The Real Payoff of the AI Boom May Be Better Medicines

U.S. healthcare spending reached $5.3 trillion in 2024 — 18% of GDP — including $467 billion in prescription drugs. Few industries offer AI a clearer path from computational capacity to economic value than medicine, argue Viswa Colluru, founder and CEO of Enveda, and Bigyan Bista, the company's head of capital formation.

The AI buildout has moved fast from the digital to the physical: chips, data centers, cloud infrastructure and new power sources. But the consequential question is what valuable work all that capacity will enable.

At the World Economic Forum's 2026 Annual Meeting in Davos, Nvidia CEO Jensen Huang described AI as a five-layer stack: energy, chips, cloud infrastructure, models, and applications. The first four layers create capacity. The application layer turns it into products, services, and measurable outcomes.

No single industry needs to carry the entire economics of the AI buildout, the authors write. Returns will emerge across industries and through thousands of products. Yet medicine stands out, because a therapy that materially changes the course of a common or serious disease creates value in several directions at once: better patient lives, reduced need for other care, healthier and more productive people, and substantial returns for the developer and its investors.

Access, affordability and real-world outcomes determine how widely that value is shared. At its best, the authors argue, a successful medicine lets commercial success and social benefit reinforce one another.

The scale of a single molecule

Eli Lilly's tirzepatide illustrates the commercial stakes. In 2025, the company reported approximately $23 billion in sales for Mounjaro and $13.5 billion for Zepbound — roughly $36.5 billion combined for the two brands built on the same molecule. Those sales say nothing about whether AI will discover the next blockbuster. The consequential question, the authors write, is whether AI can make such successes more frequent by cutting unproductive experiments and improving decisions along the way.

There are reasons to think it can. Drug discovery is a long sequence of decisions under uncertainty: which chemical scaffold or biological mechanism to pursue, which experiment to run next, which safety signals matter, which patients will benefit. AI can help researchers rank possibilities, identify relationships across large and varied datasets, and design experiments that produce more useful information.

But AI does not make weak biology or poor data disappear. Scarce data, inconsistent measurements, and the complexity of living systems remain formidable obstacles. The technology works best when computational models connect to high-quality experimental data and laboratories capable of testing predictions quickly.

The most effective systems create a learning loop in which each experiment sharpens the next decision — and the quality of that loop, the authors argue, matters far more than the sheer number of hypotheses a model can generate.

Enveda offers its own proof point. AI models for mass spectrometry allowed its scientists to read an organism's chemical code more comprehensively, leading to the discovery of a new hormone that might capture the benefits of exercise. The company developed it into a candidate medicine with positive Phase 1 trial results in just four years.

A field still awaiting its scoreboard

The field remains nascent. In an August 2026 piece in Nature Reviews Drug Discovery, leading authors assessed a decade of progress and concluded that despite extensive model development and benchmarking, evidence of clinically relevant impact remains disappointingly limited. Among their recommendations: shift toward evaluating whether AI actually improves real drug-discovery decisions.

The field should welcome that challenge, the Enveda executives write, because it gives AI-enabled drug discovery the right scoreboard — translation.

Drug development has a stubborn failure mode: results that look compelling in a model or a laboratory often do not hold up in people. Generating more molecules and hypotheses at greater speed does little to solve that if the candidates simply enter the same attrition funnel.

For executives and capital allocators, the central question is whether AI improves the odds of eventual clinical success at each stage, from early discovery through trials. That requires evidence that models enable better choices of biology or chemistry, that candidates with a higher probability of success advance to the clinic, and that those candidates produce stronger trial results and better patient outcomes.

Few outcomes, the authors conclude, would better justify the scale of today's AI investment: breakthroughs that arrive more often and reach patients sooner give millions of people more healthy years of life, hope, and meaning.