My bet that medicines will be some of the biggest consumer products of the next 20 years hinges on two questions: 1. Is the demand there? Emphatically yes. Tirzepatide, the molecule behind Mounjaro and Zepbound, already generates more in revenue than Netflix. Equally striking: more than half of these prescriptions are self-pay. People are spending their own money on medicines that meaningfully improve their lives. Even a product as good as tirzepatide leaves plenty of room for improvement. And weight loss and diabetes are just one part of the opportunity. Better sleep. Less pain. Fuller hair. Clearer skin. Better treatments for cancer and cardiovascular disease. The unmet needs are enormous. The demand question is relatively easy—the harder question is below. 2. Will we be able to build these medicines? My bet is yes. But this is where most of the uncertainty lies. Three reasons I’m optimistic: - Drug development is becoming like software. Increasingly, a lot of the research & analysis that is required at all steps of drug development can be done via LLMs. As a result, these steps can be composed within systems and automated. This allows small AI-native teams to do in days & weeks what required dozens of PhDs months to do. With billions going into AI × biology, I expect this trend to only accelerate. - Our biological toolkit is expanding. CRISPR and base editing. siRNA and mRNA. Bi/tri-specific antibodies, antibody-drug conjugates and targeted protein degradation. CAR-T and emerging in vivo CAR-T approaches. We have more ways to intervene in biology. Better delivery technologies and half-life extension can help medicines reach the right tissues and work longer between doses. Together, these advances expand the range of medicines we can build. - I have a positive regulatory bias. While regulation is always tough to predict, some of the recent steps from the FDA have been positive. FDA is developing clearer pathways for historical patient data and Bayesian methods in trials, while advancing validated alternatives to animal testing. These changes could reduce development time and cost. As unmet needs continue to grow, I expect more pressure on FDA to keep modernizing. Twenty years is a long time for these improvements to compound. Why I could be wrong: - Biology takes time. Some steps in drug development are inherently incompressible. For example, establishing long-term safety and clinical benefit still requires observing people over time. - Biology is hard to predict. Compelling mechanisms can still fail in large trials. Last month, Novartis’s long-awaited Lp(a)HORIZON trial showed that pelacarsen lowered Lp(a) but failed to demonstrate a statistically significant reduction in cardiovascular events. This contributed to a selloff that wiped roughly $30B off Novartis’s market value. These constraints are real. But teams that combine scientific rigor with commercial discipline can bring great medicines to market. Many of the biggest improvements we can still make to people’s lives are biological. Medicines can deliver them. What’s a better consumer product than that?

For more than a decade, I tried to meditate but never could. With Temple, I could finally see my slow progress, even when my racing mind felt nothing. Entropy is a delightful metric that helps me live a fuller, calmer life. And it also pushes me to workout more, and in the right ways (more on that soon). Join the waitlist for Early Access at temple.com And follow @temple for more updates.

Your heart has an age and it's rarely the one on your birthday cake. New peer-reviewed study: @UltrahumanHQ 's Cardio Age lines up with how you sleep, move & recover. When people got fitter, it dropped ~3 years in a year. frontiersin.org/journals/dig…

Inspired by Founders Fund, we shot a game of Mafia in Bangalore. Here’s Episode 1 ☺️