AlphaFold doesn't solve folding. It makes metaheuristic guesses without writing a bunch of quantum chemistry, statistical physics, thermodyanamics, and topology maths / algorithms.
I don't mean to downplay AlphaFold, but we haven't solved protein folding yet. This press is really getting ahead of itself.
I personally know several people that do research that have unlocked new possibilities through this tool. My wife is a neuroscientist and she's used this tool a few times for reasons that are above my head (even with a Msc in Microbiology). This type of work used to take a PhD student 4 years or more to do a single relatively simple protein. Getting answers within a few seconds is revolutionary.
https://www.chemistryworld.com/opinion/why-alphafold-wont-re...
Many engineers will say it doesn't code. It just regurgitates and remixes the data it was trained on. It just makes "meta-heuristic guesses."
But anyone taking an honest and objective view of it can see that Copilot does add value. It's no substitute for a real, human, engineer, but it clearly adds value.
I don't think AlphaFold would get to this level of funding, resource commitment, etc. if it was adding 0 value.
Being a domain expert, I'm curious what value, if any, you think a large transformer model could add to the domain of protein folding. Is it really zero value, in your view?
They didn't say that it added zero value, that's entirely you. They said that it doesn't solve the problem on its' own, which is true.
> Author Pamela McCorduck writes: "It's part of the history of the field of artificial intelligence that every time somebody figured out how to make a computer do something—play good checkers, solve simple but relatively informal problems—there was a chorus of critics to say, 'that's not thinking'." Researcher Rodney Brooks complains: "Every time we figure out a piece of it, it stops being magical; we say, 'Oh, that's just a computation.'"
Solving physics isn't a soft thing like making pretty art.
I'm firmly in the "AI/ML will eat the world" camp, but the praises being foisted upon AlphaFold are borderline damaging to the real field and its practice.
You can't throw AlphaFold at pharmaceutical problems and call it a day. This press feels like a "mission accomplished" victory lap when it's very clear we're only just getting started.
You won't have a perfect answer unless you want to predict its shape in a vacuum, which wouldn't be very useful either way. Having it "close enough" is already extremely useful. There are definitely edge cases where it gets it wrong, but there are always edge cases in ML. More data = better results with the same architecture.
Tons of things that won Nobel prizes weren't 100% accurate, it's not a prize for solving science, rather a prize for advancing science.
My wife is a neurobiologist and the impact of this advance is groundbreaking for her work.
