I appreciate the context, but will add that it doesn't address the replication crisis where huge amounts of papers literally cannot be replicated (likely fraud)

I appreciate the context, but will add that it doesn't address the replication crisis where huge amounts of papers literally cannot be replicated (likely fraud) and I forget the exact situation, but I think it has happened to prominent cancer researchers and so on. This goes a lot deeper than just AI BS. Someone just posted an article on HN a few months ago about junior researchers being told to move data points to make the results statistically significant.