Improbably, Microsoft emerged from the high speed train wreck at OpenAI in a much stronger position. The CEO they backed has returned. The Board has been reconstituted with adults. And Microsoft’s influence is strengthened.
But Microsoft dodged a $13 billion cannonball. It could have been much, much worse.
They could have been locked in a protracted fight to hire away OpenAI’s employees despite non-competes. They could have been tangled up in litigation about whether Microsoft had the right to use IP developed by OpenAI. They could have derailed the monstrous rally in Microsoft’s stock since their $10 billion investment in January.
(Apparently Wall Street thinks the relationship is a big deal or something: Microsoft’s market cap is up more than a trillion dollars since the January OpenAI investment.)
We noted earlier this year that Microsoft violated every convention of VC deal-making by investing without getting board representation or even guaranteed equity.
Microsoft survived that structure. Can you? Probably not.
Three takeaways for investors:
- Structure matters. Does anyone alive understand this corporate structure? I don’t, and I went to a fancy law school.
- But the consequence is that:
- The Board has sole control of the OpenAI non-profit entity;
- This non-profit entity has sole control of the for-profit entity;
- The for-profit entity has a minority “owner” that invested $13 billion but received no Board representation – in a company whose profits are capped; and
- The “owner” has to give back the entirety of their equity investment upon realizing a specified return.
- Our sage counsel: don’t do that.
- Board composition matters. OpenAI had the wrong people on the boat. And the boat was too small.
- Incentives matter. Among a long list of head-spinning developments, the most startling indication that the Board was dreadfully misaligned with its largest investor? The Board told the leadership team that allowing the company to be destroyed would be consistent with the mission.”
A capped profit structure could actually make sense, provided you’re comfortable with the level of upside.
But to stroke a $13 billion check and have no formal governance at all? That’s why OpenAI was able to fire Altman with zero input from Microsoft. (They did have the courtesy to give Satya Nadella a heads up one minute before the announcement.)
The board was a mere six members. And only four were involved in the firing, including chief scientist Ilya Sutskever who fired Altman last Friday, and then signed a letter on Monday saying Altman should um, not have been fired. The remaining three Board members include the CEO of a 12 person tech company, and a Georgetown academic. The employees’ open letter indicated that the Board “did not have the competence to oversee OpenAl.”
They’re right.
Best practices for Board composition including having diverse viewpoints, a critical mass of size, and critically, the appropriate skill sets to oversee the corporation.
The OpenAI had none of those. Our sage counsel again: don’t that, either.
Whose mission?
The OpenAI Board’s job was not designed to protect or represent shareholders. It was designed to prevent an AI apocalypse. We’re in favor of preventing a Skynet apocalypse.
But if you invest in a for-profit entity run by a non-profit Board which sees its mission as protecting humanity, not maximizing shareholder value, don’t be surprised if they’re sanguine about you flushing $13 billion.
That’s why well-constituted Boards have incentives that are well-aligned with their investors. They have equity. They have alignment on the corporation’s ultimate purpose. They want the same things. This Board, and this investor, didn’t.
Microsoft dodged this multibillion dollar cannonball. You’re unlikely to be so lucky if your structure, your Board and your incentives are a similar hot mess.
