10 October 2026
Panel says AlphaFold did not solve protein folding; more scale alone won't suffice
First reported
Latent Space ran this on .
- A recorded panel featured DeepMind's Pushmeet Kohli and Biohub's Sal Candido discussing AlphaFold, the protein-structure prediction system.
- The panelists argued that adding more computing power and data will not by itself solve biology.
- They described AlphaFold as a starting point rather than a finished answer to the problem.
Our read
Only Latent Space of the 30 tracked newsletters covered this panel. It follows our 2026-09-05 story on AlphaFold predicting protein structures with experimental accuracy.
Earlier on this story
Written by Newskeryx from which newsletters covered this story and from our own archive.
How it was covered
Latent Spaceswyx & Alessio
The bitter lesson for biological data, scaling laws, AlphaFold's limits, protein dynamics, cryo-EM, protein language models, interpretability, and AI-driven drug discovery