MIT team's Science Task Taxonomy maps 232 subfields and 208,202 scientific tasks
JMateosGarcia · x · 2026-10-06
The author highlights the methods paper behind their AI x science report as an example of creative method recombination — combining feature classification, clustering, and LLM labelling, with a network visualisation made by Gemini. A treat for data scientists.
Related event: MIT Releases Science Task Taxonomy Covering 208,000 Tasks(2 posts)→
More from AGI Musings
- Even skilled trades have their Olympics — and robots may make them our next hobby — clemnt · 2026-10-06
- Claim that AI proofs are incomprehensible to humans is questionable, argue commenters — basedjensen · 2026-10-06
- DeepMind backs 3-year Fleming Initiative programme to set standards for AMR AI evaluation — GoogleDeepMind · 2026-10-06
- Model safety works until someone puts autonomous data centers in space — VraserX · 2026-10-06
- Product Hunt CEO's 2027 calls: multiplayer agents take off, 'handmade software' emerges — alexmacgregor__ · 2026-10-06
- Anders Sandberg on Fred Polak's classic: imagined futures drive built ones — anderssandberg · 2026-10-06