Finding New Drugs Is Laborious. This AI Startup Is Skipping The Lab Work

A new British startup named Potential Sciences has launched with seed funding in the tens of millions of dollars to build large-scale synthetic chemical datasets using physics-based simulations. The company aims to train advanced AI models to identify targeted drugs and fail unviable candidates quickly, bypassing traditional lab experimentation. While academics see promise in the simulation-based approach, experts note that AI-driven discovery still faces steep competition and does not eliminate the need for costly animal and human safety trials. Open article →