New comment in Nature by me and @sayashk on the risks of the extremely rapid ongoing adoption of AI in science nature.com/articles/d41586-0… One risk is that machine-learning-based modeling is tricky to do correctly and this has led to a reproducibility crisis, which our research has exposed: reproducible.cs.princeton.ed… aisnakeoil.com/p/scientists-… The second risk is bigger but more subtle. Science is not merely a collection of facts or findings. Actual scientific progress happens through theories and paradigms. Historically, there have been many examples of fields getting stuck in a rut even as they excelled at producing individual findings, such as astronomy before the Copernican revolution. The model of the Universe with Earth at its center was extremely accurate at predicting planetary motions, because of tricks such as ‘epicycles’. Today, AI excels at producing the equivalent of epicycles. All else being equal, being able to squeeze more predictive juice out of flawed theories and inadequate paradigms will help them to stick around for longer, impeding true scientific progress. The article concludes with tentative suggestions for moving forward nature.com/articles/d41586-0…