May 20, 2025
A guide to navigating AI chemistry hype
Rafael Gómez-Bombarelli talks to Chemical & Engineering News about AI tools in chemistry, pointing to reproducibility issues and the need for benchmarking and real-world validation.
Categories:
Faculty, In the Media, Research
DMSE's Rafael Gómez-Bombarelli cautions against overreliance on AI tools in chemistry, saying that papers based on ChatGPT-generated data “don’t feel like science.”
Gretchen Ertl
Rafael Gómez-Bombarelli, associate professor in MIT’s Department of Materials Science and Engineering, weighs in on the hype surrounding AI tools in chemistry. In an interview with Chemical & Engineering News, he cautions that papers based solely on data from large language models like ChatGPT “don’t feel like science,” citing concerns about reproducibility, lack of benchmarking, and the need for real-world validation.
Read the full story in Chemical & Engineering News.