Last week, I watched Hippocratic AI raise $126 million at a $3.5 billion valuation. Then Mercor pulled in $350 million and hit a $10 billion valuation. Then General Intuition launched with a $133.7 million seed round. A seed round. Over a hundred million dollars. For a company that just launched.
The AI funding boom is reaching levels that make the crypto bubble look almost reasonable, and I can't decide if I'm excited about the innovation or worried about the inevitable crash.
The Numbers Are Absolutely Wild
Let me give you some context on what's happening right now. In just the first few days of November, AI startups collectively raised over $1.8 billion. Not for the month—for a few days. Hippocratic AI hit a $3.5 billion valuation with their healthcare AI agents. Mercor reached $10 billion helping companies find tech talent with AI. These aren't public companies with proven business models—these are startups, some barely a few years old.
But it gets wilder. OpenAI is now valued at $500 billion after their October secondary sale, making them the world's most valuable private company. Anthropic raised $13 billion in September at a $183 billion valuation. These numbers are so large they stop making sense. We're not talking about millions anymore—we're talking about hundreds of billions being poured into AI companies.
Why Investors Are Throwing Money at AI
Here's the thing: the money isn't flowing because investors are dumb. They're betting on a fundamental shift in how we work, create, and solve problems. AI isn't a feature anymore—it's becoming infrastructure. And infrastructure plays can justify massive valuations if you believe the market is big enough.
Take Hippocratic AI. They're building AI agents that can handle patient communication, appointment reminders, and basic healthcare admin tasks. They already have partnerships with 50+ healthcare organizations including Cleveland Clinic and Northwestern Medicine. Their pitch is simple: there's a massive healthcare labor shortage, and AI can help fill the gap. If they're right, a $3.5 billion valuation might actually be cheap.
Or look at Mercor. They raised $350 million to match employers with tech talent using AI. In a world where finding qualified engineers is expensive and time-consuming, an AI platform that can vet candidates and match skills sounds incredibly valuable. If it works at scale, $10 billion starts to make sense.
The Part That Worries Me
But here's where I get nervous: we're seeing massive valuations based on potential rather than proven economics. Hippocratic AI has $404 million in total funding. Are they profitable? Unknown. What's their customer acquisition cost? Unclear. How do they defend against competitors with deeper pockets? Good question.
The entire AI funding landscape feels like it's built on a series of assumptions: that AI capabilities will keep improving exponentially, that enterprise adoption will happen quickly, that these companies will achieve massive scale before burning through their capital, and that there's enough market for dozens of billion-dollar AI companies to coexist.
Any one of those assumptions could be wrong, and suddenly these valuations stop making sense.
The Foundation Model Dilemma
Here's another concern: foundation models like OpenAI and Anthropic require billions in compute infrastructure. That's why they're raising so much capital—training and running these models is absurdly expensive. Anthropic went from $1 billion in revenue to $5 billion in just eight months, which sounds great until you realize their infrastructure costs are probably eating most of that.
The application layer companies—the ones building specific tools on top of these models—have lower capital requirements but face a different problem: they're dependent on the foundation models. If OpenAI changes their pricing or API terms, or if their model quality degrades, all those application layer startups are affected. That's a lot of risk to take on at billion-dollar valuations.
When Does the Music Stop?
I tried asking a VC friend about this, and their response was basically "yeah, some of these companies will fail spectacularly, but the winners will be so massive it doesn't matter." Which is classic VC logic, but it doesn't answer the more important question: what happens to all the talent and capital that gets tied up in the companies that don't make it?
We're watching thousands of smart engineers join AI startups at these inflated valuations. When (not if) some of these companies fail, that's a lot of disruption. And unlike the dot-com crash or the crypto collapse, AI actually works. The technology is real. But that doesn't mean every company building on it deserves a billion-dollar valuation.
The Talent Arms Race Nobody's Talking About
One thing that's not getting enough attention: these massive funding rounds are largely going toward hiring. Hippocratic AI wants to scale through acquisitions and team building. Mercor is growing its engineering team. Everyone's trying to hire the same pool of AI researchers and engineers, which is driving salaries to absurd levels.
I've heard stories of fresh PhD graduates getting $500k+ packages to join AI startups. Senior researchers are commanding million-dollar-plus comp. That's great for the individuals, but it's creating a talent bubble where companies are overpaying for people because they have to in order to compete. What happens when the funding slows down and these salary structures become unsustainable?
The Enterprise Reality Check
Here's what I think will separate winners from losers: actual enterprise adoption. It's one thing to raise money at a massive valuation. It's another thing to convince large, conservative enterprises to adopt your technology, pay for it, and integrate it into their workflows.
Healthcare organizations move slowly. Banks move slowly. Governments move glacially. The gap between "we have cool AI tech" and "we have thousands of paying enterprise customers" is huge, and most startups won't bridge it before they run out of runway.
Hippocratic AI seems to be doing well on this front—50+ healthcare partners is impressive. But how many of those are paying customers versus pilot programs? How sticky is the product? How defensible is the moat? These are the questions that matter when you're asking investors to value you at billions.
My Honest Prediction
I think we're in the middle of an AI funding bubble. Not because AI doesn't work—it does. Not because these companies aren't building real products—they are. But because the valuations are based on assumptions about future growth that won't pan out for most of these startups.
Five years from now, a handful of these companies will be worth more than their current valuations. OpenAI, Anthropic, maybe a few others. But the majority? They'll either fail, get acquired at down rounds, or struggle to justify their valuations with actual revenue and profit.
And that's okay. That's how innovation works. Lots of money flows into a new space, most companies fail, the winners emerge, and the market consolidates. But in the meantime, it's wild to watch billions of dollars get deployed based on PowerPoint decks and demo days.
The Bottom Line
Should you be excited about AI startup funding? Sure. It means innovation is happening fast, and some genuinely transformative companies are being built. Should you be skeptical of billion-dollar valuations for early-stage startups? Absolutely. History suggests most of these valuations won't hold.
If I had capital to deploy, I'd be very selective. I'd look for companies with real revenue, clear paths to profitability, and defensible moats—not just cool tech and ambitious growth projections. But that's the boring answer, and boring doesn't get you into the next OpenAI at Series A.
For now, I'm watching this unfold with equal parts fascination and concern. The AI funding boom is creating real companies and solving real problems. It's also creating unrealistic expectations and unsustainable burn rates. Which effect dominates? We'll know in a few years when the music stops and we see who's still standing.