I wholeheartedly agree with this criticism of the bullshit idea of ETES (end-to-end science) done by AI-systems, especially if "breaks," as I presume, mean the
The concept of AI systems performing "end-to-end science" is criticized as a flawed idea, particularly when it involves reframing problems or theories, which current AI is incapable of. The author suggests moving beyond outdated philosophical references like Karl Popper to acknowledge more contemporary and relevant work in the philosophy of science. AI
IMPACT Critiques the notion of AI performing scientific discovery, suggesting current systems lack the capability for true problem reframing.