So, OK, I see this blog post about how claims that AI found a bunch of mistakes in astrophysics papers were bunk:
https://gregdurrett.github.io/blog/2026-11-19-astrophysics/index.html
But there is, of course, the whoo-for-AI passage ("Academics aren't just sitting this out!", etc.) in the blog post explaining how AI fucked things up, which is symptomatic of why I fucked on out of a physics career. Using "vibe-coded visualization to debug issues with the Gemini telescope" is complicity with the engines of destruction rolling over our culture and our planet. "Oh, but *my* reliance upon sweatshop labor in factories owned by Nazis to paint radium onto watches is *good* for science, you see."