Here's a paradox for you. Q1 of this year just posted the biggest global venture capital quarter in history — three hundred thirty point nine billion dollars, according to KPMG. And yet if you talk to founders right now, they'll tell you it's the hardest fundraising environment they've ever seen. Both things are true. How does that work?
That's exactly the question Daniel sent in. He's been watching the headlines about VC drying up and investor caution, and he's asking whether the media is oversimplifying — treating thousands of independent fund managers as if they're one homogenous unit with one mood. And he wants to know what the actual mechanics look like on the ground for founders trying to raise.
It's a great framing, because the disconnect between the aggregate number and the individual experience isn't a contradiction — it's the whole story. The capital isn't scarce. It's hyper-concentrated. And if you're not in the concentration zone, the drought is real.
Crunchbase reported that AI companies captured over forty percent of all VC dollars in Q1. That means the non-AI portion of that record three hundred thirty billion is actually flat to down from previous years. So you've got this bizarre situation where total numbers are screaming "boom" but most founders are living in a bust.
Which makes the media narrative of "VC is drying up" both wrong and right at the same time. Wrong in aggregate, right for the person actually trying to raise a seed round for a B2B SaaS company in Cleveland.
That's what we want to trace today — not the headline, but the plumbing. How does capital actually flow from limited partners through venture funds to founders, and why does that pipeline create such radically different experiences depending on what you're building and where you are?
Because if you're reading the news and thinking the whole market is frozen, you're missing where the money is actually moving. And if you're raising, that's an expensive mistake.
Let's start with the homogenization error itself. When a reporter writes "VC is pulling back," they're collapsing thousands of independent fund managers — each with their own LP base, their own thesis, their own deployment schedule — into a single fictional character called The Investor. It's like saying "the restaurant industry is struggling" because fine dining in Manhattan had a slow quarter while taco trucks in Austin are printing money.
Covering the covers.
The media isn't covering the capital markets. They're covering other coverage of the capital markets, and each layer of abstraction smooths out the texture until you're left with a mood, not a market.
Moods are cheap to report and expensive to act on. If you're a founder and you internalize "VC is cautious right now" as a fact about the world, you might not notice that three AI infrastructure funds just closed oversubscribed rounds and are actively deploying.
That's the thing. The KPMG number — three hundred thirty point nine billion in a single quarter — is genuinely staggering. It's not just a record, it's a record by a wide margin. But if you strip out AI and AI-adjacent deals, what you're left with is roughly a hundred ninety-eight billion. That's the non-AI portion. And depending on which sector you slice, that's flat to down year-over-year.
The money didn't disappear. It just moved. Like water finding the lowest point in the terrain, except the terrain is shaped by hype cycles and LP allocation models.
That's where the Crunchbase data gets really revealing. Over forty percent of all VC dollars going to AI companies — that's not a preference, that's a gravitational field. It means every founder who isn't building something AI-adjacent is competing for a shrinking share of the remaining pie, even as the total pie gets bigger.
Which is why the "drying up" narrative feels true to those founders. Their experience isn't a misperception. The capital available to their category actually did shrink. They're just not in the category the headlines are measuring.
And this isn't some obscure market subtlety. It's the defining feature of the current fundraising landscape. You've got two completely different worlds — the AI world where rounds are oversubscribed and term sheets come in fast, and the everything-else world where six-month fundraising cycles are becoming standard and due diligence requests have doubled.
Why does the media keep running the homogenized version? Part of it is structural. A quarterly VC total is one number you can put in a headline. "Capital is hyper-concentrated by sector with bimodal outcomes depending on founder geography and thesis alignment" doesn't fit in a chyron.
Frankly, the aggregated number serves different audiences differently. LP consultants use it to justify allocation shifts. Fund marketers use it to signal momentum. Founders use it to psych themselves up or psych themselves out. Everyone has a reason to prefer the simple story.
Daniel's question gets at something sharper. He's not just asking whether the media oversimplifies — he's asking whether the actual reality in any major capital market is much more individualized. And the answer is yes, but with structure. It's not random. The individualization follows patterns.
That's the key. It's not that every investor is a unique snowflake and you can't say anything general. It's that the generalizations that matter are at the sector and stage level, not the aggregate level. The headline "VC hit a record" tells you nothing about whether your climate-tech seed round is fundable. The headline "AI captured forty percent of dollars" starts to tell you something real.
The paradox we opened with — record totals and founder pain coexisting — isn't a paradox once you stop treating "the venture capital market" as one thing. It's multiple markets wearing a trench coat.
Which is a perfect setup for looking at the actual plumbing. Because if you want to understand why capital is concentrating this way, you have to follow the money upstream.
Let's follow the money upstream. Every venture fund has its own investors — the limited partners. Pension funds, university endowments, sovereign wealth funds, family offices. These are the people who actually supply the capital that VCs deploy.
They're not doing it for the vibes. They've got allocation targets, return requirements, and a risk-free alternative sitting right there.
That's mechanism one — the LP liquidity squeeze. With interest rates still elevated, an institutional LP can park money in treasuries or high-grade bonds and pull five-plus percent with essentially zero effort. No J-curve, no capital calls, no waiting a decade to find out if a fund returned anything.
VC has to compete with the easiest five percent an LP has ever seen.
It's losing that competition in a lot of portfolios. The typical institutional allocation to venture was already small — maybe five to ten percent of alternatives, which themselves are a slice of the total. When risk-free yields rise, the spread between what VC might return and what bonds will return narrows. The risk-adjusted case weakens.
Which means fewer LPs writing commitments to new funds, and existing LPs re-upping at lower amounts.
And that flows straight through to founders. If a GP can't close their next fund, they can't write new checks. If they close smaller, they're more selective. The LP decision made in a boardroom in Oslo or Abu Dhabi determines whether a seed-stage startup in Austin gets a term sheet.
The "drought" doesn't start with the VC saying no. It starts eighteen months earlier with the pension fund rebalancing its alternatives exposure.
The data bears this out. New fund formation has slowed significantly. The funds that are getting raised are concentrated among established names — Sequoia, a16z, Benchmark, Lightspeed. The emerging managers who might have closed a hundred-million-dollar debut fund in twenty twenty-one are struggling to hit fifty now.
Which brings us to mechanism two — the flight to quality inside the VC firms that do have capital. You're a partner at a fund. Your LPs are nervous. You've got portfolio companies that need bridge rounds. What do you do?
Fewer deals, larger checks, higher conviction. The power law of venture returns gets more extreme in a tight environment because the cost of being wrong goes up. If you can only make ten investments this year instead of twenty, every single one has to count.
"counting" means something specific right now. It means revenue traction, not vision. It means proven product-market fit, not a compelling deck. The days of funding a founder because they had a great narrative and a Stanford degree are — not gone, but severely diminished.
The PitchBook data on this is stark. AI and ML deals are commanding two to three times the valuation multiples of comparable non-AI SaaS deals. So if you're a VC with limited dry powder and LPs watching your every move, where are you going to put the money? Into the category where multiples are expanding, or the one where they're contracting?
It's not even a decision at that point. It's fiduciary gravity.
That's mechanism three — the AI magnet. It's not just that AI is getting more funding. It's that the valuation premium actively pulls capital away from other sectors. If you can pay a higher multiple and still have better exit prospects because the whole market is chasing AI, the rational move is to allocate there.
Creating a vacuum effect. The money doesn't just flow toward AI — it evacuates other sectors. Climate tech, biotech, hardtech, enterprise SaaS — these aren't bad categories. They're just not AI, and right now that's the only filter that seems to matter.
The KPMG numbers make this concrete. That record three hundred thirty billion? Strip out AI and you're at roughly a hundred ninety-eight billion for everything else. Flat to down from twenty twenty-four. So the experience of a non-AI founder isn't psychological — their addressable capital pool actually shrank.
The way this manifests for founders is brutally practical. Fundraising cycles that used to take three to four months are stretching to six to nine. Due diligence requests have multiplied — investors want customer references, detailed unit economics, cohort retention data, things they might have glossed over when capital was abundant.
There's also a sharp shift toward revenue traction over vision. Two years ago you could raise a Series A on a compelling narrative about the future of work. Now you need to show real revenue growth, ideally with AI somewhere in the story, even if it's just a feature.
The "AI-adjacent" tax. Founders are bolting chatbot interfaces onto products that don't need them, not because it makes the product better, but because it makes the fundraising deck fundable.
Here's the thing — compare two founders in Q1 of this year. One is building AI infrastructure, model orchestration, something in the stack. She's getting inbound interest from partners who haven't done a cold outreach in years. Term sheets in two weeks. Valuation multiples that make her previous round look quaint.
The other founder is building a good B2B SaaS product in supply chain logistics. Growing revenue, happy customers, solid unit economics. He's been fundraising for seven months. His deck has been through seventeen revisions. He's heard "we love the team but we're not doing new SaaS bets right now" from fourteen different funds.
Same macro environment. Same headline about record VC totals. Completely different realities.
That's the LP-to-founder chain in action. The pension fund's allocation shift becomes the VC's smaller fund, which becomes the partner's heightened selectivity, which becomes the founder's seven-month fundraising slog. Every link in the chain transmits and amplifies the pressure.
Those mechanisms don't just affect fundraising — they create knock-on effect that reshape the entire startup ecosystem. And one of the strangest is the zombie fund phenomenon.
Sounds like a B-movie.
It kind of is. Hundreds of small VC funds raised in twenty twenty and twenty twenty-one are technically still in their deployment period. They exist on paper. Their websites are up. But they can't raise a follow-on fund, so they're hoarding their remaining capital for follow-on investments in existing portfolio companies. They're not writing new checks.
A founder sees a fund that invested in their space two years ago, reaches out, and gets... Or worse, a polite meeting that goes nowhere.
It creates this illusion of available capital that isn't actually flowing. The fund count in Crunchbase and PitchBook looks healthy. But the number of funds actively making new investments is much smaller. It's like walking through a supermarket where half the products on the shelf are empty boxes.
Which makes the spray-and-pray approach especially dangerous right now. A founder blasts their deck to a hundred funds, gets ninety-five no-responses or soft passes, and concludes the market is frozen. When really they just aimed at a lot of zombies.
That's the practical implication. With capital this concentrated, targeted outreach to the right ten to fifteen funds — ones actively deploying in your specific sector — yields far better results than broad shotgun approaches. You need to know who actually has dry powder and a mandate to use it.
Which requires research that most founders don't do. They download a list from Crunchbase and start emailing.
The second knock-on effect is the rise of alternative financing. Revenue-based financing, venture debt, rolling funds — these have grown from about five percent of early-stage startup financing in twenty twenty-one to roughly fifteen percent now.
Platforms like Pipe and Capchase basically said, traditional VC is too slow and too selective for a huge swath of viable companies, so we'll advance capital against recurring revenue and skip the equity dilution entirely.
It's not just a stopgap. For a SaaS company with predictable revenue, revenue-based financing can be better than selling equity. You pay it back out of cash flow, you keep ownership, and you don't spend six months fundraising.
The catch is you need revenue. Pre-revenue startups can't access it, which means the selectivity problem we've been describing hits earliest-stage founders hardest. They're stuck with traditional VC or angels, both of which are concentrated in AI.
The third effect is geographic. The Bay Area and New York now capture about sixty-five percent of all US venture dollars, per PitchBook data. That's up from already high levels. The concentration isn't just by sector — it's by zip code.
Take a climate-tech founder in the Midwest. Same month, same macro environment, building something important. She's competing for a shrinking pool of non-AI capital, most of which is concentrated on the coasts, from funds that are being more selective than ever. Her AI counterpart in San Francisco is fielding inbound term sheets.
It's not that the Midwest founder is doing anything wrong. The terrain is just tilted. And that tilt has consequences. If the best climate-tech ideas can't get funded because they're in the wrong city and the wrong sector relative to the AI magnet, we lose more than returns — we lose solutions.
Which is where the "spray and pray is dead" point really lands. If you're that Midwest founder, sending cold emails to two hundred coastal VCs is a waste of time. What works is identifying the ten funds that explicitly invest in climate tech outside the coasts, that have raised recently enough to still be deploying, and that aren't zombie funds pretending to be active.
Those funds exist. There are specialized climate-tech investors, agtech funds in the Midwest, hardtech funds that explicitly avoid the AI hype. But you have to find them. The shotgun approach doesn't just fail — it demoralizes you into thinking no capital exists anywhere.
When really it's just hiding in plain sight behind the aggregate numbers.
What do you actually do with this understanding? Let's make it practical, because Daniel's question really has three answers depending on who's asking.
Right — and the first answer is for founders. Stop reading aggregate VC news. The quarterly total is not your market. Your fundraising environment is determined by your sector, your stage, and your geography. Not by the three hundred thirty billion dollar headline.
It's almost worse than useless. If you're a non-AI founder and you see "record VC quarter," you either feel gaslit or you start doubting your own pitch. Neither helps you raise.
The actual move is to research which specific funds are actively deploying in your niche. Not which funds exist, not which funds invested in your space three years ago — which ones have dry powder and a current mandate. That's a list of maybe ten to fifteen names, not two hundred. And those are the only ones that matter.
The way you find them is not by downloading a Crunchbase export. You look at who closed a fund recently. You look at who's been writing first checks in the last six months. You talk to founders who raised in your sector and ask which partners actually showed up with a term sheet, not just a coffee.
The zombie fund problem makes this essential. If you don't filter for active deployers, you're going to waste months talking to funds that are technically alive but functionally dead. They'll take the meeting — they don't want to look inactive — but they're not writing checks.
Which is demoralizing in a way that's completely avoidable. It's not that the market rejected you. It's that you pitched a fund that hasn't made a new investment in eighteen months and didn't tell you.
The second answer is for investors — and this one's uncomfortable. The current environment is punishing generalism. Funds with a clear thesis — AI, climate, defense tech — are raising easily. Funds that lead with "we invest in great teams" are struggling.
Because "great teams" isn't a strategy LPs can underwrite anymore. They want to know why you'll see deal flow others miss, why your portfolio construction makes sense, and why you won't get steamrolled by the AI gravitational field.
The data backs this. The funds closing oversubscribed right now are the ones that can say, we invest in X at Y stage with Z advantage. The ones that can't articulate that are seeing their limited partners quietly reduce commitments.
The market isn't just selecting for AI — it's selecting for clarity. If you're a generalist fund, the question isn't whether you'll survive. It's whether your LPs will re-up when the alternative is a specialized fund with a sharper edge.
The third answer is for everyone else — the observers, the journalists, the policy people. When you hear "VC is drying up," the only useful response is to ask: for whom, and in what sector?
That's the whole game. The headline "VC hit a record" and the headline "founders are struggling to raise" are both true, but they're true about different populations. The media's homogenization error isn't just sloppy — it actively obscures the capital concentration that's reshaping who gets to build.
If you're a policymaker worried about innovation, this distinction matters enormously. If capital continues to concentrate in AI and on the coasts, you'll see a lost generation of non-AI startups — not because the ideas are bad, but because the funding infrastructure for them is atrophying.
Which is a policy choice, not a law of nature. The LP squeeze, the flight to quality, the AI magnet — these are all responses to incentives. Change the incentives, and the capital flows differently.
The practical through-line for all three audiences is the same: disaggregate. Stop treating "the venture capital market" as one thing. The number that matters isn't the global total. It's the number that applies to your sector, your stage, your geography. Everything else is noise.
Noise is expensive.
There's one question this whole analysis leaves open, and it might be the most important one. Is the AI capital concentration a bubble that redistributes when the next hype cycle hits, or is it structural — a permanent reorientation of where venture dollars live?
That's the fork in the road. If it's cyclical, then non-AI founders just need to survive until the money rotates back. Biotech, climate, hardtech — they've had their moments before, and they'll have them again. The drought is painful but temporary.
If it's structural?
Then we're looking at something closer to a lost generation of non-AI innovation. Startups with strong fundamentals in important sectors that simply can't get funded, not because they're bad bets, but because the entire capital allocation machinery has reorganized around one category.
That's not a market failure in the narrow sense. The money is going where returns look best. But it's a policy problem. If venture capital is how we fund the future, and the future we're funding is exclusively AI, we're making a very large bet with very little diversification.
The historical parallel that worries me is what happened to cleantech after the twenty ten bust. Capital fled, and it took nearly a decade for climate-tech funding to recover. In the meantime, a lot of good ideas died on the vine — not because the technology didn't work, but because the funding infrastructure collapsed.
This time the evacuation isn't from one sector — it's from every sector that isn't AI. The gravitational field we described is pulling capital out of biotech, out of materials science, out of everything that doesn't have a machine learning component in the pitch deck.
Which is why I keep coming back to that PitchBook number — AI deals at two to three times the multiples of non-AI SaaS. That's not a market signal about relative value. That's a market distortion that makes non-AI investing look irrational even when the underlying businesses are perfectly sound.
The open question for anyone watching this space is: what breaks the magnet? Does AI have to underperform for capital to redistribute, or does the next hype cycle — synthetic biology, quantum, whatever — simply pull money in a new direction?
Or does nothing break it, and we just accept that venture capital has become, functionally, the AI financing industry with a few niche exceptions around the edges?
That's the version that should make policymakers nervous. Because if that's the equilibrium, the startups solving problems AI doesn't touch — and there are a lot of them — need a different funding infrastructure entirely. Government grants, corporate R and D partnerships, something that fills the gap venture used to fill.
We're already seeing early signs of that. The CHIPS Act, the Inflation Reduction Act's climate provisions — those are, in effect, public-sector responses to the private sector's capital concentration. The government is stepping in where venture stepped out.
Whether that's a feature or a bug depends on your politics. But it's happening, and the capital concentration we've traced today is the mechanism driving it.
The question Daniel's prompt ultimately points toward isn't just "is VC drying up." It's "what kind of innovation are we choosing to fund, and what are we choosing to starve?" The answer to that question is being made right now, deal by deal, in the gap between a record headline and a founder's seventh month of fundraising.
Now: Hilbert's daily fun fact.
Hilbert: In the eighteen eighties, a Russian naval officer stationed on the Kamchatka Peninsula transcribed a series of whale songs into musical notation — the only surviving record of a specific North Pacific humpback population's vocal patterns before commercial whaling collapsed their numbers. The transcriptions sat unexamined in a Saint Petersburg archive until two thousand nineteen, when bioacoustics researchers realized they'd captured a dialect that no living whale still sings.
A ghost dialect.
That's haunting.
This has been My Weird Prompts, produced by Hilbert Flumingtop. If you want more episodes, find us at my weird prompts dot com.
If you've got a question like Daniel's — something where the headlines and the reality don't quite square — email the show at show at my weird prompts dot com. We'll dig into the plumbing.
