Daphne Koller argues AI will not transform drug discovery simply by designing molecules faster because the bigger bottleneck is understanding human disease biology. More than 90% of clinical-stage drugs fail, while novel targets advanced annually fell from roughly 100 in 2015 to 30 in 2024.
Drug Discovery Has No Magic Wands
Daphne Koller
Research
14 Pages
Key Takeaways
Wrong targets dominate: More than 90% of drugs entering clinical trials fail, while 37 targets each have more than 50 active programs.
Novelty is shrinking: Novel targets advanced annually fell from about 100 in 2015 to 30 in 2024, despite rapid progress in AI-enabled molecular design.
Trials stay slow: About 50% of clinical development remains in-life biological observation, where feedback takes years and costs millions despite AI-driven operational efficiencies.
Join our newsletter to have all of this content + Exclusive Newsletter Bonus Content delivered to your inbox every week
