The options Stanislas Polu left behind at OpenAI were worth more than all of Dust

Stanislas Polu, founder of Dust, has put a number on what it cost him to leave OpenAI. The stock options he walked away from were worth more than the total valuation of Dust at its own Series B, a round that had closed just six months before. That single comparison carries more weight than any general claim about how quickly AI lab valuations have moved.

The arithmetic deserves a moment. A Series B is not a seed note or an angel check. It is a round at which a company has already demonstrated enough to attract institutional capital at a defensible price. For the unvested options at a single employer to exceed that entire company valuation, the gap between what AI labs are worth and what a well-funded AI startup is worth has to be extraordinarily wide. Polu is not making an abstract point about the industry. He is describing a personal financial decision with a specific, concrete cost.

What makes the claim notable is its structure. It does not rest on a public filing, a leaked valuation memo, or an analyst estimate. It rests on Polu’s own knowledge of two numbers he is in a unique position to know: what his options at OpenAI were worth when he left, and what Dust was worth at its Series B. Both figures belong to him in a way they do not belong to anyone else commenting on the AI funding environment from the outside.

Both series B, but the uh stock option that I gave up were more than the whole company as of six months. Stanislas Polu

The implication Polu draws is about valuation growth at AI labs specifically. He is not saying the options were valuable in isolation. He is saying the relationship between the two numbers, set six months apart, tells a story about the speed and scale of that growth. A company valued at Series B is already a going concern with real capital behind it. The fact that one person’s unvested options at a lab could exceed that figure suggests the labs are operating in a valuation register that is categorically different from the broader startup ecosystem around them.

There is no claim here about whether that gap is sustainable, rational, or likely to persist. Polu is not offering a forecast. He is offering a data point: one that is precise, firsthand, and self-evidently verifiable to anyone who can see both cap tables. That is a different kind of evidence than a market observer’s characterization of the AI funding climate. It is a founder accounting, in concrete terms, for what he gave up and what that amount meant relative to something he built himself.

The candor is itself worth noting. Founders who leave well-compensated positions do not always quantify the cost in public. Polu’s willingness to frame his departure in terms of a specific financial comparison makes the claim more tractable than most of what gets said about AI lab compensation. It also puts a floor under the conversation. Whatever one believes about where AI lab valuations go from here, Polu’s account establishes that at the moment he left, the number was large enough to clear a Series B valuation with room to spare.