The Future of Generative AI

BMS has gone from less repurposed legacy data and see if we can make it usable for machine learning. Sometimes you kind of have to do things that you wouldn't normally do because it's what a system or predictive model needs to see, right? There could be an example where an algorithm would say that you need to put this compound together and any scientist that knows what they're doing is going to say that that is definitely going to fail. We are currently using AI pretty heavily in our protein-degrater programwhere it's really been very helpful in helping us sort through different types of designs.