I’ve been wanting to write about the interesting work that FutureHouse has been doing on reasoning models for biology, but I’m torn about how to frame it: On the one hand, I’m worried that the problems may not be very urgent/expensive to their prospective users.