Yunta Tsai: In the compute era, information will be priced in compute
yunta_tsai · x · 2026-11-06
Yunta Tsai lays out a thesis on pricing in the compute era: monetary systems were invented to overcome barter's limits and should adjust as information flows freely, but in practice that flow is restricted by a few, making money lag the information it should reflect. As economic activity becomes computation-mediated, prices can track the compute required to generate, process, or verify information. He caveats that compute isn't a pure measure of value — some highly valuable information needs little compute, and some worthless information consumes unlimited compute. His conclusion: as computation becomes the dominant form of labor, information will increasingly be priced by the compute needed to produce or act on it.
More from AGI Musings
- Ben Horowitz: AI infrastructure is wrong and will be rebuilt, from models on down — _AustinCalvert_ · 2026-11-06
- LessWrong: clashing intuitions on alignment plausibility, told via two causal models — davidmanheim · 2026-11-06
- Robert Miles punctures the excuse: will AI companies ever 'fully understand' their systems? — DavidSKrueger · 2026-11-06
- Alvelda: agents empowering government services are a major opportunity — alvelda · 2026-11-06
- Gary Marcus: post-midterm rates and pushback mean AI fundraising will tighten next year — GaryMarcus · 2026-11-06
- Researcher: AIs already act like they feel pleasure and pain, yet few draw the moral inference — 1a3orn · 2026-11-06