Bandaged Up OpenAI Faces a Tougher Task Now

“I have never been more excited about the future,” wrote Sam Altman, the reinstated chief executive officer of OpenAI, to his subordinates on Wednesday, in a statement formally announcing his return and the rebuilding of the company’s fractured board. “I am sure books are going to be written about this time period, and I hope the first thing they say is how amazing the entire team has been. Now that we’re through all of this, we didn’t lose a single employee.”

“We’re through all of this”? Hardly. It is, at best, only the end of the beginning. You can’t clear up an atomic blast with a dustpan and brush; not when it shook lead investor Microsoft Corp. and the wider AI community. The confirmation of Altman’s return, and the appointment of an interim board, still leaves many huge questions unanswered. Radical transparency is needed if we are to regain trust in both the company and its technology.

The bandaged up version of OpenAI looks like this: As expected, an interim, three-person, all-male board has been installed. Neither Altman, nor co-founder Greg Brockman, who resigned in solidarity, are on it — they will instead take the CEO and president roles, respectively. The chief technology officer will once again be Mira Murati, who you’ll remember was briefly the stand-in CEO when Altman was kicked out.

The three board members are former Salesforce Inc. co-CEO Bret Taylor, who will be chair; former US treasury secretary Larry Summers (a paid contributor to Bloomberg Television); and Quora CEO Adam D’Angelo. D’Angelo is the only board member to survive the cleanse, despite being one of those who voted for Altman’s ousting. Losing his seat, however, is OpenAI co-founder and chief scientist Ilya Sutskever, whose decision to join Altman’s opponents gave the board the fateful majority it needed.

Sutskever is remaining at the company, though his position is unclear. “I think he’s a guiding light of the field and a gem of a human being,” said Altman of Sutskever on Wednesday. “I harbor zero ill will towards him.”

Crucially, newly added to the board, in a non-voting role, is Microsoft. Given just a literal minute’s heads up of Altman’s firing, the company was suddenly an over-exposed outsider despite being OpenAI’s biggest investor. CEO Satya Nadella made it clear he wouldn’t let such a situation arise again, but a solution had to be found carefully. A board observer with no direct power is an attempt to thread the needle of keeping tabs on its AI Golden Goose, while staying distant enough so as to not attract added attention from antsy competition regulators. A Microsoft spokesman declined to share any more details on the precise nature of this observer’s access, or who specifically would take up the position.