In a new paper published in ACS Editors' Choice, ChatGPT was trained by BIDMaP faculty and students via precise prompt engineering to efficiently text mine the academic literature on metal-organic frameworks (MOFS). The resulting "ChatGPT Chemistry Assistant" successfully produced highly accurate synthesis condition predictions for over 800 MOFs. Congratulations to Zhileng Zheng, Oufan Zhang, Christian Borgs, Jennifer T. Chayes, and Omar Yaghi for this groundbreaking collaboration!
ACS Editors' Choice Initiative is a program that selects one scientific article per day from a portfolio of over seventy-five journals and makes it openly available for a limited period of time. The purpose of the program is to highlight important, newly published work of an author and research team that is of broad public interest.
See the full article here: https://data.berkeley.edu/news/chatgpt-accelerates-chemistry-discovery-climate-response-study-shows
Image reference:
"ChatGPT + Chemist". Adapted from Zheng et al. (2023), Journal of the American Chemical Society. DOI: 10.1021/jacs.3c05819