OpenAI CEO Sam Altman and his husband Oliver Mulherin have just joined the likes of Bill Gates and Warren Buffett in pledging to give away half their fortune, which is nice, because Altman’s reputation is otherwise taking a real battering right now.
Remember the chaos last November when OpenAI’s board abruptly fired Altman, and he got hired by Microsoft for the blink of an eye before returning to OpenAI, as the board members behind his brief defenestration had to leave the building instead? At the time, the lack of clear reasoning behind the ouster—Altman had been “not consistently candid in his communications with the board”—painted a picture where Altman has been wronged by a bunch of clumsy amateurs for no good reason.
But now we know a lot more about what went down, thanks to ex-board member Helen Toner and her appearance on the Ted AI Show podcast yesterday. Here are Toner’s most eye-catching allegations:
—OpenAI announced ChatGPT in 2022 without giving its own board any warning.
—Altman didn’t tell the board he owned OpenAI’s Startup Fund (over which he relinquished control last month,) while claiming to have no financial interest in OpenAI.
—“On multiple occasions, he gave us inaccurate information about the formal safety processes that the company did have in place, meaning that it was basically impossible for the board to known how well those safety processes were working or what might need to change.”
That is, indeed, a lack of consistent candidness. Or, put another way, it seems that Altman did his damnedest to stop OpenAI’s board from exercising its mission of ensuring safe AI development. If Toner’s allegations are true, then I’m not at all surprised that the board fired Altman.
OpenAI said in a statement that it was “disappointed that Miss Toner continues to revisit these issues,” and pointed out that the company’s post-fiasco independent review “concluded that the prior board's decision was not based on concerns regarding product safety or security, the pace of development, OpenAI's finances, or its statements to investors, customers, or business partners.”
Altman certainly emerged from the November episode as the winner, with his enemies vanquished, and there is no question that the man knows how to drive business. Just today, OpenAI scored consultancy giant PwC as both ChatGPT’s biggest customer, with 100,000 users, and its first enterprise reseller.
But the aphorism that someone is “only as good as their word” is still something that people take seriously, and for good reason. Coming so hot on the heels of the Scarlett Johansson AI voice drama, the exodus of OpenAI employees who say the company isn’t taking safety seriously enough, and the revelation that OpenAI never followed through on its promise to give proper resources to a key safety team, Toner’s allegations provide yet another reason to view Altman as fundamentally untrustworthy.
As my colleague Jeremy Kahn noted in yesterday’s Eye on AI newsletter, Altman’s lack of probity and the inability of OpenAI’s board to properly oversee him have (together with other things) major implications for AI regulation: “We desperately need a regulator with enough expertise and authority to look over the shoulder of these companies and ensure they aren’t building systems that pose extreme risks.”
Jeremy also recently pointed out that Microsoft CEO Satya Nadella, whose company continues to be OpenAI’s biggest partner, can’t be happy with the amount of ethical unease surrounding Altman and his operation. After what Toner just revealed about how Altman deals with his supposed overseers, I’ll bet Nadella is thanking his lucky stars that he’s not Altman’s boss.
More news below.
David Meyer
Want to send thoughts or suggestions to Data Sheet? Drop a line here.
