Fired CEO Sam Altman will return to run OpenAI - the company he co-founded, following days of speculation and turmoil at the leading generative artificial intelligence start-up.
In a dramatic reversal, Altman, who was fired by OpenAI’s board of directors last week, will be reinstated under the supervision of a new board.
Greg Brockman, the co-founder and president who quit the company on Friday after Altman was fired, will return alongside him.
Under an “agreement in principle”, Altman will serve under the supervision of a new board of directors.
Patrick's Books:
Statistics For The Trading Floor: https://amzn.to/3eerLA0
Derivatives For The Trading Floor: https://amzn.to/3cjsyPF
Corporate Finance: https://amzn.to/3fn3rvC
Patreon Page: https://www.patreon.com/PatrickBoyleOnFinance
Buy Me a Coffee: https://buymeacoffee.com/patrickboyle
Visit our website: www.onfinance.org
Follow Patrick on Twitter Here: https://twitter.com/PatrickEBoyle
Fired CEO Sam Altman will return to run OpenAI - the company he co-founded, following days of speculation and turmoil at the leading generative artificial intelligence start-up.
In a dramatic reversal, Altman, who was fired by OpenAI’s board of directors last week, will be reinstated under the supervision of a new board.
Greg Brockman, the co-founder and president who quit the company on Friday after Altman was fired, will return alongside him.
Under an “agreement in principle”, Altman will serve under the supervision of a new board of directors.
Patrick's Books:
Statistics For The Trading Floor: https://amzn.to/3eerLA0
Derivatives For The Trading Floor: https://amzn.to/3cjsyPF
Corporate Finance: https://amzn.to/3fn3rvC
Patreon Page: https://www.patreon.com/PatrickBoyleOnFinance
Buy Me a Coffee: https://buymeacoffee.com/patrickboyle
Visit our website: www.onfinance.org
Follow Patrick on Twitter Here: https://twitter.com/PatrickEBoyle
Sam Altman – The CEO of Open AI - told the Wall Street Journal last year that “AI means that a lot of people are going to lose their jobs”, so he was probably not awfully surprised to lose his job as the CEO of Open AI last week. There was probably no one better prepared for such an outcome. I imagine that he kept his personal items in a box next to his desk at work, so that nothing would be accidentally left behind in the office when he eventually got the call from HR, and that he fled immediately to his bunker in New Zealand – probably expecting all of the other Silicon Valley CEO’s to be waiting there for him.
It turns out that all of the drama at Open AI was a storm in a teapot. Yesterday morning we learned that in a dramatic reversal, Altman, will be reinstated under the supervision of a new board, that contains only one member from the prior board and OpenAI’s largest investor, Microsoft, is expected to have a larger voice in OpenAI’s governance going forward.
A lot of the goings on at Open AI relate to its strange corporate structure, which relate to how it was initially funded. There are lessons that can be drawn from the chaos at Open AI that apply to the recent trend we have seen towards “stakeholder capitalism.”
Open AI was founded in 2015 by a group of high-profile entrepreneurs and researchers – among them were Sam Altman, Reid Hoffman, Elon Musk, Peter Thiel, Amazon Web Services and a few others. Its stated mission was “to advance artificial intelligence in a way that would benefit society as a whole, unconstrained by a need to generate a financial return.” They announced that they would "freely collaborate" with other institutions and researchers by making their patents and research open to the public. The founders pledged over one billion dollars to the venture, but actually only contributed around $130 million dollars- the majority of which came from Elon Musk.
Open AI were able to hire some of the top researchers in the space early on, despite paying a lot less than Facebook and Google partially because researchers said that they were excited by the opportunity of working with some of the top people in the space and because they believed in the mission of the company.
After three years, Elon Musk left his board seat claiming a conflict of interest with Tesla's AI R&D endeavors relating to autonomous driving. Insiders say that Musk pitched that he should be running the company and when Altman and OpenAI’s other founders rejected this proposal, Musk walked away — reneging on a massive, planned donation which left the nonprofit with no ability to pay the astronomical fees associated with training AI models on supercomputers.
Unable to balance its research goals with its budget constraints, Open AI made the announcement of creating OpenAI LP, a separate legal entity that operates as a "capped profit" corporation in 2019. The profits being capped at 100 times any investment. The company explained this decision saying, “We need to invest billions of dollars in the coming years into large-scale cloud compute, attracting and retaining talented people, and building AI supercomputers.”
This transition from nonprofit to for-profit required OpenAI to balance its desire to make money with its stated commitment to ethical AI development. They said that the cap on profits was there to prevent investors from being enticed to try and grab limitless wealth by deploying software that could be very harmful. I’ll let you decide whether that was a real concern they had, or a good way of marketing an investment in the firm.
This unconventional structure meant that Open AI had a board of directors which in theory controls the entire corporate structure (which includes the charity and the capped profit company) – but which unlike other boards is not accountable to shareholders. The directors are in fact not allowed to own any stock to prevent a conflict of interest, because they are specifically not supposed to be aligned with shareholders. The Nonprofit’s principal beneficiary is “humanity”, not OpenAI investors, according to their website – which may mean that they are accountable to everyone – including you – or it may mean that they are accountable to no-one.
The companies operating agreement – to investors – says – in writing… “It would be wise to view any investment in OpenAI in the spirit of a donation, with the understanding that it may be difficult to know what role money will play in a post-AGI world.” Documents like this – that were written by an actual lawyer - highlight the problems we are starting to see from the combined popularity of science fiction in Silicon Valley and widespread microdosing of hallucinogens.
OK, so, while the Open AI investor documents write about the unknowable role that money will play in a post AGI world… we haven’t gotten there yet, and in the real world, where the role of money is reasonably well defined, Open AI is an unprofitable company and is expected to need to raise a lot more money over time from investors like Microsoft, to keep up with the high costs of building more sophisticated chatbots.
Despite this lack of profitability, the company is valued by investors at 86 billion dollars, and Bloomberg reported last weekend that “some investors were considering writing down the entire value of their OpenAI holdings to zero.” Eighty-six billion dollars of value evaporated in a weekend and the reason given by the bord was that Altman had not been consistently candid with them.
With billions of dollars in jeopardy, some investors started exploring legal action, but the problem with that is that it is not clear how to sue a board whose legal duty was to the safety of humanity, not to investors.
Microsoft may have been willing to sign the wacky documents like they did and pretend that the board which answers only to “humanity” were in charge, but because Open AI requires investors to write large checks to fund the ongoing losses at Open AI, Microsoft were easily able to exert control – which is what they did. They pushed for Altman to be reinstated while also offering him a role at Microsoft.
Altman quickly confirmed that he and Greg Brockman, another co-founder who left on Friday, would join Microsoft, OpenAI’s key partner. Former colleagues would have an open door to follow and join a new AI unit, according to Microsoft chief Satya Nadella.
As much of a win as this might have appeared for Microsoft – people were saying that they had managed to buy the hottest AI firm for zero. This might not have been the optimal outcome for them as they would likely have had to deal with antitrust regulators and lawsuits from other Open AI investors.
Now, the board doesn’t answer to the staff at Open AI either, (other than that the staff make up some of humanity) - but the staff do all of the work, while the board attend occasional meetings where they worry about the future of humanity. Not a lot would be going on at Open AI without employees.
The majority of Open AI’s 700 or so employees signed an open letter to the board demanding that the board resign and that they rehire Altman. The letter stated that the board had told the employee leadership team that allowing the company to be destroyed “would be consistent with the mission.” The employees said that unless their demands were met, they would resign from Open AI and join the new subsidiary of Microsoft being headed up by Altman and Brockman.
As a side note – to all of this… You have to wonder what the employee contracts at Open AI look like that the entire staff could leave to work for a major investor in the company leaving Open Ai as an empty shell. Typically, executives like Altman would have contracts that prevent them from hiring away key staff once they are no longer at the firm, and staff would have signed NDA’s preventing them from taking any technology with them.
The press are reporting that this mass revolt by OpenAI staff was the primary reason for the board’s about face. So, while the board (on paper) only report to humanity, investors and employees won out.
The open AI story is a bit of a crazy one where Microsoft and a number of other sophisticated investors agreed to put billions of dollars in, and employees got stock grants all at an 86-billion-dollar valuation without the contractual or fiduciary rights that investors might normally expect.
Open AI is not the only company with this type of structure either. Anthropic, a rival company started by a group of former OpenAI employees, has set itself up as a public benefit corporation, a legal structure that is meant to insulate it from market pressures.
The Open AI case highlights a trend that we have seen in recent years in favor of unusual corporate governance. I wrote about this issue in my corporate finance textbook as while there has been a general move towards better and better corporate governance over time – bad corporate governance has been a growing issue particularly in Silicon Valley where companies like Google, Facebook and Snap structured their IPO’s such that founders were left with unchallenged power to do almost anything that they want.
At these firms’ founders were given shares with special voting powers meaning that they didn’t have to listen to investors at all. Investors looked at these companies and felt that the founder/CEO had such a great track record that there was little consequence to having poor corporate governance.
Berkshire Hatheway – Warren Buffett’s company has a similar unusual structure, where the A shares get more votes than the B shares do. The different voting rights were allegedly put in place so that the company and its leadership would not be beholden to investors seeking short-term profit.
This structure at Berkshire Hatheway has kept Warren Buffett in firm control even as he has given away more and more of his shares to charity. The A shares can be converted to B shares, but B shares can’t be converted to A shares. Starting in 2006, Buffett began giving away large amounts of Berkshire Hathaway stock to charities, but prior to donating them, he converts his A shares to B shares. Since Class B shares have diminished voting rights, this has the effect of preserving his voting control even as his percentage ownership of Berkshire declined over the years.
While investors might feel that there is little consequence to having poor corporate governance this could be short sighted. Since you are buying shares of companies in perpetuity, leadership who are not accountable to shareholders can take value destructive paths without answering to anyone.
Meta’s Reality Labs division, which houses its efforts to build the metaverse, has lost around $46.5 billion dollars since 2019. Would Mark Zuckerberg have been able to waste this much money if he was accountable to investors? Under a normal corporate governance structure there is a good chance that he would have been pushed out by shareholders a few years ago as the losses were mounting.
There have been numerous studies on the relationship between corporate governance and investment performance. Gompers et al found that companies with stronger shareholder rights have higher profits and trade at higher multiples than companies with weak governance.
A few years ago, the Business Roundtable, a nonprofit lobbyist association whose members are CEO’s of major US companies, put out a press release saying that corporations should be run to protect all corporate stakeholders, defined to include customers, society and employees, rather than following the conventional objective of maximizing shareholder wealth. This is what has become known as “Stakeholder Capitalism.”
Critics of shareholder wealth maximization celebrated this announcement as an acceptance of their long-term belief that focusing on shareholder wealth had brought about income inequality, the outsourcing of manufacturing jobs and other societal costs and externalities. Traditionalists viewed this announcement as foolish and caving in to the woke mob. Cynics argued that it meant nothing, was just PR and that everything would continue on as normal.
To most people, these arguments might appear to not matter whatsoever, but they do matter because the various stakeholders in a company have very different interests. Actions that benefit one stakeholder group can be reasonably expected to make other stakeholder groups worse-off. These conflicts of interest mean that a board that is supposed to represent everyone often represents no one other than themselves and their own specific objectives.
Aswath Damodaran wrote a great blog piece on this topic a few years ago. He looked at the different models of corporatism: Cutthroat corporatism where the end game is stockholder wealth maximization at almost any cost; Crony Corporatism, where companies focus less on efficient operations than on connections to government with the objective being tilting the scales of competition in the company’s favor; Managerial Corporatism where the interests of executives drive the direction of a company and investors and other stakeholders are ignored and finally Constrained Corporatism where companies preserve the primacy of shareholders, while constraining how they interact with other stakeholder groups. Constrained Corporatism is what we mostly see where companies are restricted in their actions by the legal & regulatory constraints, self-imposed constraints and market-driven constraints
Damodaran describes “Stakeholder Capitalism” as confused corporatism which on the surface might look like constrained capitalism, but due to the multiple objectives, with no clear ranking of importance – leaves you with a structure which is destined to fail. The events at Open AI over the last week really highlight this issue.
OpenAI’s new board will consist of three people, at least initially: Adam D’Angelo, the chief executive of Quora (and the only holdover from the old board); Bret Taylor, a former executive at Facebook and Salesforce; and Larry Summers, the former Treasury secretary. The board is expected to grow from there.
OpenAI’s largest investor, Microsoft, is also expected to have a larger voice in OpenAI’s governance going forward. This may include a board seat.
If we want companies to behave better in their interactions with society, putting boards in place that are not accountable to anyone other than a vague concept of “the greater good” is probably not the best approach. The Volkswagen emissions scandal from 2015 highlights that bad corporate behavior can harm shareholder value. Volkswagen was hit with huge fines, recall costs and a damaged brand because of their illegal behavior. A board focused on shareholder value is able to make decisions that are good for society as a whole, especially when constrained by the legal system and a reasonable code of ethics. It may be far from perfect, but it does appear to be better than the alternatives.
Thanks for tuning in to this week’s podcast. If you have a friend who might find these interesting, I’d appreciate if you could send them a link to help the podcast grow. Talk to you again soon, bye.