when somebody says "of course even I can admit that ai is good for X"
X is often something *outside their area of expertise* they've heard it's good at. like, for example, protein folding.
spoiler: it turns out that actually there are serious problems with using it for protein folding
Anthony Moser
Here is a rather approachable essay on the subject just published.
Good scientists reveal how they do their experiments and report their results; so should any machine-driven research
