Compute-for-token model guarantees future inference revenue
Solving the AI energy crisis | Greg Osuri on what it takes to power AI (Changelog)
“The business model for distributed training — almost all of the models I’ve looked at — the common thing they have is you contribute compute, you get some representation of that contribution in form of a token, and that token will guarantee you future revenues from the model, from inference.” — 01:06:10
Context: Explaining why decentralized training needs a token; he notes a 5090 on Akash earns ~92 cents/hour today and predicts intermediaries will take token risk before consumers do.