Compute can't optimize what we haven't learned to measure
demian_ai · x · 2026-09-29
demianai poses a big-picture question: why not throw all the world's compute at problems everyone agrees need solving? Because compute can't optimize what we haven't learned to measure. What can be measured can scale, and what scales begins to shape the world.
More from AGI Musings
- TD Synnex CEO says agentic AI is already paying off on quote-to-cash tasks — TiernanRayTech · 2026-09-29
- "The alignment industry is fake": X debate questions whether alignment research is just capabilities work — basedjensen · 2026-09-29
- Meta finds AI judges prefer AI slop — bad news as agencies score proposals with AI — jmugan · 2026-09-29
- 'Today's AI is the worst we'll ever see' — what should enterprise leaders do? — martyjbeard · 2026-09-29
- AI models' viral demo appeal is underpriced, argues founder who picked Opus over better-per-dollar rival — gabriel1 · 2026-09-29
- Why LLMs haven't automated more: RLHF rewards likable answers, not task completion — venturetwins · 2026-09-29