46 min
Seminar, Hosted by Lucy Family Institute for Data & Society, University of Notre Dame — Soc(AI)ety Seminars
AI for molecular design and synthesis, with implications for accelerated discovery & dual use
Massachusetts Institute of Technology
Recording
Abstract
The role of AI in scientific discovery keeps expanding, particularly in chemistry and biology, where machine learning has become nearly ubiquitous in early-stage small-molecule discovery. Connor Coley presents his group's methods and applications in synthesis planning, generative molecular design and structure elucidation by mass spectrometry, with an overview of what these tools can and cannot yet do and of their dual-use implications. AI does not remove the complexity of molecular discovery; it adds new tools for working through it.