# Home compute is usable once distributed training/inference is cracked

- **Speaker:** Greg Osuri
- **Attribution:** attributed
- **Said:** 2024-09-27
- **Theme:** Local AI — https://www.gregosuri.com/predictions/local-ai/
- **Source:** IOSG OFR 13th Panel \| GPU Symphony: Decentralized Compute Power (IOSG Ventures)
- **Watch at:** 00:07:08 — https://www.youtube.com/watch?v=_UG21_OaZLo&t=428s
- **Canonical:** https://www.gregosuri.com/predictions/local-ai/2024-09-27-home-compute-is-usable-once-distributed-training-i-uoyf/
- **Cite as:** Greg Osuri, "Home compute is usable once distributed training/inference is cracked," IOSG OFR 13th Panel \| GPU Symphony: Decentralized Compute Power (IOSG Ventures), 27 Sept 2024, 00:07:08. https://www.youtube.com/watch?v=_UG21_OaZLo&t=428s

## Quote — verbatim

> Yes, we do have a lot of compute at home that's very usable, but we haven't quite cracked the code in distributed training and inference that can leverage various sets of devices.

## Context — site annotation, not spoken

Naming the remaining technical blockers — distributed training/inference ("getting solved") plus verifiability for permissionless, trustless access to home devices.