I'm not super familiar with this, but I think many of the gene therapy advances did involve some form of machine learning to find relevant genes.

Some might also use protein folding simulations, and the best ways to do that involve narrow AI.

I'd love to hear from somebody in the medical research field how much AI/ML actually speeds up their progress.

The general acceleration in research and outcomes, we're now fixing deep issues more and more often. And hardware / software improvements keep coming too, even outside of computing. Sorry if that sounded pompous but there's a slight feeling that we're not plateauing, quite the contrary.