Jensen Huang says AI will hit life sciences hardest; Lilly building on it

The column's starting point is a familiar complaint about drug development: most compounds that enter clinical trials never reach the market, and those that do can take a decade, with average spending of about $2.6 billion before approval. Against that backdrop, Eli Lilly built LillyPod, described as the most powerful supercomputer in the pharmaceutical industry, to apply AI to the earliest and most expensive stages of research. The columnist stresses that AI cannot replace rigorous clinical trials, but even modest gains — a 5% to 10% reduction in time and cost — would be significant for a company with many candidates across various phases.

A second thread is data. Eli Lilly's chief scientific and product officer, Daniel Skovronsky, has noted that AI model quality depends heavily on training data, and the company holds extensive records of clinical trial successes and failures, especially in diabetes and weight-loss markets. That archive, the column suggests, may be hard for rivals to replicate.

Even so, the columnist declines to call the stock a buy on AI alone, pointing to other large drugmakers pouring money into the technology. The case for Eli Lilly instead rests on recent results — third-quarter revenue up 39% year over year to $24.4 billion, led by Mounjaro and Zepbound — a broad pipeline, and a dividend that has risen 239% over the past 10 years despite a forward yield of 0.5%.