ENIGMA Q&A: Why Open Datasets Made Alzheimer's AI Classification Nearly Solved
PTenigma · x · 2026-11-23
The author shares a transcript (produced with Claude, which handled his heavy accent) of an unusually long 20-minute Q&A after an ENIGMA-related talk:
- Why Alzheimer's classification advanced so fast: The answer credits open, labeled datasets like ADNI and OASIS that let the whole community work on them, plus the disease's very large effect size; the binary classification task is close to solved (0.9 accuracy/AUC).
- Extension: Thompson agrees that with more open datasets, other problems could reach similar performance quickly, and is asked whether ENIGMA plans to release open datasets.
- Slides: The talk frames AUC(N) via a Strasbourg Γ-calculus — the whole learning curve captured by one zeta value — and derives a variance zeta law decomposing uncertainty into class-mean gap, dimension, and covariance estimation terms.
- References include Thompson, "How Much Data is Enough? The Zeta Law of Discoverability in Biomedical Data" (arXiv:2604.17581, 2026).
More from AGI Musings
- Arohan asks Anthropic to publish full human-agent trace of a Millennium Prize solve — _arohan_ · 2026-11-23
- User asks Anthropic to publish a full human-agent trace of solving a Millennium Prize problem — _arohan_ · 2026-11-23
- LeCun cites Socrates' opposition to writing to counter AI capability worries — ylecun · 2026-11-23
- Mustafa Suleyman and Yuval Harari discuss AI, consciousness and humanity — krishnan · 2026-11-23
- Polymarket: Only 10% odds US enacts AI safety bill by end of 2026 — Polymarket · 2026-11-23
- OpenAI Chief Scientist warns AI could spark 'avalanche' of breakthroughs by 2027 — Polymarket · 2026-11-23