I believe, the first instance of an open market for GPUs. There is no open source market right now available, centralized or decentralized, so this is the first time ever the world is experiencing what an open market for GPUs looks like, and it is coming amid the biggest supply crisis we have for GPUs.” — Greg Osuri, 02:49:28

Context: Pre-upgrade feature overview; he frames the GPU supply crisis as “a symptom of inefficient distribution” where cloud providers failed, creating the opening for an open network.


8.4 million data centers, 85% unused

"#3 - Akash Network with Greg Osuri" (Weapons of Mass Adoption)

“Most data centers, there are about 8.4 million data centers in the world. Most data centers are not used to the point they’re supposed to be used. 85% of the time, they’re unused. So, if you can leverage the super cloud and connect these data centers and actually get a price point that’s significantly lower than what you would otherwise pay to these hyperscalers.” — Greg Osuri, 00:16:44

Context: Quantifying the latent-supply thesis, then pivoting: “It’s no longer just cheap cloud. It’s about access… to the GPUs” ([00:17:27]).


The invisible 50% cloud tax

"#3 - Akash Network with Greg Osuri" (Weapons of Mass Adoption)

“Anderson Horowitz wrote a very thorough, thoughtful research piece… they pointed out over 50% of the margins that these online services get goes to a cloud provider. Like, if you use Dropbox or Asana or Netflix, 50% of the payment on the profits are going to go into the cloud providers. So, there’s an invisible tax we pay like it or not.” — Greg Osuri, 00:14:24

Context: Citing a16z’s “The Cost of Cloud” piece while explaining supercloud economics and BigQuery/Aurora lock-in.


"We knew supercloud was going to be a thing" — five years early

"#3 - Akash Network with Greg Osuri" (Weapons of Mass Adoption)

“We stuck to our values and stuck to our thesis. Akash is exactly what it is in the white paper we wrote… Narratives are like, yeah, I’m going to take off. Like, yeah, we know. We knew that five years ago, right?… we know SuperCloud is going to be a thing.” — Greg Osuri, 00:56:31

Context: On why Akash “didn’t take off” in the 2021 DeFi/NFT cycle — no narrative fit plus Cosmos liquidity constraints; he frames the AI narrative finally aligning with the original thesis.


Open systems go far, closed systems go fast

Building the Super Cloud of GPUs with Akash Founder Greg Osuri | EP #102 (Frictionless Podcast by Logan Jastremski)

“Decentralizing and open systems can go far, whereas centralizing closed systems can go fast… that’s why I believe open and decentral systems will reach [places] where centralized systems just cannot.” — Greg Osuri, 00:24:36

Context: His recurring coordination-cost argument for why the demand-supply connection for idle GPUs should be a decentralized marketplace.


Next milestone: cloud parity

Akash Network - Chat With a Founder Greg Osuri (Don Cryptonium)

“The next big milestone for Akash is called cloud parity. That means we can get to a point where we can do… anything possible that’s on the cloud today.” — Greg Osuri, 01:02:45

Context: Asked about the next few years; the theme is “adoption, distribution” with a claimed 98% conversion rate among AI devs shown Akash.


Akash's one thing Amazon doesn't have

Greg Osuri: Redefining Success Against All Odds | BTC Ep.7 (Michael Huynh)

“We technically are an alternative to Amazon Web Services, but we don’t have all the bells and whistles… we have one thing that Amazon doesn’t: that’s open source, and that translates to everything we need. Now we have GPUs — the high-end GPUs Amazon doesn’t have. It’s really about identifying the one thing that nobody else has and you have, and getting 100% behind the one thing.” — Greg Osuri, 00:54:31

Context: Applying Bruce Lee’s “one kick a thousand times” to startup strategy — one differentiating feature beats a thousand features.


Community is Akash's competitive moat

Greg Osuri: Redefining Success Against All Odds | BTC Ep.7 (Michael Huynh)

“I keep saying community is the competitive moat for Akash. It’s one thing that no one else has — no big cloud provider has a large, open community.” — Greg Osuri, 00:55:18

Context: Closing point: Amazon can’t scale the way an open community can; his recurring argument for why open-source networks outlast corporate clouds.


Everyone pays an invisible cloud tax

How Akash Network is Democratizing Access to Cloud GPUs | Greg Osuri (0xResearch)

“All of us pay a cloud tax — you and me, we just don’t realize it. For context, next [nearly] 50% of all the subscription fees you pay for online services goes to one of the cloud providers, you just don’t know it.” — Greg Osuri, 00:02:11

Context: His recurring opening argument: cloud is an oligopoly — “an extremely important layer that is extremely opaque” — and Akash’s mission is to reverse that opacity.


Cloud is a public utility — people first, not shareholders

1on1 Greg Osuri - Akash (Jerry V Hall)

“A corporation is going to optimize for corporation; community is going to optimize for [community]. That’s why decentralization leads to communal public utilities… if you consider cloud to be a public utility… it is very, very important that it is decentralized.” — Greg Osuri, 00:16:13

Context: The promise of decentralized networks is reversing hyper-centralization by giving cloud users ownership and equity; cites Akash being 70-80% cheaper with chips unavailable on the cloud.


Post-parity: 10,000 regions instead of Amazon's 30

Akash: Crypto's AI Supercloud W/Greg Osuri ($1 To $1 Million Podcast)

“Once we get to parity we get to post-parity, which I call innovation beyond imagination. You and I cannot sit here and imagine what kind of innovation [there] will be… Amazon has about 30 regions right now; instead of 30 regions you’re going to have 10,000 regions.” — Greg Osuri, 00:59:09

Context: Roadmap framing (pre-parity vs post-parity with cloud). He later explains 10ms latency enables applications impossible today (cloud gaming, distributed ML) since “Akash essentially makes any computer a data center” (01:04:05).


Open systems go far, closed systems go fast

Akash Network founder Greg Osuri - Building a Decentralized Computing Marketplace (Crypto Nuggets)

“The whole notion of a decentralized system is not how fast you can move but how far you can go… with open systems you can go far, while with closed systems you can go fast. I’m not here to build closed systems.” — Greg Osuri, 00:52:09

Context: His recurring argument for why decentralized coordination costs are worth paying — network effects compound; Akash had nine chain upgrades that year without a security incident.


Win state: hyperscalers integrate Akash

Akash Network founder Greg Osuri - Building a Decentralized Computing Marketplace (Crypto Nuggets)

“Our best-case scenario is when Google, Microsoft and Amazon integrate Akash on the demand side — that’s our ideal win state… we do have some efforts in place to attract the big cloud onto Akash.” — Greg Osuri, 00:45:36

Context: Describing the three-sided ecosystem strategy (demand, supply, crypto); precursor to his later “hyperscalers will join Akash” refrain.


Supercloud thesis: decouple control plane from resource plane

Akash: $300 Million Incentives w/Greg Osuri | CEO of Overclock Labs,Founder & Akash Network (Bare Metal Podcast)

“The idea of supercloud was just getting formulated — I think Cornell wrote a paper in 2015 — and we really liked the idea of supercloud, that you can decouple the control plane from the execution plane or the resource plane, and reduce the cloud providers, both public as well as private, to resource providers and not give away control.” — Greg Osuri, 00:03:36

Context: Origin story of Overclock Labs/Akash; he analogizes hyperscaler cloud to a rental-car oligopoly versus ownership.


Akash reaches cloud parity within two years

"Akash: The Crypto-Powered Decentralized Supercloud" (The Edge Podcast)

“Every time we launch, we’re very, very, very cognizant about the value proposition and how we differentiate ourselves… So we believe we’ll achieve parity in the next two years. And after that, the phase we like to call is really innovation beyond imagination.” — Greg Osuri, 00:17:20

Context: “Parity” = matching traditional clouds’ critical capabilities (observability, automation, etc.); dated January 2024.


Permanency is only possible with a protocol

"Akash: The Crypto-Powered Decentralized Supercloud" (The Edge Podcast)

“Akash five years down the line and be decentralized, gives Akash a unique angle that traditional companies doesn’t, which is permanence. Like permanency is not possible with the corporation, only possible with a protocol… with permanency, you get time that you don’t get with traditional corporations.” — Greg Osuri, 00:18:46

Context: Expanded later at [00:40:00]: software lost permanency in the web era, “with crypto… the creator or the company… can go away and the software still is alive. Look at Bitcoin, right? It’s not going to die.”


Akash as a supercloud over any cloud-capable computer

Revolutionizing GPU Access for AI Workloads | Greg Osuri - Akash Network (ETHDenver)

“Akash is first and foremost a supercloud. A supercloud is a cloud that sits on top of private, public, hybrid — doesn’t matter — any cloud-capable computer, and gives a familiar user interface… whereas the compute comes from all over the place using a marketplace mechanism.” — Greg Osuri, 00:06:36

Context: His recurring supercloud definition, here emphasizing marketplace bidding between tenants and providers.


Akash is the world's first decentralized supercloud

Decentralizing AI: Compute Protocols and Proof of Work - Panel Discussion | Proof of Data 2024 (Recall Network)

“Akash is world’s first supercloud that’s decentralized. A supercloud is essentially a cloud that sits on private, public, hybrid, single computer, and effectively provides you capabilities and features that would look like a normal cloud, and with a marketplace where you can source resources from places that you normally can’t find in your typical cloud marketplaces.” — Greg Osuri, 00:01:27

Context: His self-introduction on the panel; canonical statement of the supercloud framing.


If we don't decentralize, we die as a society

Decentralizing AI: Compute Protocols and Proof of Work - Panel Discussion | Proof of Data 2024 (Recall Network)

“We are in a very crazy political and social-economical environment. If we don’t decentralize, we will die as a society, and feudalism over freedom is going to prevail — and that’s the reality we are faced with right now.” — Greg Osuri, 00:08:38

Context: After arguing cloud innovation has stalled (the promised edge computing and 10ms latency never arrived because hyperscale was more profitable) and censorship is “getting out of control.”


The cloud tax is creating a feudal society

Decentralizing AI: Compute Protocols and Proof of Work - Panel Discussion | Proof of Data 2024 (Recall Network)

“What that is trending towards is a feudalistic society that we work so hard as a humanity to escape — few people with the resource that everyone needs have control over those resources and charge a tax. All of us pay a cloud tax without even realizing we are paying a cloud tax: an average 50% of all the income that online services make go to a cloud provider, and these providers are unelected companies with shareholders, profits… as a primary driver.” — Greg Osuri, 00:07:11

Context: Answering why decentralized compute matters now; he adds that with hyperscalers as GPU gatekeepers, “try getting a GPU from Amazon — it’s impossible.”


Akash adoption: HIPAA data, NSA/DoD, IP-address marketplace

Agents Unleashed - Progress In The Decentralized AI Stack (Training, Fine-Tuning, Inference) (Olas)

“Not too many people know Akash is one of the largest IP address marketplaces… Akash is used by healthcare companies to store HIPAA compliant data. Akash is used by NSA — they’re doing a project inside the Department of Defense… Akash is so robust and so simple, it is composable, part of a larger protocol.” — Greg Osuri, 00:19:36

Context: Arguing that doing one thing well (“Akash is a supercloud”) plus simplicity drives adoption; “more things we do as a protocol means bigger attack surface… more brittle.”


Akash is the world's first open supercloud

AtomDenver panel: AI 🤝 Crypto (Secret Network)

“Akash is the world’s first open supercloud. A supercloud is a cloud that sits on private, public — doesn’t matter — a cloud that essentially decouples the control plane from the resource plane… to take advantage of resources that are globally distributed.” — Greg Osuri, 00:02:17

Context: His self-introduction; the recurring “supercloud” framing of Akash’s architecture.


Permissionless censorship resistance is the core guarantee

Agents Unleashed - Progress In The Decentralized AI Stack (Training, Fine-Tuning, Inference) (Olas)

“Protocols or systems that build upon these simple, permissionless, censorship-resistant primitives is key, because when you build on Akash the guarantee you get is: you own it, no one’s going to stop you, and you can change it.” — Greg Osuri, 00:20:19

Context: His recurring thesis on why open infrastructure beats centralized clouds; “we remove the friction for coordination across the globe.”


A new operating system for the internet on 85% unused capacity

Overcoming the Challenges of Distributed Networks with Greg Osuri CEO of Overclock Labs & Akash N... (Techpost)

“We believe in the unstoppable web, and what we’re doing with Akash is essentially creating a new operating system for the internet. And how we’re doing that is by tapping into unused 85% capacity in 8.2 million data centers… Akash is essentially a peer-to-peer network of computation clusters designed for global-first applications, planetary-scale applications.” — Greg Osuri, 00:02:59

Context: His standing framing of the Akash thesis, repeated twice in the episode; anchored to internet-era permissionless principles (“you don’t need anybody’s permission to go to a website or host a website”).


What AWS did for Amazon, Akash does for the world

Hash Rate - Ep 050 - Akash Decentralized Cloud - Greg Osury (Hash Rate Podcast)

“What AWS did for Amazon is what Akash is doing for the rest of the world.” — Greg Osuri, 00:02:52

Context: After the TurboTax/holiday-surge analogy: enterprises use ~5% of their capacity, and Akash monetizes the idle 95% just as Amazon monetized its post-dot-com overbuild.


Cloud parity: moving Walmart-scale workloads off AWS

Akash Accelerate '24: Official Livestream (Akash Network)

“The goal really is to get to cloud parity, in the sense in a few years from now we should be able to go to Walmart or some large company and be like, hey, you can move all your stuff from Amazon to Akash. We’re not there yet — we need to build a lot of capabilities.” — Greg Osuri, 00:27:45

Context: Roadmap section of the keynote (reserved instances, managed backends, access control, secrets management as the missing pieces).


Regulation is the only thing that can stop Akash

Akash Accelerate '24: Official Livestream (Akash Network)

“There are lots of risks, but I think the biggest one would be — Akash is too early right now to be shut down by regulation. I think if regulation is out of the way, I think Akash is unstoppable. But if regulation comes and basically says you cannot exist, I think that’s going to be a big problem.” — Greg Osuri, 01:34:27

Context: Audience question “what’s the biggest risk Akash faces” during the Voorhees coffee chat; Greg adds “if you’re a builder you can build as long as you’re allowed to build — we’ll build our way out of any problems.”


The 2018 supercloud thesis

Akash Accelerate '24: Official Livestream (Akash Network)

“We wrote a paper… the subsequent version, 2018, that introduced a notion of a supercloud — the Akash supercloud platform — where we said, well, you can decouple the control plane from the resource plane, where the resources could effectively be any cloud-capable compute.” — Greg Osuri, 00:08:21

Context: Keynote history segment; Greg credits Cornell with formalizing the “supercloud” term and claims Akash’s whitepaper articulated it in 2018.


Centralized control planes can't scale — the internet's decentralized roots are the answer

Akash Accelerate '24: Official Recap (Akash Network)

“We quickly realized that you cannot scale to any level of high performance using a centralized control plane. The only way to really massively scale is go back to the roots, how internet was founded — a decentralized network.” — Greg Osuri, 00:05:08

Context: Keynote origin story: why Overclock pivoted from centralized cluster orchestration to blockchains (the “only data structure” enabling decentralized coordination) in 2016–2017.


Cloud parity: moving Walmart off Amazon "in a few years"

Akash Accelerate '24: Official Recap (Akash Network)

“The goal really is to get to cloud parity, in the sense, in a few years from now you should be able to go to Walmart or some large company and be like, hey, you can move all your stuff from Amazon to Akash. We’re not there yet — we need to build a lot of capabilities.” — Greg Osuri, 00:25:48

Context: Roadmap section of the keynote (reserved instances, managed backends, access control, secrets management as the gaps to close).


Centralized control planes can't scale — the internet's roots are the answer

"Keynote from Greg Osuri: \"Journey to Acceleration & Beyond\" - Akash Accelerate '24" (Akash Network)

“You cannot scale to any level of high performance using a centralized control plane. The only way to really massively scale is go back to the roots, how the internet was founded — decentralized… instead of a single control plane we can have a decentralized control plane where there’s no dependency on a single point of failure… blockchains happen to be the only data structure that would allow us to have a decentralized control mechanism, or decentralized coordination mechanism.” — Greg Osuri, 00:05:07

Context: Explaining how a Kubernetes-tooling company “stumbled upon blockchains” in 2016-17 — his foundational architectural thesis.


First general-purpose decentralized compute network

"Keynote from Greg Osuri: \"Journey to Acceleration & Beyond\" - Akash Accelerate '24" (Akash Network)

“We launched the mainnet in 2020 — first time ever a decentralized network could launch general purpose compute. There was nothing like it. The only thing that was close to Akash was Golem, and Golem was not general purpose… Akash was the first general purpose compute [where] you can actually demonstrate a Docker container running in a decentralized setting… they call it the Airbnb for data centers, without the cleaning fee.” — Greg Osuri, 00:09:19

Context: Positioning claim about Akash’s place in decentralized-compute history.


Goal: cloud parity — move Walmart off Amazon

"Keynote from Greg Osuri: \"Journey to Acceleration & Beyond\" - Akash Accelerate '24" (Akash Network)

“The goal really is to get to cloud parity, in the sense that in a few years from now you should be able to go to Walmart or some large company and be like, hey, you can move all your stuff from Amazon to Akash. We’re not there yet — we need to build a lot of capabilities.” — Greg Osuri, 00:25:46

Context: Setting the roadmap horizon (reserved instances, managed backends, access control, secrets management); he stresses Akash never pivoted — the 2019 roadmap already listed private containers.


1,000 providers and 10ms latency for 98% of the internet

A Decentralized Compute Marketplace with Greg Osuri (Software Engineering Daily)

“We want to be at least 500 providers by end of the year; end of next year we want to have at least a thousand providers… the goal is to get to 10 millisecond latency for 98% of active internet population… the kind of stuff you can do with such a low-latency network is beyond imagination.” — Greg Osuri, 00:45:43

Context: His “cloud parity” roadmap — managed services plus a globally distributed provider base; quote spans into the [00:46:28] block.


Data-center concentration is reversing; no winner-take-all for hyperscalers

"Greg Osuri: Akash – Decentralizing Cloud Computing and Revolutionizing GPU Access (#10)" (Fluence)

“There was like 8.2 million data centers 10 years ago, which reduced to 7.5 [million] now, and increasingly concentration happening… but that’s reversing. We’re seeing smaller data centers now that are popping up everywhere with GPUs on them. So I don’t think hyperscalers are going to be the winner-take-all in this market.” — Greg Osuri, 00:37:48

Context: On whether cloud giants will eventually absorb GPU supply; he credits Nvidia’s self-interest in fragmenting distribution.


Largest distributed network: 500-600 new regions by year-end

A Decentralized Compute Marketplace with Greg Osuri (Software Engineering Daily)

“It’s already the largest distributed network in terms of regions — it’s got more regions than Amazons and Googles of the world. By end of the year we anticipate at least 500 to 600 new regions, so by end of the year it’ll be the largest distributed network in terms of physical distribution. So you can imagine the edge computer applications that can be finally enabled on Akash.” — Greg Osuri, 00:07:52

Context: Describing network scale (~70+ providers growing ~20/month at recording); a dated, quantified 2024 forecast.


Airbnb for data centers with trust via decentralized auditors

Tech Snippets Today - Greg Osuri - Founder at Akash with Joseph Raczynski (Joseph Raczynski)

“The best analogy for Akash is Airbnb for data centers… you have about 80 providers now, all different sizes and shapes… and the way you know someone is telling the truth is through auditing. There is a decentralized network of auditors… [that] verify this compliance offline and post a proof online… Akash gives you more trust than traditional Amazon because Amazon will not show you the certificates.” — Greg Osuri, 00:07:55

Context: Answer to the classic enterprise-security objection to decentralized cloud; claims proof-backed visibility beats hyperscaler opacity.


Akash as the world's first supercloud, spanning data centers and homes

Tech Snippets Today - Greg Osuri - Founder at Akash with Joseph Raczynski (Joseph Raczynski)

“Akash is the world’s first supercloud. What a supercloud is, is essentially a cloud that connects any cloud-grade compute — be it in a data center, be it in a home, or be it in a private setting or a public setting — and brings [it] together in a way that’s easily consumable by developers around the world.” — Greg Osuri, 00:00:02

Context: Opening answer to “what are you working on”; the supercloud framing explicitly includes home compute as first-class supply.


Everyone pays an invisible cloud tax — 50% of online-service revenue goes to the cloud

"Greg Osuri | Founder of Akash Network: Liberty, Community, and the Future of Decentralized AI" (Beacon Podcast)

“Cloud infrastructure, which arguably is one of the most important layers that holds humanity together… It is an invisible layer that we don’t talk about much. I mean, all of us pay a cloud tax. 50% of all the monies we pay to an online service goes to a cloud.” — Greg Osuri, 00:13:25

Context: His recurring “cloud tax” argument, here with an unusually high 50% figure (elsewhere he has cited lower numbers); leads into why an open-source cloud is needed.


Open-source clouds failed until crypto could attach resources to software

"Greg Osuri | Founder of Akash Network: Liberty, Community, and the Future of Decentralized AI" (Beacon Podcast)

“We tried open source cloud several times before, and none of them really took off or succeeded or even survived because they didn’t have the resources that you can actually acquire through open source. So for the first time, I had an opportunity to create not only open source software, but also give access to the resource in an open manner.” — Greg Osuri, 00:15:07

Context: His thesis that a token marketplace is what makes open infrastructure sustainable — “software as a coordination layer and the hardware as the value” (00:14:08). Follow-on claim at 00:16:01: 500+ contributors, and Overclock Labs could disappear with Akash unaffected.


Post-parity Akash: 1,000+ regions at cell towers with 10ms latency

"Greg Osuri | Founder of Akash Network: Liberty, Community, and the Future of Decentralized AI" (Beacon Podcast)

“A big portion of the roadmap current focus is to get to parity to Amazon web services. Post-parity is when, I think, Akash will start really seeing stuff that are not possible… How about, you know, 1,000 plus regions close to cell phone towers offering 10 millisecond latency…? What kind of possibilities do you unlock when you have such high-performance cloud? Well, you can do cloud gaming… that’s just the tip of the iceberg.” — Greg Osuri, 00:19:49

Context: His two-phase framing of Akash’s future (quote spans into the 00:20:33 block): pre-parity serves what it can today (GPUs); post-parity unlocks hyper-local edge compute impossible for centralized clouds.


Web3 restores the liberty the web lost

"Greg Osuri | Founder of Akash Network: Liberty, Community, and the Future of Decentralized AI" (Beacon Podcast)

“What immediately comes to my mind when you say about Web3 is liberty. Right? The liberty that the Web originally promised but somehow got lost in the way and now we are bringing back the original values… Liberty, sovereignty, freedom.” — Greg Osuri, 00:00:51

Context: Answer to the opening “why Web3?” question; he adds that permissionless composability makes systems “a lot less fragile and a lot more reliable.”


Akash as a multi-trillion-dollar internet of compute

Akash: The Internet Of Compute W/ Founder Greg Osuri ($1 To $1 Million Podcast)

“If Akash is successful, this is a multi-trillion dollar network you’re talking about — like an internet of compute… It’s a bold vision, but if they’re successful you guys are thinking too small.” — Jesse, 00:53:46

Context: Speaker: the host (Jesse), not Greg — he tells listeners who say Akash is “already too big” that the upside is a multi-trillion-dollar network; Greg does not dispute the framing.


Decentralizing compute is the wedge for decentralizing the whole stack

Akash: The Internet Of Compute W/ Founder Greg Osuri ($1 To $1 Million Podcast)

“If we can decentralize at least the compute layer, we have a much better chance of decentralizing the entire stack. Versus if you don’t decentralize this base-level layer, we’re going to be having trouble.” — Greg Osuri, 00:47:18

Context: Explaining why both decentralized (Venice, Nous, Prime Intellect) and centralized stacks (Brev) building on Akash matter for long-term demand.


Akash never breaks — more alive than AWS

[Solo Talk] Decentralization is saving AI where centralization failed by Greg Osuri (Kryptoplanet[Official])

“I’m always confident because Akash never breaks. It never does. It’s been around for 4 years. It is more alive than Amazon Web Services, and I take that.” — Greg Osuri, 00:15:48

Context: Setup for the live demo — deploying Mistral on an A100 from his phone on stage in about 90 seconds.


Compute controlled by unelected few is not the future

"[Panel] ​“Decentralized Computing as a Business”" (Kryptoplanet[Official])

“Really founded Akash to democratize cloud and compute, because compute is the fabric that connects the modern society, and that controlled by a few people that are unelected, that take tax from us, is not a future that I want for me.” — Greg Osuri, 00:02:53

Context: Greg’s self-introduction; he also credits Cornell for the supercloud idea and calls Akash “the world’s first supercloud.”


No other way to scale globally than decentralization; hyperscalers keep failing

More than Decentralised Compute | Converge @ EthCC 2024 (Caladan)

“When you remove the need for a centralized scheduler or a centralized control plane… and you go to a fundamental peer-to-peer manner, you can scale unlimited… One of the reasons Akash has been around for 4 years and hasn’t had a single outage except for the planned upgrades is because it’s foundationally [decentralized]… During last one year Amazon went down 100 times — when it goes down you can’t order your Uber or your DoorDash. Is that the foundation that we want to build our AI [on]?… From a technological standpoint, I think there is no other way to scale at a global scale other than decentralization.” — Greg Osuri, 00:08:35

Context: Second pillar of his argument; he adds he confidently demos live on stage because “this will never fail.”


The cloud is an unelected tax — "we need decentralization to avoid feudalism"

More than Decentralised Compute | Converge @ EthCC 2024 (Caladan)

“The a16z did a study… that said something like 50% of your money that you pay for online services goes to a cloud provider that you have no idea who it is. Compute is the fabric that holds a modern society together, and compute is foundation for AI. If you think that you’re going to have to pay a tax to someone that you have no idea who it is, you’re not elected — we used to do that when society was feudalistic. So we need decentralization to avoid feudalism, that’s the simple answer. And we believe compute should be a common good; it should not be in the hands of few.” — Greg Osuri, 00:07:12

Context: Answering “why do we need decentralized compute?” — his recurring feudalism framing plus the demand for permissionless access (“a world that I’m not willing to accept” is needing massive deals with cloud providers).


A ~2-year window to decentralize before hyperscalers regain the advantage

BASS 2024 10 Ben Fielding, Greg Osuri, Guy Wuollet, Steven Willinger (Stanford Blockchain Club)

“We’re going to fix the supply chain problem at some point… but once the supply chain is fixed, you get the economy of scale that large hyperscalers have that may overpower the network. So I think we have maybe about 2 years to really decentralize, because we have one thing the hyperscalers don’t: Akash is the only place to get on-demand H100s right now — and that’s going to be not forever.” — Greg Osuri, 00:10:24

Context: Answering whether decentralized networks can compete with AWS/Google once GPU scarcity ends; he frames crypto incentives and gamification as the tools to grab distributed supply during the window.


Akash has run four years without a single outage

Akash Network - The Decentralized Compute Marketplace (The Crypto Conversation (Brave New Coin))

“It’s been online for about four years, has not gone down a single time, like not a single time. Amazon goes down 200 times a year, right?… because Akash is open source, because you have so many eyeballs watching this code base 24/7, the chances of bad code getting in thereby causing instability is very, very low.” — Greg Osuri, 00:19:50

Context: “Power of the crowd” differentiator, citing CrowdStrike’s bad-code outage as the centralized counterexample and the South Korean security firm that caught an Akash bug pre-exploit (00:20:31).


Akash is the first implementation of an open source supercloud

Decentralized solutions for cloud computing and AI 📌 Greg Osuri, Akash @ DePIN Day Austin (Fluence)

“The white paper we wrote in 2017, 2018… introduced the notion of a supercloud. Supercloud, the term was coined by Cornell in 2015, but we implemented the supercloud. So Akash is the first implementation of an open source supercloud.” — Greg Osuri, 00:04:22

Context: Historical claim anchoring the supercloud framing; the same white paper targeted machine learning as the primary workload years before the GPU crunch.


Goal: cloud parity with Amazon

Decentralized solutions for cloud computing and AI 📌 Greg Osuri, Akash @ DePIN Day Austin (Fluence)

“As we keep delivering, the goal is to get to cloud parity. Cloud parity is when we can sufficiently say that you can actually move all your workloads from your Amazon of the world to Akash. We’re not there yet… our biggest capabilities, be it reservations or managed backends, I think will put us to the parity.” — Greg Osuri, 00:10:47

Context: Forward-looking roadmap statement; also notes Console 2.0 with a “DigitalOcean-like experience” and one-click, crypto-free deploys was due “in the next 3 months” (from Sept 2024).


Next 12 months: managed-services marketplace makes Akash a "network of networks" and pays open-source creators

Akash Network - The Decentralized Compute Marketplace (The Crypto Conversation (Brave New Coin))

“Akash is introducing [a] Managed Services Marketplace. Now it transitions from a service that offers compute… to a network of networks… the creators of Postgres, or creators of MySQL… offering their services… and get a cut for their creation. So we’re empowering open source creators with a business model because open source cloud needs open source software to function.” — Greg Osuri, 00:29:26

Context: The mid-term path to AWS parity (Amazon’s ~500 managed services); quote spans into the 00:30:06 block, with the 12-month shipping timeline stated at 00:30:06.


The biggest decentralization risk is who controls the roadmap

Decentralized solutions for cloud computing and AI 📌 Greg Osuri, Akash @ DePIN Day Austin (Fluence)

“Decentralization — I think if you remove enough layers, you’ll see the biggest risk is not the actual software but actually people building the software and the roadmap. So who sets the roadmap and who delivers the roadmap, all that is done in the open. That’s a big part of Akash’s decentralization approach.” — Greg Osuri, 00:06:31

Context: Explaining Akash’s ~500 contributors, on-chain funding proposals, working groups, and seven contributing companies.


Three companies controlling 80% of the cloud is feudalism

Akash Network - The Decentralized Compute Marketplace (The Crypto Conversation (Brave New Coin))

“It turns out there are three companies that control these computers, 80% of it, at least. Now, that, to me, is a little troublesome. Why? Because so much power in the hands of a few. You know, we had something like that. It was called feudalism, right? If you have so much power in the hands of a few, that few dictate the future of this public utility.” — Greg Osuri, 00:03:54

Context: Opening framing of why decentralized cloud matters; “cloud is just someone else’s computer.”


Compute should be a public utility like air and water

Compute: Past, Present, and Future by Akash Network founder Greg Osuri at deAI Summit (Pundi AI)

“Imagine a world where computation is free and accessible as the air we breathe. We believe compute should be a public utility — like air, like water, like all the resources that we don’t need to seek someone’s permission to use.” — Greg Osuri, 00:11:05

Context: Decentralized compute framed as a return to the internet’s founding principles.


The cloud tax: half of every dollar goes to cloud providers

Compute: Past, Present, and Future by Akash Network founder Greg Osuri at deAI Summit (Pundi AI)

“Today if you use an online service, 50% of every penny you pay goes to one [of] the cloud providers… All of us, no matter what sector we are in, half of every dollar we spend goes to an invisible layer — or the cloud tax, we call it.” — Greg Osuri, 00:06:44

Context: Citing an a16z study of public-company S-1 filings (Asana 63%, Palantir ~40%).


Compute should be a public utility like air and water

Compute: Past, Present, and Future by Akash Network founder Greg Osuri at deAI Summit 2024 (Pundi X Labs)

“Imagine a world where computation is free and accessible as the air we breathe. We believe compute should be a public utility — like air, like water, like all the resources that we don’t need to seek someone’s permission to use.” — Greg Osuri, 00:11:05

Context: Transition into the DePIN landscape slide; decentralized compute framed as “a return [to] the internet’s founding principles.”


The cloud tax: half of every dollar goes to cloud providers

Compute: Past, Present, and Future by Akash Network founder Greg Osuri at deAI Summit 2024 (Pundi X Labs)

“Today if you use an online service, 50% of every penny you pay goes to one [of] the cloud providers… Look at companies like Asana, where 63% of the fees actually goes to one of the cloud providers; companies like Palantir… about close to 40% goes to a cloud provider. All of us, no matter what sector we are in, half of every dollar we spend goes to an invisible layer — or the cloud tax, we call it.” — Greg Osuri, 00:06:44

Context: Citing an a16z (captioned “hor[owitz]”) study of public-company S-1 filings; a recurring quantified claim in his talks.


Akash becomes a network of networks

AWS at a Fraction of the Price – Greg Osuri | Akash Network (We are DePIN)

“Akash transitions from this simple cloud to this network of networks, which I think is very, very powerful, and I think that’s how we get to mass adoption.” — Greg Osuri, 00:41:50

Context: His stated future for Akash: a multi-service marketplace where model/software creators (e.g., Nous hosting Hermes 3, Postgres creators hosting Postgres) offer services with royalties, matching AWS’s 500 services without building them all.


Cloud is a public utility that must be decentralized

AWS at a Fraction of the Price – Greg Osuri | Akash Network (We are DePIN)

“Cloud is ultimately a public utility that should not be closed off and they should not be this opaque, and that’s the reason why you want to decentralize a cloud.” — Greg Osuri, 00:11:32

Context: Capping his critique of centralized providers, referencing the CrowdStrike outage that “got the planet to a standstill” as proof of centralized fragility.


Everyone pays a hidden cloud tax

AWS at a Fraction of the Price – Greg Osuri | Akash Network (We are DePIN)

“We all use a cloud and we all pay a cloud tax without realizing — somewhere north of 50% of every penny we pay to an online service goes to a cloud provider.” — Greg Osuri, 00:10:45

Context: Arguing the AWS/Google/Microsoft oligopoly stifles competition; the quantified “north of 50%” cloud-tax figure is one of his recurring claims.


Akash as the world's first decentralized supercloud; Nvidia its biggest user

IOSG OFR 13th Panel | GPU Symphony: Decentralized Compute Power (IOSG Ventures)

“Akash is the world’s first decentralized supercloud, and the core essence of a supercloud is where you can decouple the control plane from resource plane, where the resource plane can consist of any cloud-capable computer, orchestrated in a decentralized and open manner.” — Greg Osuri, 00:02:08

Context: Self-introduction; he adds Akash serves users “the biggest user being Nvidia, integrated directly into Nvidia products” and decentralized AI protocols like Venice AI ([00:02:52]).


Compute should be as free as air and water

Greg Osuri - Democratizing Access to AI Resources - TOKEN2049 Singapore 2024 (TOKEN2049)

“We believe compute should be a public utility. Compute should be as free as the air we breathe and water we drink, and that’s the future we’re all building towards… the access should be ungated, [it] should be open, [it] should be public.” — Greg Osuri, 00:08:48

Context: The affirmative vision after the feudalism warning; ties to restoring sovereignty over data and compute.


The cloud tax: 50% of every online-service dollar

Greg Osuri - Democratizing Access to AI Resources - TOKEN2049 Singapore 2024 (TOKEN2049)

“The costs are getting out of control. Over 50% of every single penny we spend on online services goes to one of the three cloud providers. All of us pay a cloud tax without knowing who we [are] paying the tax to… even Department of Defense, which uses Palantir, clearly pays a cloud tax… Andreessen Horowitz did this study, it’s called the Trillion Dollar Paradox.” — Greg Osuri, 00:05:12

Context: Recurring quantified claim, here with the explicit a16z study name.


Akash 50–80% cheaper than Amazon, with H100 access

Interview with Akash Network founder/CEO Greg Osuri at deAI Summit TOKEN2049 (Pundi X Labs)

“Akash is the world’s first decentralized supercloud where it connects underutilized resources with people that need these resources… the obvious advantage is cost — anywhere from 80% to 50% cheaper than what you normally pay on Amazon — and more importantly access, like GPUs which are very very hard to get on your regular cloud. Akash offers advanced GPUs like H100s for like $150.” — Greg Osuri, 00:00:00

Context: Opening self-introduction. Auto-caption garbles the GPU pricing figure (“40 uh h100s for like $150” — plausibly “$1.50” per hour).


Akash is the world's first decentralized supercloud, 80% cheaper, used by Nvidia

Key Challenges in Building Decentralized AI Infrastructure | AI / ALL Summit (Sahara AI)

“Akash is the world’s first decentralized supercloud. Supercloud, like the name suggests, is a cloud that federates any cloud-capable compute bit, a private or public cluster… where users can effectively access compute using a marketplace mechanism at significantly lower cost, usually 80% cheaper than what you would normally otherwise pay on traditional cloud… one of the biggest users for Akash is Nvidia, and Akash is directly integrated into Nvidia products.” — Greg Osuri, 00:01:26

Context: Greg’s self-introduction; “supercloud” is his recurring framing for federating the world’s idle compute. Quote begins at the end of the prior caption block.


Cloud should be as free as air and water

Sovereign AI's Battlefield: Compute, Storage, & Running On The Edge | Crypto x AI Event (Delphi Digital)

“Cloud, which is perhaps the most important fabric that holds a modern society together, and a very important public utility, should be as free as the air we breathe and the water we drink, and the access to which should not be gated. That was the whole premise of starting Akash.” — Greg Osuri, 00:38:17

Context: His founding vision statement when asked why he focused on this problem; frames cloud as a public utility.


First milestone: 1% of the cloud market

Sovereign AI's Battlefield: Compute, Storage, & Running On The Edge | Crypto x AI Event (Delphi Digital)

“For us success we measure in market share, right? So our first milestone would be 1% market share of the cloud business.” — Greg Osuri, 00:39:01

Context: Quantified success metric for Akash, justified by compute becoming a necessity people will own.


Akash: four years with zero downtime vs Amazon's ~200 outages a year

The Supercloud - by Greg Osuri, Akash (Cosmoverse)

“Akash has been around for about four years and has not gone down a single time, and you can’t say quite the same for networks like Amazon, where Amazon has reportedly about 200 outages on average per year.” — Greg Osuri, 00:10:59

Context: Reliability argument for peer-to-peer architectures with no single point of failure; also claims up to 85% cost savings and GPUs at 1/4 hyperscaler cost.


Distributed data centers need a decentralized connective layer

AI Made in Cosmos - with Greg Osuri, Murthy Vitwit, Valery Litvin & Dean Tribble (Cosmoverse)

“Now there’s enormous amounts of compute spread across hundreds and hundreds of data centers across the world… that really brings about the need for a layer to connect these distributed data centers, and that layer cannot be centralized. I think that’s critical for this layer to be decentralized, and Akash is one of the decentralized networks, if not the leading decentralized network, that’s connecting these distributed data centers.” — Greg Osuri, 00:13:05

Context: Since the GPU crunch, Nvidia allocates smaller quantities (<5,000 units) to many players rather than lots to one, dispersing global compute.


The internet's decentralized origins ended with four companies in control

The Supercloud - by Greg Osuri, Akash (Cosmoverse)

“Picture this: in 1969 the first and most widely adopted decentralized network was created to withstand a nuclear attack, and today we end up with four companies controlling the internet… Have you really forgotten the very principles that gave birth to this digital age?” — Greg Osuri, 00:00:45

Context: Opening framing; the talk is a back-to-first-principles argument for decentralization rather than an Akash pitch.


The invisible cloud tax on every online service

The Supercloud - by Greg Osuri, Akash (Cosmoverse)

“Andreessen Horowitz did this study: half of every penny you spend on online services goes to a cloud provider — in some cases like Asana, where over 63% of all the fees you pay to the cloud service goes to a cloud provider. In a way, we’re paying this invisible tax, the society, all of us together — we don’t even realize we’re paying this invisible tax.” — Greg Osuri, 00:05:13

Context: Recurring quantified claim; followed by the CrowdStrike outage as evidence of centralized fragility.


Akash as the world's first decentralized supercloud built from underutilized compute

Pioneering DePIN Growth (Cosmos)

“I’m the co-founder for Akash Network. Akash is the world’s first decentralized supercloud, and it leverages underutilized compute to create a super cluster and offers compute at significantly lower cost in a permissionless and decentralized manner.” — Greg Osuri, 00:00:00

Context: Greg’s self-introduction as panel moderator; the only substantive Akash claim he makes in this session — the remaining content is from the Sentinel and Pocket Network panelists.


Compute should be as free as air and water

Akash Network: A New Era of Affordable, Decentralized Cloud Computing with Greg Osuri | Varuni (Thecoinrepublic)

“In a world where we have heavy censorship, having a layer that is considered to be the public utility — right, compute — should be as free as the air we breathe and water we drink, and that is a foundation that humanity can accelerate on, and that foundation needs to be permissionless.” — Greg Osuri, 00:08:03

Context: The third and “most important” benefit of Akash after GPU access and 80%-cheaper cost: permissionless compute as public utility.


Five-year vision: cloud parity and a services network with thousands of regions

Akash Network: A New Era of Affordable, Decentralized Cloud Computing with Greg Osuri | Varuni (Thecoinrepublic)

“The goal is to get parity with the current cloud… Akash will transition from a resource network to a services network where you have hundreds if not thousands of services all offered in a decentralized manner… The question I’ll leave you folks with is: what happens when you have thousands of regions that are spread across the world with thousands of services? Amazon has 32 regions right now — imagine that as a thousand regions.” — Greg Osuri, 00:19:39

Context: Answer to “how do the next 5 years look?”; he also cites 50% quarter-over-quarter growth in on-chain leases and a goal to 10x growth.


Society pays an invisible "cloud tax" of 50%

Greg Osuri | Trump's impact on crypto x AI, why DePIN is inevitable, and Akash Network revenue ATH's (Proof of Coverage Media)

“We all, all of us as a society, pay something like cloud tax without even realizing we’re paying it… 50% of every single penny you pay for online services goes to [a] cloud provider that is not elected, that’s not selected, that is there because the users really don’t have too much options.” — Greg Osuri, 00:32:27

Context: Answering how Akash fits the AI supply chain; recurring cloud-as-invisible-monopoly argument.


Akash must outlive its company — systems that run as designed

What is Akash Network? Greg Osuri on Decentralized Cloud Computing | Greg Osuri - Founder of Akash (Genzio)

“The whole point of building open source decentralized systems is to remove [single] centers of liability. Anything could happen [to] Overclock Labs, and Akash will run as designed — that’s the future we want. So we are very reluctant to collect user information.” — Greg Osuri, 00:18:06

Context: Explaining why upcoming credit-card payments are designed without logins or stored user data.


Beat Amazon at a game they cannot play — inside the home

Greg Osuri - Akash: Decentralized Compute Marketplace - ep 158 (Zima Red)

“The only way to compete is to do something they cannot do. That’s why decentralization comes into play. They cannot go inside a home, where networks like Akash can go inside a home… You got to beat Amazon at a game that they cannot play — open source is something that Amazon can’t play.” — Greg Osuri, 00:13:46

Context: Answering why startups can’t just raise $200-300M to build data centers (“Amazon will put $2 billion on top”); home compute and open source as the structurally undefendable flanks.


Akash: four years, zero unplanned downtime; AWS down ~200 times a year

Akash's Greg Osuri on the Future of Cloud Computing | Mainnet 2024 (Messari)

“Akash has been alive for about four years; it has not gone down a single time. Whereas if you look at networks like Amazon Web Services, it goes down about 200 times a year… The reason why Akash is incredibly reliable is because there’s not a single company or a single cluster that controls the network, but rather a distributed network of nodes that interact with each other in a peer-to-peer manner without a central point of failure.” — Greg Osuri, 00:13:50

Context: Reliability argument; he pairs it with the CrowdStrike incident and an anecdote of an Akash auth exploit fixed by an open-source user before anyone was impacted (“fault tolerance, not fault avoidance”).


Compute should be a public utility, "as free as the air we breathe"

Akash's Greg Osuri on the Future of Cloud Computing | Mainnet 2024 (Messari)

“Compute should be a public utility. Compute should be as free as the air we breathe and water we drink, so that innovators everywhere, from every corner of the globe, have the tools to build and create and transform without barrier… It’s very important that we return to sovereignty — control over what gets built, control over what goes and touches our daily lives.” — Greg Osuri, 00:10:14

Context: His positive thesis after the feudalism warning; decentralized computing as a return to the internet’s 1969 founding principle.


The risk is a return to feudalism — including secret AGI captured by the few

Akash's Greg Osuri on the Future of Cloud Computing | Mainnet 2024 (Messari)

“The big risk as a society as a whole we have is the return of feudalism. Feudalistic society is where few have power over the many… the few today are the ones that are running AI… There’s this discussion about artificial general intelligence: what happens when somebody cracks it and no one knows that they cracked it? Because there’s no incentive for them to share this AGI — they can very well go and play in the stock market or do things that are significantly advantageous to the few. So if we don’t do something about this layer, we go back to feudalism… I call the term digital feudalism.” — Greg Osuri, 00:08:48

Context: The keynote’s central warning, framing hyperscaler concentration as a civilizational risk; “a future where corporations, or the digital lords, control humanity as a whole.”


Compute should be as free as the air we breathe

Akash Network Explained: Everything You Need To Know Before Investing by Founder Greg Osuri (Founder School)

“We believe compute should be a public utility. Compute should be as free as the air we breathe and water we drink, and that’s the future we’re all building towards. What that means is restoration [of] sovereignty: we need to have control over our data, we need to have control of our compute, and we need to empower communities rather than corporations to drive this technological progress.” — Greg Osuri, 00:08:48

Context: His recurring public-utility thesis, presented as the industry’s “new era of decentralization.”


Four years live with zero downtime — open source reliability

Akash Network Explained: Everything You Need To Know Before Investing by Founder Greg Osuri (Founder School)

“In a case like Akash, we have been live for about 4 years and there hasn’t been a single time the network went down or slowed down. It’s always available, it’s always live, and that’s possible because it’s open source — people are looking at the source code all the time, people are pushing updates all the time — and it’s decentralized: there’s not a single system that controls the entire network.” — Greg Osuri, 00:10:14

Context: Contrasted with the CrowdStrike incident — “somebody pushed bad code to production” taking down airlines — as proof of centralized opacity and fragility.


The internet was born decentralized — the cloud reversed it

Akash Network Explained: Everything You Need To Know Before Investing by Founder Greg Osuri (Founder School)

“[Personal computing] created the internet on the foundation of decentralization, connecting these individual devices in a hive network, sharing intelligence and removing the need for a centralized coordinated system to communicate… historically, decentralization is what really spun the innovation and the accessibility that we have today. But eventually somehow the centralization came, with large hyperscalers… close to 70% of all the IT spend is spent by four companies.” — Greg Osuri, 00:02:15

Context: His mainframe → PC → internet → cloud re-centralization narrative; the 70% figure is at [00:02:59].


Decentralization means removing systemic failure risk

Akash Network (AKT) Explained: Everything You Need To Know Before Investing by Founder Greg Osuri (Founder School)

“It’s really about removing the risk of a system failure — removing systemic failure risk. Today you have centralized systems like Amazon and Google of the world: if the company collapses, the product exits, no matter how important it is to you or to the society. If the company dies, the product dies.” — Greg Osuri, 00:12:28

Context: His onion model of decentralization: consensus, governance-attack economics, company influence (citing Terra), and — the deepest layer — whether development survives the founding lab, which motivated the radically-open reorganization.


20 hyperscalers is the mainframe problem all over again

Scaling Web3 and AI through DePIN with Greg Osuri! (Secret Network)

“Internet was created in the ’60s to combat the centralization of data centers… because we had large mainframes and people were like, this is not good, because one mainframe gets attacked, we’re done. The same thing is happening — we have 20 hyperscalers. It’s not good, because they get attacked, you’re done. So the best thing you can do is, number one, distribute AI as much as possible. AI should be with everyone, not with anyone.” — Greg Osuri, 00:40:36

Context: Answering “best case scenario in five years”; concentration of AI in ~20 known US hyperscaler regions is a geopolitical/military vulnerability.


Optimistic future: 10,000 anonymous data centers, not 20

Scaling Web3 and AI through DePIN with Greg Osuri! (Secret Network)

“Instead of having 20 data centers, how about 10,000 data centers, where it’s impossible for adversary to compromise our AI infrastructure in the future… one data center goes down, who cares, we have 10,000 of them, even more.” — Greg Osuri, 00:43:26

Context: He wants these to be unidentifiable, trust-minimized contributors (via TEEs) — “the more anonymous, the more distributed the infrastructure is, the more likely we’re going to survive as humans.”


Enterprises will eventually be forced to go distributed

Akash Accelerate 2025: Official Livestream (Akash Network)

“A lot of places don’t have energy now… right now it’s about 10 years you’ve got to wait before you can get an interconnection request approved… enterprises are in a very big trickle where you can’t really build out massive data centers… and you have to eventually go distributed.” — Greg Osuri, 04:49:31

Context: Keynote Q&A on enterprise adoption; he also says Nvidia promised Akash GB200 NVL72 machines (“we haven’t gotten them yet”) and that Star Cluster Alpha targets ~25 GB200 clusters with ~$220M purchasing power.


The cloud of the future is a planet-wide mesh

Akash Accelerate 2025: Official Livestream (Akash Network)

“To build one gigawatt AI cloud, we need around 8,300 of these [GB200 NVL72] racks. But here’s the key: they don’t have to live in one facility. They can live everywhere… This is hyperscale rebuilt from the ground up to be permissionless, flexible, and resilient. The cloud of the future isn’t a place. It’s a planet-wide mesh of compute, and Star Cluster is how we build it.” — Greg Osuri, 04:34:16

Context: The “math” section of the keynote — each GB200 NVL72 rack draws ~120 kW; distribution replaces the single-campus hyperscale model.


Tomorrow Akash trains and hosts the defining models; the cloud should be owned by all of us

Beyond GPUs: How Decentralization Can Solve AI's Biggest Scaling Bottleneck | Day 2 | Crypto x AI (Blockworks)

“Today, Akash runs everything from websites to AI inference, but tomorrow it’ll help train and host the models that define the new era. Because the cloud of the future shouldn’t be owned by one company — it should be owned by all of us.” — Greg Osuri, 00:02:16

Context: Forward-looking positioning of Akash from inference to training and hosting frontier models.


Build a cloud no one can shut down

Akash Accelerate 2025 - Greg Osuri Keynote (Akash Network)

“Infrastructure is destiny, but only if it’s in the right hands — of everyone. Let’s build a cloud no one can shut down. Let’s build the intelligence layer of a free civilization.” — Greg Osuri, 00:20:25

Context: Closing call to action to builders (deploy on Akash), operators (run star nodes), and investors (reserve Starbonds).


Starcluster Phase 1: hyperscale from distributed star nodes (2025)

Akash Accelerate 2025 - Starcluster (Akash Network)

“We call it Akash Starcluster: a product roadmap that brings hyperscale performance to a globally distributed network. Phase one launches in 2025 — data center grade star nodes. These are high-density racks running in existing colo spaces, powered by renewable energy, operated by independent node keepers. They form the foundation, the raw power layer.” — Greg Osuri, 00:00:01

Context: Opening of the Starcluster announcement; VMs (Aug 2025) and confidential computing (Sep 2025) are named as the supporting releases for node keepers.


The cloud of the future is a planet-wide mesh, not a place

Akash Accelerate 2025 - Greg Osuri Keynote (Akash Network)

“They don’t have to live in one facility. They can live everywhere — distributed across regions, climates, and use cases. This is hyperscale rebuilt from the ground up to be permissionless, flexible, and resilient. The cloud of the future isn’t a place. It’s a planet-wide mesh of compute, and Star Cluster is how we build it.” — Greg Osuri, 00:10:11

Context: After the math that a 1 GW AI cloud needs ~8,300 GB200 NVL72 racks (~120 kW each) which need not be co-located.


The cloud of the future is a planet-wide mesh

Akash Accelerate 2025 - Starcluster (Akash Network)

“This is a hyperscaler rebuilt from the ground up to be permissionless, flexible, and resilient. The cloud of the future isn’t a place — it’s a planetwide mesh of compute, and Starcluster is how we build it.” — Greg Osuri, 00:02:09

Context: Closing thesis of the clip, restating Greg’s recurring cloud-decentralization argument in its 2025 form.


The future isn't hyperscale — it's hyperdistributed

Superintelligence Needs The Supercloud Why the AI Revolution... | PMLS 2025 | Day 3 | Open Source AI (Blockworks)

“This is the architecture of the future: a planetary AI mesh, alive with local energy and local compute. Redundant by design because there’s no single points of failure. Open to anyone, anywhere… The future isn’t hyperscale, it’s hyperdistributed.” — Greg Osuri, 00:08:12

Context: The keynote’s thesis line, tying decentralized training breakthroughs to “intelligence more resilient, more inclusive, and more aligned with democratic values.”


Massive centralized data centers are obsolete

Decentralized Infrastructure Allows America to Compete on AI—Greg Osuri (Crypto World Daily)

“In this context, massive centralized data centers are obsolete. They’re rigid, expensive, and confined to one geographic location. Even worse, they create single points of failure. If one power grid goes down or is overheated, an entire segment of the country is plunged into a technological dark age.” — Greg Osuri, 00:02:11

Context: Core contrarian thesis of the op-ed, set against the hyperscale build-out narrative.


500 US data centers are sitting ducks in a conflict

Solving AI's Energy Crisis with Decentralized Compute, w/ Akash CEO Greg Osuri (The People's AI: The Decentralized AI Podcast)

“In case of a national conflict tomorrow, international conflict, what do you think is going to get targeted? These little sitting ducks, large targets… These are massive data centers. We all know where they are. There are about 500 of them in America… When someone takes out your infrastructure, they’re going to target this infrastructure.” — Greg Osuri, 00:38:23

Context: Follows “the more you study about the grid, the more prepper you become… the more you realize how fragile we are” [00:36:50], citing the Texas freeze; decentralization removes fragility the way Bitcoin’s design does [00:39:12].


The 2017 whitepaper called it: cloud hyper-centralization is failing

The Truth About Decentralized AI and the Future of Compute (TEACHMEDEFI)

“AWS went down four times, took almost all the internet with it. It causes such a disruption every time a data center goes down that it’s almost like decentralization sells for itself.” — Greg Osuri, 00:05:50

Context: He precedes this by quoting his own 2017 paper: “the cloud is getting hyper inefficient, hyper centralized — we’re going to start seeing problems with stability, with prices and accessibility. And look what happened now.”


Akash built the first-ever supercloud

The Remote Work Tribe Podcast: Greg Osuri (The Remote Work Tribe Podcast)

“Cornell really had a vision of creating a cloud where you can take this compute and actually deliver a similar product, if not better than what we have with Amazon. And we really took that idea and we implemented the first ever, or we built the first ever supercloud. I mean, no one actually built supercloud before us. And we took it one step further. We made it open source. We made it decentralized.” — Greg Osuri, 00:03:10

Context: Credits Cornell University with coining “supercloud” [00:02:23]; the premise is vast underutilized compute — data centers used ~15% of the time and powerful home gaming GPUs. Decentralization is framed as embedding core principles “into a system that cannot be changed, no matter how big it grows” [00:03:57].


"We predicted that" — the 2017 white paper called the cloud's cost and reliability crisis

The Remote Work Tribe Podcast: Greg Osuri (The Remote Work Tribe Podcast)

“We noticed that cloud tooling has been not only getting extremely complicated, it’s getting extremely pricey… we prophesied saying that, look, if the cloud continues the way it is going, it’s going to lead to the massive reliability issues, massive cost issues, like we see today. I mean, people complain about AWS going down every other month… And we predicted that. We wrote a white paper. We published that in 2017.” — Greg Osuri, 00:01:40

Context: Recounting Overclock Labs’ founding thesis; he claims AWS “went down four times this month” (January 2026) as vindication.


Cloud should be priced like a utility — by a free market, not the state

🔴LIVE: WLFI In Trouble? Winter Is Over. Brian Armstrong Pushing DC HARD, FTDA then Lighter & Akash (The Rollup)

“We have fair pricing and transparent pricing for electricity and water and all the commodities that we use and depend on — why not have the same mechanism for the cloud? And we don’t want a state actor to enforce that, but rather a free market to enforce it. That’s the thesis of decentralized compute.” — Greg Osuri, 02:21:38

Context: His recurring “cloud as commodity/utility” argument, here framed as the middle layer of the decentralized-AI stack.


Clouds are picking AI's winners and losers

🔴LIVE: WLFI In Trouble? Winter Is Over. Brian Armstrong Pushing DC HARD, FTDA then Lighter & Akash (The Rollup)

“For the first time, we saw cloud no longer works in your best interest, but rather gives compute to people they think are going to be successful… one might say — this is speculation — Amazon overwhelmingly supporting Anthropic might have resulted in their success. So who’s to say cloud companies are not picking winners and losers?” — Greg Osuri, 02:20:14

Context: His case for decentralizing the compute layer, recalling the 2023–24 GPU squeeze when startups couldn’t get quota.


Cloud demand will outpace compute supply

From AWS to Akash: Greg Osuri on Building a Decentralized Compute Marketplace (Smart Economy Network)

“There will come a point where cloud consumption outpaces the supply for the compute… at the rate at which the data was growing, there is no single company in the world that can catch up and scale. So the obvious answer is a grid model… in a peer-to-peer manner where there is compute everywhere.” — Greg Osuri, 00:26:13

Context: His founding conviction for Akash; he says the thesis was first proven in 2024 when clouds literally ran out of GPUs while people had GPUs in their data centers and homes.


Four-year outlook: decentralization takes center stage

From AWS to Akash: Greg Osuri on Building a Decentralized Compute Marketplace (Smart Economy Network)

“If you ask four years from now: world will be energy rich, dollar strong compared to other fiats… GDP will be much higher growth, employment will be significantly lower… and decentralization will take a more center stage, because moratoriums on data centers and all these will start kicking in… the only way to scale will be decentralized systems.” — Greg Osuri, 01:04:31

Context: Closing synthesis; he also predicts an energy innovation wave (turbines, desalination) enabling cheap energy to “solve hunger” and water scarcity.


Half of every SaaS dollar goes to cloud rent-seekers

The Infrastructure Behind Agentic Finance, with Shashank Yadav and Greg Osuri of Akash Network (Fraction AI)

“You would see a lot of these financials of a lot of SaaS companies — on an average, about 50%, every half of every dollar you pay was going to a cloud provider for the services. And in some cases like Asana… about 60 to 70%… I think like Netflix was 70%… and the margins were 60, 70% for these cloud providers… The access to the cloud should be open, and the software that governs cloud should be open. It should be built by the people that use it.” — Greg Osuri, 00:05:46

Context: Origin story — “the idea that total cost of ownership is going to be lower on the cloud was just a fallacy”; Amazon as “a giant rent seeker” killing open-source business models.


Every company will have their own AI cloud

This Crypto Turns GPUs Into Income For Everyone!!!! (AllinCrypto)

“Every company will have their own AI cloud. And that’s a prediction I made a long time ago, and that prediction still stands true today… The obvious answer is to own your compute.” — Greg Osuri, 00:13:40

Context: Citing xAI selling excess compute, Meta mulling cloud services, and ~100-120 “neoclouds” in Akash’s database; owned compute sits idle, so a marketplace for latent supply becomes essential.


Hyperscalers will join Akash

This Crypto Turns GPUs Into Income For Everyone!!!! (AllinCrypto)

“Akash is a supercloud. Supercloud is a cloud of clouds… I would imagine Google and Amazon and Azure of the world will become part of Akash Network at some point.” — Greg Osuri, 00:12:55

Context: Answering whether Akash competes with hyperscalers; he reframes it as a distribution channel, like hotels listing on Airbnb.


Akash: a supercloud built from places clouds don't go

AI Data Centers Are Eating the Grid. Is There Another Way? (The People's AI: The Decentralized AI Podcast)

“Akash is a decentralized super cloud. What that means is it is a cloud network where the compute comes from places that normally cloud providers don’t offer. Places like homes, buildings, small buildings. Pretty much anywhere there’s compute can be contributed to Akash network for the greater good.” — Greg Osuri, 00:37:18

Context: Host reintroduces him as the person who made “the decentralized bet long before it was fashionable”; Greg says constrained compute is driving more usage to Akash.


Core thesis: spread compute; Akash hit 10 billion tokens a day

AI Data Centers Are Eating the Grid. Is There Another Way? (The People's AI: The Decentralized AI Podcast)

“That’s the core thesis of Akash Network is to, instead of concentrating this compute power in small places, let’s spread across as much as possible. For that, just yesterday, we reached 10 billion tokens a day. That’s quite a lot. It’s more than some of the centralized players.” — Greg Osuri, 00:40:41

Context: Milestone claim dated to the day before recording (~July 2026); “spread the prosperity and let people go figure out the best way to source compute and energy.”