™The Bitter Lesson of Biology: Why Scaling AI Isn't Enough — Pushmeet Kohli (DeepMind) & Sal Candido (Biohub)
A 32:41 podcast panel: Google DeepMind's Pushmeet Kohli and Biohub's Sal Candido discuss AI for biology after AlphaFold — scaling laws, data quality, the virtual cell, and drug discovery.
Key points Tap a time to jump
Opening: AlphaFold has not fully solved protein folding
Finding the right scaling laws and data for biology matters more than blindly scaling up
AlphaFold's data limits; building a virtual cell needs fundamentally different datasets
AlphaFold's limits: structure, function, dynamics; protein dynamics and disorder remain unsolved
Moving from models of individual proteins to whole biological systems
Protein language models may hold scientific knowledge not yet unlocked
When AI could bring 10x–100x acceleration to drug discovery and reach the clinic
Curing all disease needs 10x breakthroughs, not 10% incremental gains
Comments
More to watch
All →Latest news
All →- Oct 13Suno Speech vs ElevenLabs Eleven v4: One-Pass Soundtrack vs Expression Control
- Oct 13PageIndex vs Graphify: Structured Retrieval Tools Compared
- Oct 13Claude Haiku 5.5 vs GPT-6 Instant: Lightweight Speed Models Compared
- Oct 13Drex 1.5 vs Decision-1: The Models That Score, Not Write
- Oct 13Muse Wonder vs GPT-6: Two Flagships Meet in One Week
- Oct 13Drex 1.5 Review: A 9B Decision Model That Scores Options, Not Text
Today
Oct 13 Tuesday- Brief
- Quiz
- Call