The models are already good enough.

In the 90s I worked in a genetics lab cloning DNA into bacteria. Success rate was often one-in-a-million. Didn't matter. We'd plate millions of bacteria on antibiotic dishes and wait. Only the ones th…