Post

Not super surprised by this. Alphafold is at a deadend. 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

I like to put it this way... There are plenty of successful A.I. disciplines, that have been providing value to the sciences and society as a whole, for decades.

Curious, what is your take on Apple's exposure to the systemic risks to other players which your work highlights?

macrumors.com/2026/07/27/e...

Google once again demonstrating its unparalleled ability to pioneer groundbreaking, useful technology only to set fire to it in an astounding act of shareholder-appeasing self-harm.

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.