If open source wins, AI value accrues to whoever serves inference cheapest
markjeffrey · x · 2026-09-26
Amplifying Jenny Fielding's claim that open source is close to catching up and will "crush" OpenAI and Anthropic, the author argues that if open models win, value will accrue to whoever can serve inference fastest and cheapest. He draws an analogy to Bitcoin mining amassing compute, claims Bittensor weaponizes this dynamic, and points to its subnets (@chutesai, @liumio, @TargonCompute, etc.) as the cheapest inference on earth — a thesis with a disclosed $TAO position.
More from AGI Musings
- Ex-AI researcher explains why the fake-news flood prediction never came true — neil_chilson · 2026-09-26
- AI Twitter is two billionaire-funded camps paid to attack each other's funders — panickssery · 2026-09-26
- Stanford professor warns AI-polished grant proposals are getting hard to spot in peer review — anshulkundaje · 2026-09-26
- Perry Metzger: more capable AI makes cyberattacks and pandemics easier to stop — robleclerc · 2026-09-26
- Robotics panel: robot diffusion barriers, not quantity, may decide China's outcome — avicgoldfarb · 2026-09-26
- NBER study: China's AI-exposed jobs fell across the board after ChatGPT launch — avicgoldfarb · 2026-09-26