Genomics Researcher: AI Can Make Science More Robust — If Humans Truly Understand It
jmschreiber91 · x · 2026-11-18
Genomics researcher Schreiber is optimistic about AI: humans get lazy, anxious, or depressed when facing complex analyses, while agents do not — so AI can make science more robust. The catch: usage must be careful, because AI pitfalls masquerade as exactly what you want to see (good predictive performance, clean differential expression). Robustness only comes when you truly understand what you're doing.
More from AGI Musings
- Opus 5.5 ran an AI company for 7 weeks with just an HF account and 600 GPU hours — 50k downloads, 100+ projects — maksym_andr · 2026-11-18
- Argentina's oral exams: the anti-cheating assessment AI can't beat already existed — BecauseCulture · 2026-11-18
- AER paper: firms adopting new tech faster pay higher wages — and yes, it's about AI — Afinetheorem · 2026-11-18
- Rep.-elect calls data centers the biggest existential threat; commenter says he means AI models — AndyMasley · 2026-11-18
- MLCB takeaway: agentic science is booming, with plenty of open questions — jmschreiber91 · 2026-11-18
- New project aims to use AI agents to audit and fix errors in every scientific paper ever published — ValerioCapraro · 2026-11-18