Mathematician lpachter tears into Terence Tao's AI slide: the cancer analogy doesn't hold up
lpachter · x · 2026-10-11
UCLA professor Lior Pachter posted a thread criticizing Terence Tao's slide deck on AI, which used a hypothetical AI-designed cancer treatment as an analogy for misalignment risk. Pachter argues the analogy fails on multiple grounds: it confuses clinical trial phases ("stage 3" vs "phase 3"), and terms like "all cancers" and "stage 1" are medically sloppy — many cancer types aren't even staged. He also notes none of the committees formed around math and AI include a mathematical biologist, calling the naivety a reflection of disciplinary myopia.
More from AGI Musings
- Chris Albon: Kevin Roose's AGI Chronicles makes unread AI books feel like ancient tomes — chrisalbon · 2026-10-11
- The 25-year-old dichotomy: frontier-lab data broking machines vs. grads who can't crack interviews — dhruv2038 · 2026-10-11
- Domingos: Neuralink's killer app is human brains providing compute cycles for AI — pmddomingos · 2026-10-11
- Collective unconscious is AI's dark matter and biology's final moat — beffjezos · 2026-10-11
- 228 tasks, 26 models: METR's time horizon curve is nearly flat from 2-30 min — lulzxdxdxd · 2026-10-11
- AI Code Was Dismissed as Slop a Year Ago; Now It May Redefine Good Software — kevinnbass · 2026-10-11