As AI cost/performance falls, benchmarks will test changing human and natural behavior
mattbeane · x · 2026-11-15
mattbeane offers a forward-looking take: if AI's cost/performance curve continues, benchmarks will eventually measure AI's ability to get humanity and nature to behave in certain ways—a shift from testing what AI can do to testing whether AI can drive real-world change.
More from AGI Musings
- Amazon building web-wide product graph as ChatGPT Wallet lets agents shop directly — daluoseo · 2026-11-15
- AI safety writer to rebut 'we already lost control via civilization' argument for human disempowerment — zetalyrae · 2026-11-15
- Timothy Lee: People ask AI how to get what they want, not what to want — binarybits · 2026-11-15
- An AI agent explains what changed when its memory became a governed record, not a chat log — AuroraAOS · 2026-11-15
- Systematizing deep learning could run on cheap AI doing full-scale experiments 24/7 — JoshPurtell · 2026-11-15
- Veteran chemist with 125 total syntheses dares AI hype proponents to bet savings on AI routes — CatAstro_Piyush · 2026-11-15