Discussions about artificial intelligence tend to focus on its risks, but there is also excitement on the horizon. AI tools, like the models beneath ChatGPT, are being increasingly used by scientists for everything from finding new drugs and materials to predicting the shapes of proteins. Self-driving lab robots could take things even further towards making new discoveries. As it gets ever more useful, could AI change the scientific process altogether? Jane Dyson, structural biologist at the Scripps Research Institute in La Jolla, California, explains why Google DeepMind’s AlphaFold tool is useful, but scientists should be aware of its limitations. This week, Google DeepMind released a new tool to unpick the link between genes and disease, as Pushmeet Kohli, head of the company’s “AI for Science” team, explains. Also, Kunal Patel, one of our producers, meets Erik Bjurström, a researcher at Chalmers University of Technology and Ross King, a professor of Machine Intelligence at Chalmers University of Technology and at the University of Cambridge. They explain why self-driving lab robots could make research more efficient. Alok Jha, The Economist’s science and technology editor hosts, with Abby Bertics, our science correspondent and Tom Standage, deputy editor.  Sign up for Economist Podcasts+ now and get 50% off your subscription with our limited time offer: economist.com/podcastsplus. You will not be charged until Economist Podcasts+ launches. If you’re already a subscriber to The Economist, you’ll have full access to all our shows as part of your subscription. For more information about Economist Podcasts+, including how to get access, please visit our FAQs page. Hosted on Acast. See acast.com/privacy for more information.