The VC Who Left $4 Billion Behind to Bet Small on AI Biology

Key points

• Vijay Pande left Andreessen Horowitz's (a16z) roughly $4 billion biotech fund in 2024 to launch VZVC, a far smaller AI-focused firm.
• Pande argues biology is shifting from a science of discovery to a science of engineering, with AI as the key reason why.
• Clinical trials, the required human testing before any new drug reaches patients, remain brutally expensive even with AI cutting earlier costs.
• Pande believes open, shared datasets will do more to advance AI medicine than privately locked-up data.

What actually changed Pande's mind about big funds?

Running billions is prestigious. Pande walked away anyway. After managing a16z's roughly $4 billion biotech practice, one of the largest in venture capital, he launched VZVC with a fraction of that capital and a very different philosophy.

The logic is simple: big funds make many bets. VZVC makes fewer, and only in companies where AI is genuinely central, not decorative.

He told TechCrunch AI the firm is deliberately not doing 30 bets a year. Fewer bets means each one gets more time, more attention, and a harder look.

So what does 'biology as engineering' actually mean?

For most of history, biology has worked like detective fiction. Scientists observe nature, spot patterns, and slowly piece together how life works. That is the discovery model.

Engineering flips it. You define an outcome, design something to achieve it, build it, and test it. Pande's argument, backed by a career that included founding the groundbreaking Folding@home protein-research project at Stanford, is that AI is finally making that flip real in biology.

AI systems can now predict how proteins fold, how a drug molecule will behave, or which genetic tweak is most likely to work, all before anyone touches a test tube. That compresses years of lab work into weeks.

If AI is so powerful, why are drugs still so expensive to develop?

Here is the part that catches people off guard. AI can dramatically speed up early research. But before any drug reaches a patient, it must pass clinical trials: long, tightly regulated studies that test safety and effectiveness in real human volunteers.

Those trials cost hundreds of millions of dollars and take years. AI has not yet cracked that problem. Pande is candid about it. The discovery pipeline is getting faster; the human-testing pipeline is not keeping pace.

Does locking up data help or hurt?

It hurts, in Pande's view. Many biotech companies treat their datasets, the collections of biological and clinical information AI learns from, as a competitive weapon to be guarded.

Pande argues the opposite approach works better. Open, shared datasets let more researchers train better AI models, which produces better science for everyone, including the companies that share. The analogy is the internet: it became valuable because information moved freely, not because one company hoarded it.

For patients, that argument carries weight. Faster, cheaper drug development means treatments reaching people sooner.

Common questions

What is VZVC?

VZVC is the new venture capital firm Vijay Pande launched after leaving a16z. It is smaller and focuses specifically on companies that use AI as a core part of their biology or medicine work, rather than as a marketing addition.

Will AI actually make medicine cheaper for ordinary people?

AI can cut the cost of early drug research, which is a meaningful first step. Clinical trials remain expensive, so the full saving has not yet reached patients, but most experts expect AI to reduce total development costs over time.