OpenAI
Debacle of the decade?
This is a fast moving story and it may well have changed by the time you read this post, so check the press for recent developments. But as I write on Tuesday afternoon, this is where things stand.
In what is being called “the debacle of the decade”, this past Friday four members of the governing board of OpenAI—the company which created ChatGPT and whose corporate motto is “creating safe AGI that benefits all of humanity”— unexpectedly fired the company’s popular CEO, Sam Altman, and removed the company’s President, Greg Brockman, as chairman of the board. On Sunday, Altman and Brockman both accepted jobs at Microsoft, OAI’s 49% shareholder and joint venture partner, which made clear that it was open to hiring additional OAI employees in the coming days. By Monday, OpenAI had hired a new CEO from outside the company, replacing the internal interim CEO named just two days earlier, in what must be one of the fastest CEO ‘searches’ in corporate history. And to top things off, Ilya Sutskever, OAI’s chief scientist and one of the four board members who voted to fire Altman, apparently changed his mind and along with hundreds of other OAI employees signed a public letter to the OAI board (on which he still sits) demanding Altman’s immediate return to the company.
Which of course is not how things are normally handled on well managed corporate boards.
But normal or not, what are we to make of the OpenAI debacle? Are there any lessons here for those of us interested in the study and practice of corporate finance and corporate governance? I have no doubt that there are indeed important lessons to be learned from the OpenAI fiasco, although I am not yet sure exactly what they are or how widely applicable they may be. The OpenAI story is after all a bit of a ‘man bites dog’ case, due to the company’s unorthodox funding and governance structure. The commercial entity we know as OpenAI is a for-profit corporation which has private shareholders who have invested billions of dollars of risk capital in the company in exchange for a capped amount of future profits. Microsoft alone has reportedly invested $13 billion and owns 49% of the shares in the for-profit company. Other private shareholders include the prominent VC firms Sequoia Capital, Andreesen Horowitz, Tiger Management, Thrive Capital and Khosla Ventures. Employees at OpenAI receive stock based compensation, and until last Friday were expecting to sell some of their shares to third party investors in a private placement at an implied valuation of $90 billion, three times the valuation of the previous funding round.
Despite this private funding structure, however, the for-profit OAI entity is legally owned and governed by a non-profit organization which by charter is devoted to the future development of AI technology for the benefit of humanity, rather than to maximizing future profits or economic value for itself or OAI’s private shareholders. Legally, OAI’s private shareholders have (limited) economic rights but apparently not much in the way of governance rights.
It remains unclear exactly why the OpenAI board fired Altman, which the board has not seen fit to explain publicly in any detail. In a statement issued on Friday, the OAI board said that “Mr. Altman’s departure follows a deliberative review process by the board, which concluded that he was not consistently candid in his communications with the board, hindering its ability to exercise its responsibilities. The board no longer has confidence in his ability to continue leading OpenAI.” The board subsequently issued a weekend communication to employees in which it stated that its decision to fire Altman was “not about product safety or security, the pace of development or OpenAI’s finances. This was not about any singular incident.” Once again, however, the board declined to explain with any specificity why it fired Altman, for reasons that remain unclear.
One of the primary responsibilities of any governing board—perhaps the primary responsibility— is that of hiring and firing the CEO. And when the board loses confidence in the CEO, for whatever reason, it is generally time for the CEO to go. At most business corporations, shareholders have the legal right to replace members of the governing board if they do not like the board’s decision(s) on this or any other matter. But this is not the case at OpenAI, where the company’s private shareholders have little if any legal authority to demand changes at the board level, no matter how much economic value they believe the board’s decisions may have destroyed. And so while OAI’s private shareholders are now in open rebellion, the company’s board is standing pat, for now at least, protected by its unique legal governance mandate and authority. As a result, OpenAI’s future as a company is now at risk and the company’s $90 billion valuation seems like a pipe dream.
If there is one immediate takeaway from the OpenAI saga, perhaps it is this: Strong corporate governance rights do not always matter much, but occasionally they matter greatly. This is a principal which even the most successful VC investors sometimes ignore, occasionally with disastrous consequences, as we have seen recently at Twitter, FTX and now OpenAI.
It may well be that the powerful potential of AI technology to revolutionize our world, for good and/or ill, presents unique challenges which demand some new and unusual corporate governance and funding structures, which may or may not look like those employed by OpenAI. The promotors of AI technology still have a lot of hard work to do on many fronts—technological, commercial and otherwise—but if they can’t come up with the right governance and funding structures, all of their hard work may prove to be for naught.
Which, come to think of it, may also be true for the shares of OpenAI.
Links
Altman is Out after OpenAI Board Skirmish, WSJ, November 17, 2023
OpenAI Announces Leadership Transition, November 17, 2023
Altman to Join Microsoft Following OpenAI Ouster, WSJ, November 20,2023
Ilya Sutskever, the OpenAI Genius Who Told Sam Altman He Was Fired, WSJ, November 21, 2023
Meet the Board of OpenAI Who Pushed Out Sam Altman, WSJ, November 21, 2023
OpenAI Investors Keep Pushing for Altman’s Return, WSJ, November 21, 2023
OpenAI’s Ineffective Altruists vs Sam Altman, WSJ Editorial, November 20, 2023
Open AI Seeks $90 Billion Valuation in Sale of Existing Shares, WSJ, September 26, 2023
Does AI have an Inherent Governance Problem? NY Times Dealbook, Nov. 21, 2023
The Long Shadow of Steve Jobs Looms Over the Turmoil at OpenAI, NY Times, November 21, 2023