Post

finally got a link to share with people who try and say "oh you DON'T LIKE AI??? what about ALPHAFOLD!!!!" as if it has anything to do with any of this AI capex

Nah Alphafold isn't really very useful. It is sometimes able to replicate a crystal structure if you handhold it enough. But it's not a physics model so it's never been able to predict the conformational dynamics that are important to understanding realworld solution phase structure

It's never been useful for predicting a drug structure because proteins aren't statues in a vacuum! They're machines in constant motion carrying out widely variable tasks depending on their environment. You need a physics model, not a language model. And no one is anywhere near producing one.

Sure, but I don't think ploughing the resources into Gemini is going to get to that either