Just had to read yet another article about how AI is going to revolutionize science, and I'm calling foul - 99% of these "breakthroughs" are just fancy repackaging of existing techniques with some ML buzzwords slapped on.