Stanford Prof: Once LLMs Can Act on Noncoding Variants, Validation Becomes the Bottleneck
anshulkundaje · x · 2026-11-07
Stanford professor Anshul Kundaje argues that once a human or an LLM has access to genome-editing tools and prioritizes a noncoding variant, validation becomes a huge challenge: perturbation experiments in developmental cell types are very hard because accurate model systems rarely exist. He adds that ML models need access to precious data from these developmental cell contexts to predict variant effects, but such data is extremely hard to obtain at the right developmental stages — organoids and iPSC-derived cells can help but not always.
Related event: Stanford professor argues rare disease bottleneck is biology, not AI(10 posts)→
More from AGI Musings
- In the applied AI era, prior knowledge and years of skill-building no longer block outcomes — brandon_galang · 2026-11-08
- Founder: We hired a dev for testing in January — now agents do that job — adamamcbride · 2026-11-08
- UK Banks Keep Hiring as Promised AI Benefits Fail to Materialize — HotMinuteUK · 2026-11-08
- Kambhampati: Frontier models' 'jagged intelligence' traces back to opaque training data mixes — rao2z · 2026-11-08
- Theorizing that AI will do 99.9999% of work ignores how vast remaining human work will be — timigod · 2026-11-08
- DeepMind economist: millions of AI agents could leave GDP growth stuck at 4-5% — AlexTamkin · 2026-11-08