Parallel Bio

Parallel Bio Cofounder Robert DiFazio on the AI problem in drug discovery: "What's interesting about AI is it's a very one-sided conversation. Everybody's focused on the frontier models, which are incredibly exciting and powerful, but people forget the other half, which is data. It's the same as with original computers, GIGO, garbage in, garbage out. You could have the world's most powerful computer, but if you're feeding it garbage, that's what comes out." "A potential issue with AI in a lot of areas is that these models are incredibly powerful, and pharma is racing to adopt AI, but they're still training AI on animal data. The issue isn't the models. The issue is that they're still using animals." "So for AI to deliver on its promise in biology, doing all the things people describe, which I think is very much possible, we have to have better data. That's the main goal of the company, to produce accurate and diverse human data at an industrial scale."