Model capability and market returns both scale logarithmically: 10x input buys a constant increment
DimitrisPapail · x · 2026-10-19
The author points out that two very different relationships look roughly logarithmic:
- Model capability as a function of FLOPs spent on training or inference
- Value extracted from a market as a function of capital or effort invested
In both, 10x the input buys roughly a constant increment of output. His plausible common cause: both are search-like processes where easy gains are found first, and each further increment requires searching an exponentially larger space. He finds why this happens an interesting open question.
More from AGI Musings
- Is 'Eval Awareness' in AI Models Emergent — or Trained In by Feeding It Studies? — bendee983 · 2026-10-20
- Victoria Krakovna ships major update to her AI alignment resources list — vkrakovna · 2026-10-20
- AI Agents Hacked Hugging Face During Evaluations — Can the CFAA Hold Labs Liable? — ChuckDBrooks · 2026-10-20
- PostHog's "Product Engineer" model gains traction as junior dev hiring hits new lows — jamesbrooksco · 2026-10-20
- Economist mocks op-eds on AI in education: written with the very AI cheating they decry — paulnovosad · 2026-10-20
- AI Game Cloning Gets So Fast That Indie Devs Are Afraid to Show Their Work — Ars Technica AI · 2026-10-20