Something is happening this summer that I did not expect, and I want to write it down while it is still surprising — because in a year it will feel obvious, and the useful thinking always happens before things become obvious.
In the middle of August, which is the deadest month in academic science, our Head of Global Strategic Partnerships has opened eleven serious conversations with oncology researchers at USC, Stanford, UCLA, UC Davis, UC Irvine, and UCSF. Not introductions. Not polite acknowledgements. Material collaboration discussions, material transfer agreements, an active drug discovery partnership, and one researcher who liked the conversation enough that he has begun opening doors for us himself.
These are not people who need us. They are people with tenure, laboratories, reputations, and options. And they are answering emails from a small family office most of them had never heard of six months ago.
I have been thinking about why.
The machine that stopped cycling.
The venture capital industry has spent a decade running a self-reinforcing loop: raise money from pension funds and endowments, invest it in companies, sell those companies or take them public, return the proceeds, raise again. It worked well enough that it became the default assumption about how innovation gets funded.
That loop has broken. The exit routes closed — the IPO window shut, regulators made large acquisitions harder — and the money stopped coming back. Venture funds have drawn something close to two hundred billion dollars more from their investors than they have returned. For the funds raised at the peak, investors have received back roughly eight cents on the dollar. The institutions that supply this capital have responded rationally: they have stopped supplying it. Fundraising is down something like seventy percent from its high.
At the same time, artificial intelligence has absorbed most of what capital remains. An AI infrastructure company raises its Series A at a median valuation around three hundred million dollars. A functionally identical company that does not have AI in its description raises at fifty-five million. This is not a small distortion. It is a gravitational event, and everything that is not AI has been pulled off course by it.
Biotech has taken this worse than most sectors, for reasons that were always structural. Life sciences needs more capital, more time, and more tolerance for failure than software ever did. A drug candidate can take fifteen years to reach a patient. That timeline was always in tension with a fund that must return capital in ten. In a generous market, the tension was manageable. In this one, it is not.
What this did to the scientists.
Here is the part that rarely gets discussed outside the industry, and it is the part that matters most to us.
A senior researcher at a top American university does not operate alone. Their work advances through partnerships — with companies that provide sponsored research funding, that license discoveries and develop them, that co-author papers and share platforms and infrastructure. For twenty years, the reliable source of those partners was the venture-backed biotech sector.
That sector is now in trouble. The company that was going to take a lab's discovery forward has shut down, pivoted to AI, or is quietly unable to honour what it promised. Government funding is competitive and politically exposed in ways that make long-horizon planning difficult. And so researchers who are accustomed to being courted find themselves in an unfamiliar position: looking for a partner who will still exist in five years, and finding fewer of them than they expected.
We did not become more attractive. The alternatives became less available. It is important to be honest about which of those two things happened.
When our outreach arrives in that environment, it does not read as a pitch. A family office with no fund life, no exit clock, platforms that already exist and have been tested, and a scientific framework that treats human and animal cancer as one problem rather than two — for someone who has watched three partners disappear in eighteen months, that is not a solicitation. It is an answer to a question they had already been asking themselves.
This is why the response rate makes sense. And it is why I want to be careful about how we interpret it.
The mistake I want us not to make.
The comfortable reading of this moment is that we were right all along and the market has finally caught up. There is something to that. We built BioFund deliberately outside the venture model, and the venture model is now in visible difficulty. It would be easy to treat this as vindication and to relax into it.
I think that would be a mistake, for a simple reason: markets recover. Biotech public offerings are already showing signs of life in 2026. Capital will loosen. The funds that are struggling today will raise again, or new ones will take their place, and in three or four years the researchers who are talking to us now will have options again.
If our advantage is only that we are the last credible partner standing, then our advantage expires the moment someone else stands up.
So the question is not how to enjoy this window. It is what to build inside it that survives its closing.
What survives the recovery.
Relationships do. That is the whole answer, but it deserves unpacking.
A collaboration that begins now — a material transfer agreement, a shared cell line, a co-designed study — does not evaporate when the funding environment improves. It matures. It becomes sponsored research, then a licensing conversation, then in some cases a spin-out. And when a university or a new venture goes looking for growth capital in 2029, the question of who was there first is not just a nice story. It shapes who holds the platform relationship, who holds the license, who has a small stake in what comes next.
Here is the part I keep coming back to. In a tight market, we matter because we are one of the few stable partners left standing. In a recovering market, we matter because we were already there — already known, already trusted, already working alongside these labs before anyone else came knocking. Either way, we come out ahead. Not because we planned it that cleverly, but because patience was always going to pay off eventually, and it happens to be paying off now.
Which means the in-kind work happening right now — the platform access, the pilot support, the collaborative studies that generate no revenue and take real time and real people — is not charity, and it is not us waiting our turn. It is exactly the kind of building we said we would do when we started BioFund. Relationships now, science now, and whatever revenue, licenses, or shared ventures follow later will follow because the work itself was worth doing first.
Which brings me to capacity.
There is a very real question underneath all of this optimism, and it deserves an honest answer.
Eleven conversations is more than BioFund, at its current size, can properly attend to. Each one that matures wants something real: material, platform time, study support, scientific attention from a team that is already fully committed. Credibility and capability are no longer the question — the response rate has settled the first, and the platforms have settled the second. What's left is capacity. And unlike the other two, capacity is not something we have to prove. It is something we get to grow into.
We closed BioFund at seven families deliberately, and that decision stands. But the friends who have asked, repeatedly, whether there is a way to stand alongside this work are asking at precisely the moment when the answer has become genuinely exciting. The window is open wider than we can currently reach through it, and reaching further is simply the next thing worth building.
That is the origin of The BioFund Circle, and it is why the timing is not accidental. A small, named group of people who understand what this moment is, who want to be structurally inside it rather than adjacent to it, and who are content to be silent partners in something whose returns arrive slowly and whose meaning arrives immediately.
The distributions will be modest for years. I have said so plainly everywhere it is written down. But the people who join now are joining before the compounding begins — before the collaborations become licenses, before the licenses become platforms, before the platforms become the quiet infrastructure underneath a decade of cancer research. That is not a financial argument. It is a timing one. And for the right person, timing is the more interesting of the two.
A note on gratitude.
None of the above happens without someone doing the work. Neeki has spent this summer making calls that most people in her position would have postponed until September, to people who had every reason not to answer, and she has done it with a combination of scientific fluency and personal warmth that cannot be taught and cannot be delegated.
The market conditions explain why the doors were unlocked. They do not explain why anyone walked through them. That part is hers.