Venture is where Wall Street was

Venture still picks winners the way Wall Street did in the 1980s: warm intros, gut feel, and who you know. Wall Street changed when a math professor with no network built a system that saw what the relationships missed. Venture's turn is coming, and the open question is who builds the machine.

A few weekends ago I watched a documentary about the quant revolution on Wall Street. I expected a finance story. What I got was a story about who gets let into the room.

In the 1980s, Wall Street was a relationship business. Know the right people, read the room, trust your gut, and you did well. Jim Simons fit none of that. He was a math professor with no finance background and no network on the Street. What he had was a bet that a system could see what all those relationships were missing.

So he built a different kind of team. Mathematicians, physicists, speech-recognition scientists. People nobody on Wall Street would have thought to call. He pointed them at decades of market data most firms weren't even looking at, and the industry mostly shrugged.

Here's the part I love. Simons didn't need to be right every time. He needed to be right a little more often than chance, across an enormous number of trades. His Medallion Fund went on to post some of the strongest returns in the history of investing.

He didn't beat Wall Street at its own game. He played a different one.

When the credits rolled, one question stuck with me. Which industry looks today the way Wall Street looked then?

The uncomfortable answer, for those of us who work in it, is venture capital. Mine included.

Venture still runs on warm intros, gut feel, and pattern-matching on pedigree. A founder's odds of getting a meeting often depend on who they know before what they've built. Most funds see a small fraction of the companies worth funding, and they see them late. The warm intro is our trading floor: loud, personal, and built for a market that is disappearing.

That model held up when the number of fundable companies was manageable. It is harder to defend now. AI has collapsed the cost and time it takes to build, and a small team can ship in weeks what used to take a funded company a year. The supply of credible startups is growing faster than any partner's calendar can absorb.

So here is the question. If the number of companies worth looking at keeps compounding, and the way investors find and judge them stays the same, does the model hold? And if it breaks, what replaces it?

Three posts crossed our team's feeds recently that point at the same pressure from different angles.

Ben Leventhal, who built Resy, shared a cold email he'd received from an AI assistant that handles investor scouting for a founder. The agent had studied Ben's own public analysis, matched the company to how he thinks, laid out the traction, and framed the whole thing as a deal alert rather than a pitch. His caption: "I have no idea who is pitching whom what here." The agents are already in investors' inboxes. The real question is whether the investor on the other end has one too.

Sarah Guo, founder of Conviction, was blunter: "please stop sending me your ai-generated decks. they smell like total lack of thought." In a follow-up she clarified that AI isn't the problem. The problem is a lack of original thought and the inability to steer the models toward something good. My read is that AI has made a polished deck nearly free to produce, so the deck alone no longer tells an investor much. The signal has to come from somewhere else.

Christian Keil put a number on it: "Roughly half of the decks I receive these days are Claude-generated, with ~default settings." When half the inbound looks the same, judging companies on presentation stops working. Funds need a way to compare companies on what they've actually built, not on how their slides look.

None of them are saying venture is dead. They are saying the inputs changed faster than the process.

When we started talking about this internally, the answer depended on where you sat.

From the product side, our colleague Farzad Nazifi put it plainly: "VCs have funded the disruption of taxis, hotels, and real estate, but almost never their own industry." Venture has spent decades backing companies that turned relationships into systems. It has rarely turned that lens on itself.

From the founder side, Josh May framed it as a race for access: "Right now a founder has to stop building to go pitch, take this meeting and that meeting, and hope someone gets a feel for them. The ones acting on the quick are the ones that are going to win. Whoever owns the bridge between founders and capital owns the market."

From the investor side, Rick Spencer starts with the proof: "Venture isn't broken. It's evolving. I can point to checks where I verified some things and missed others, and the signals we're surfacing now would have caught what I missed. It's getting close to impossible for one person to find the right deal at the right moment and know it's the right deal. The point is increasing the probability of success and decreasing the time."

Put those together and the pattern is plain. Founders are moving faster than the people funding them. Investors know their sourcing is narrower than it should be. And the tools that reshaped every other market haven't been built for this one yet.

Be precise about what the Simons story says. Renaissance didn't eliminate human judgment. It changed what that judgment had to work with, so its people stopped relying on what one person could see from one desk.

Venture doesn't need to stop trusting conviction or relationships. At the early stage they matter, and they always will. What it needs is a wider field of view: every company worth seeing, seen early, in a form that can be compared and acted on. That makes good investors better and faster. It doesn't make them obsolete.

Market after market has moved from relationships to systems. Public equities did. Advertising did. Lending did. The interesting question isn't whether venture follows. It's who builds the rails when it does, and which funds choose to run on them.

If you allocate capital and you can feel your sourcing narrowing, or you're a founder who has watched a good company get skipped because nobody made the intro, reply and tell me. I read every reply.

Xella Health plugs Oura ring data into clinical care. Xella, the Fund I women's precision-health company, now lets members choose to share Oura ring data, including sleep, heart rate variability, temperature trends and cycle data, so its clinicians read it alongside Xella's own diagnostics and care plans. The focus is fertility, perimenopause, cardiometabolic and hormonal health. The read for founders: wearables have been collecting this data for years, and the value now sits with whoever turns it into a clinical decision, a position a lab-backed care company can hold and a consumer app cannot. Read more

Supersede's Indiana plant is open. Supersede opened a 100,000 sq ft plant in Bristol, Indiana, in Elkhart County, with more than 50 jobs planned, making recycled-polypropylene structural panels that replace wood sheathing and vinyl for RV makers, plus marine, modular housing and construction. It is the follow-through on the debt facility we covered in Issue 3, and it puts production next to its buyers: Coachmen already uses its Marine Board in Catalina travel-trailer floors. The read: the fastest route to scale for a materials company is to build where its customers already are. Read more

Azra Games' UnGodly on screen at Apple's iPhone launch. Azra's dark fantasy RPG UnGodly was shown during Apple's launch of the iPhone 18 Pro Max and iPhone Duo, running at 120 Hz ProMotion with a new Ultra graphics setting, and it is open for pre-order on the App Store, Google Play, Galaxy Store and ONE Store. The read: platform owners pick the titles that show off new hardware, and being chosen for that moment is distribution a studio cannot buy. Read more

Fundraising was built for a world of inboxes, pitch decks, and warm introductions. As AI agents become the new interface for how we discover, evaluate, and act on information, the fundraising process is changing too. At Pitch Protocol, we’re building the infrastructure for agents to connect founders with the right capital based on what they’re actually building, not just who they happen to know. From the studio, we’re building toward a world where better companies get discovered by better systems. See how it works at pitchprotocol.vc.