AI Didn't Make Our Jobs Easier. It Made Them Harder. And We're Here For It.

What the most disruptive shift in software history actually means for early-stage health tech investing.

If you’re plugged in at all, your Twitter feed the last few weeks has been a pretty specific thing: OpenClaw tutorials, people blasting out their setups, mac mini hard drives everywhere, everyone running agents and posting about what they built over the weekend. And then, intermittently, someone pointing out that 135,000 OpenClaw instances are exposed to the internet right now and Gartner just called it an unacceptable cybersecurity risk. The vibes are extremely “we are so back” and “this might be fine” at the same time.

And then yesterday, the guy who built it - solo, talking to AI instead of typing, as a playground project - announced he’s joining OpenAI. Not selling the company. Just going. Because he wants to change the world and that’s the fastest way to do it. He left money on the table from Meta, from Microsoft, from OpenAI itself. Just picked the mission.

That’s the moment we’re in.

Everyone Can Build Now. That’s The Problem.

What this means for people building in health tech is something we think about constantly. Because the same forces that let one person build OpenClaw and get Zuckerberg texting him on WhatsApp are available to every founder we talk to. A two-person team can now ship something that looks and functions like a real product in weeks. A founder with clinical insight and zero engineering background can build internal tools that would have required a funded engineering team not long ago.

For healthcare, that’s a genuine unlock. The sector has always been brutal to build in - long sales cycles, reimbursement labyrinths, credentialing requirements, workflows that resist change by design. Anything that lets a small team move faster and prove more with less capital is good for the ecosystem.

And it makes our job harder. Which is exactly why we’re here.

Because AI also makes it dramatically easier to build the appearance of a solution. The demo is more polished. The deck tells a more complete story earlier. The pitch sounds sharper. And sometimes beneath it, there isn’t much there yet. The floor on what looks fundable is rising fast while the floor on what’s actually fundable stays exactly where it always was. Finding that gap is the job.

No Shortcuts Here

There’s a version of this argument that goes: it doesn’t matter, because building fast to get acquired is a legitimate strategy. And honestly, in healthcare, that’s sometimes true.

Clinical documentation, revenue cycle automation, the AI layer inside health systems - these are categories where a small team can build something real, get traction inside a few health systems, and get acquired by Epic or Oracle or a large payer before they’ve had to prove long-term durability. The acquirer already has the distribution and the institutional relationships. They just needed the tech and the team. That’s a real exit path and there’s nothing wrong with it. That wave is coming and it’s going to move fast.

But that path is narrow. And it’s almost entirely in the workflow and infrastructure layer of healthcare - not in the consumer health, longevity, women’s health, and metabolic categories where we actually invest.

In those spaces, you can’t fake your way to acquisition. The acquirer - a payer, an employer, a scaled consumer health platform - wants proof that people actually got healthier and came back. Outcomes data, retention curves, evidence that the behavior change stuck past month three. You cannot vibe code your way to that. It takes time and trust earned at the individual level, and it requires a founder who understands why people fall off and how to design against it.

The Thesis Gets More Interesting, Not Less

Our thesis at Pave has always been that the biggest health companies won’t pick a side between consumer wellness and the healthcare system - they’ll connect the two. Consumer trust at the individual level, layered into reimbursement pathways as they scale. Noom and Omada proved the model works. The question now is who builds it in women’s health, metabolic disease, mental health - the categories where clinical evidence and consumer willingness are both present but the bridge hasn’t been built.

What AI actually changes here is timing. Building that kind of dual-channel company used to require capital and operational sophistication that pushed it to later stages. A lean team can now build more of the infrastructure earlier. That doesn’t make it easy. It makes it executable at seed and pre-seed where it previously wasn’t. That changes what we can fund and how early we can get there.

Where We're Placing Our Bets

The pre-seed valuation environment in health is about to get messier. More founders can get to a convincing demo faster, with less capital, which means early rounds are going to inflate on vibes before the market corrects. That’s not a reason to slow down - it’s a reason to get sharper. We invest at ideation. First check. Before the demo exists, before the deck is polished, before anyone else is in the room. That’s always been our posture and it’s never been more important than right now.

At that stage, the signal was always founder over traction. Now it’s founder over everything - because traction can be manufactured faster than ever. What can’t be manufactured is genuine pattern recognition about how healthcare actually works, a network that tells you which founders are the real deal before they’ve raised, and the kind of lived experience inside health that lets you stress test a thesis at the idea stage. That’s what we bring. That’s how we cut through the noise.

We think the founders who define health over the next decade are building right now, probably underestimated, probably pre-revenue, almost certainly pre-product in any traditional sense. AI has made it possible to find them earlier and back them with more conviction. The ones who understand both the tools available to them and the structural realities of what they’re building in are going to move fast and build things that last.

That’s who we’re looking for. We know how to find them.