From Private to Public and Back Again

When VCs start acting like crossover funds

There’s an interesting inversion happening in venture right now. For years, public market investors tried to act more like VCs by chasing growth rounds, writing big checks into private company, calling themselves “crossover” funds. This made sense when venture capped out around Series D. But now it’s going the other way. The venture firms are the ones acting like crossover funds.

Thrive, Andreessen, General Catalyst, and others are all deploying billions into what are basically public securities that just haven’t listed yet, like OpenAI.

Meanwhile, the traditional crossover players – the Coatues and Tigers of the world – suddenly find themselves in a strange position. They were supposed to own this space, but the AI boom has made the public market possibly more exciting than the private market. NVIDIA and its peers have created venture-like returns without the illiquidity or the markup games. If you’re a public markets investor and your’e bullish on AI, you have innumerable opportunities in the public markets. NVIDIA is the obvious one, but there are plenty of “NVIDIAs” in terms of returns (though not scale) out there to choose from.

That said, FOMO doesn’t die easily. Even at half-trillion-dollar valuations, investors can’t help but look at OpenAI or Anthropic and think: what if this really is the next trillion-dollar company? The problem is, dilution and scale eventually catch up. A $500 billion company that needs to raise billions each year for compute is on a treadmill. For every “tripling” you dream about, you might need a quadrupling just to stand still.

Then there’s the risk calculus. On the private side, AI investing right now feels a lot like ad tech did in the 2010s: one decision from Google or Meta could vaporize your business overnight. The hyperscalers are building fast, deploying faster, and sitting on distribution moats no startup can match. If you’re funding an AI company today, you’re betting it can survive in a world where its most powerful suppliers could also become its biggest competitors.

So investors are forced into a weird choice. Do you back the giants like OpenAI and Anthropic and risk overpaying? Or do you back the upstarts, knowing most could be disintermediated before they ever reach scale? Both paths are risky; they just hurt in different ways.

At $500 billion, OpenAI is no longer an underdog. It’s being priced like a sovereign economy, not a startup. That doesn’t make it a bad company. It’s extraordinary, maybe once-in-a-generation; but even the best companies eventually have to reconcile with math. To justify that price to most investors, it would need to become a $2 trillion company. There aren’t many of those in the world, and every one of them owns global distribution and multiple revenue streams.

Venture capital used to be about taking a view on where the world is going and finding the next great company aligned with that vision. Now, at least for the mega/platform funds, it’s about deciding which already-great company might still have one more doubling left in it, and how much you’re willing to pay for that belief.