AI is everywhere these days, and though there’s debate about how useful it is, one area where experts think it could be game-changing is scientific research. It promised to be particularly useful for speeding up drug discovery, an expensive and time-consuming process that can take decades. But so far, it hasn’t panned out.

The few AI-designed drugs that have made it to clinical trials haven’t been approved, venture capital investment in these efforts has cratered in the last few years, and many startups have shut their doors. So why has it been so hard to make AI-designed drugs? What are the fundamental issues, and what does the future of this research look like?

Joining Host Ira Flatow with some answers is Peter Coveney, who studies how chemistry discoveries can be sped up with algorithms and computers.

Guest: Dr. Peter Coveney is a professor and director of the Centre for Computational Science at University College London.

Transcripts for each episode are available within 1-3 days at sciencefriday.com.

Subscribe to this podcast. Plus, to stay updated on all things science, sign up for Science Friday's newsletters.

In new Brexit row, Britain details post-EU plans.