An OpenAI co-founder is preparing for a dramatic comeback just days after being fired by the board. Photo: Justin Sullivan/Getty Images North America
Sam Altman has a habit of comparing artificial intelligence to nuclear weapons.
A few months after ChatGPT launched, OpenAI's co-founder signed a declaration saying governments should take AI as seriously as they would an atomic bomb. He has since led calls for the creation of a global body, similar to the International Atomic Energy Agency, to address the risks posed by the technology.
Altman is also prone to drawing personal comparisons to Robert Oppenheimer, the father of the atomic bomb. He is 38 years old, the same age Oppenheimer was when he was recruited into the Manhattan Project. Altman also quotes the scientist in an interview and notes that they have the same birthday.
After watching a biopic about the nuclear scientist this year, he was disappointed by its anti-nuclear message. «I was hoping Oppenheimer's film would inspire a generation of kids to become physicists,» Altman tweeted after seeing the film, «but it really missed the mark.»
For the four board members who fired OpenAI's CEO last week, comparisons between Altman and Oppenheimer may have been too close for comfort. Although the board was unable to publicly explain its actions, it is believed that their many grievances included concerns that Altman was paying too little attention to the dangers posed by AI.
Altman's resignation appears to have been an attempt to seize control of artificial intelligence and slow down the development of world-changing technologies while protective barriers are in place.
However, if this was the motivation, it backfired spectacularly. Altman returned to OpenAI on Tuesday after an uprising from investors and employees over his ouster, and the key movers behind his ouster have now been fired themselves.
Instead of wresting control from Altman, there was a clumsy coup. handed him even more power.
Dozens of employees streamed into OpenAI's San Francisco offices for an impromptu party Tuesday night after Altman was reinstated, joined virtually by hundreds more. OpenAI Chairman Greg Brockman, who was also fired and is now back, tweeted a midnight selfie surrounded by cheering employees.
we are back pic.twitter.com. com/YcKwkqdNs5
— Greg Brockman (@gdb) November 22, 2023
Microsoft chief executive Satya Nadella, who swung into action over the weekend to protect his $13bn (£10.4bn) investment in OpenAI, said that this news “inspired” him.
Chaos followed. Within days, Altman and Brockman were briefly hired by Microsoft, the former head of Twitch appointed a new chief executive of OpenAI, and more than 700 OpenAI employees threatened to quit.
The storm began Friday night when it was reported that Altman had been fired without any explanation. The shocking decision sparked anger among both employees and investors, including Microsoft, which worked behind the scenes to persuade the board to change course.
On Sunday, senior staff invited Altman back to OpenAI's offices for talks about his possible return, and at one point it appeared he was on the verge of being reappointed.
However, the board ultimately backed out and appointed Emmett Shear , co-founder of video streaming service Twitch, is the new head of OpenAI. . Microsoft immediately snapped up Altman and Brockman.
Instead of solving the problem, these steps provoked open rebellion on the part of the staff. A letter began to circulate calling on the board of directors to resign and Altman to return, eventually gaining the support of 95% of the staff. Even Ilya Sutskever, a member of the board of directors who is considered the initiator of Altman's dismissal, recanted his testimony and signed the letter.
Shear also questioned why exactly the board took action and promised to launch an investigation.
< img src="/wp-content/uploads/2023/11/feed64b81ae5aa710bf29531055c79f9.jpg" />Twitch co-founder Emmett Shear was appointed to replace Altman shortly after his ouster. Photo: Eoin Noonan/Web Summit/Getty Images Europe
By Tuesday, with the board's position increasingly untenable, the directors began negotiations with Altman about his return to OpenAI. to prevent the company from falling apart.
The entrepreneur may have returned to his job, but everything has not returned to normal. The events of the last few days mean that the balance of power within OpenAI has changed.
The momentum has gone against those who call for slower progress in the name of security. Instead, ChatGPT's development may now well be leveraged by those looking to cash in on innovation.
OpenAI is Silicon Valley's hottest startup, with an estimated valuation of $86 billion. ChatGPT is the most exciting new product in over a decade.
However, unlike past successes like Facebook or Google, the company has a unique corporate structure.
OpenAI is structured as a non-profit parent organization that owns a separate for-profit subsidiary that hires staff and raises funds. The for-profit subsidiary is controlled by the nonprofit's board, which is not accountable to shareholders but instead has a mission to «benefit all humanity.»
It is this structure that has led to clashes at the top, disagreements over what exactly it means to «benefit» benefit all humanity.»
Artificial intelligence researchers strongly disagree on whether AI is like the printing press, a revolutionary invention that will benefit humanity and create wealth, or like the atomic bomb, a threat to civilization.
They are divided into philosophical camps with different jargon incomprehensible to outsiders: decels, EAs, e/accs, doomers, x-risk deniers, and so on.
Two of the people who left OpenAI—Helen Toner, an academic at Georgetown University, and Tasha McCauley, a robotics entrepreneur and scientist at the RAND Corporation think tank—are associated with effective altruism, a movement deeply rooted in concerns about the threat of AI to humanity. Sutskever, who personally informed Altman of his resignation last Friday, has also become increasingly concerned about the risks posed by AI. His status at OpenAI was unclear Wednesday, but he was not named on the new board of directors.
Ilya Sutskever, who personally informed Altman that he was being fired, has become increasingly concerned about the risks associated with AI. Photo: JACK GOOSE/AFP
Of the four directors who voted to oust the OpenAI founder just days earlier, only one — tech entrepreneur Adam D. «Angelo — will remain on the new board.»
Those who departed were replaced by Larry Summers, the former US Treasury secretary, and Bret Taylor, a well-connected Silicon Valley executive who led Twitter through its sale to Elon Musk as the company's chairman. Both appear to be focusing more on AI's potential to impact business and employment than on its risks to humanity.
Altman is also expected to return to the board of directors, although he has not been reinstated at this time.
Altman is also expected to return to the board, although he has not been reinstated at this time.
p>
One well-connected technology investor says Altman's return is unlikely to mean risk concerns will be put aside. “It has been painted as an anti-security acceleration, but it is more of a security-style civil war. Sam is genuinely concerned about the safety of AI, but he has a very different approach.”
However, the company almost certainly has a more commercial bent.
The board will eventually increase to nine people, with Microsoft lobbying for a seat, according to tech news site The Verge. OpenAI's nonprofit future may depend on an independent investigation into the events of the past few days.
While much remains to be seen, there was widespread speculation Wednesday morning that Altman was stronger than a week ago. One entrepreneur who dealt with Altman even suggests that the coup worked so well for him that it could have been faked.
“This whole story could have been planned by Sam himself. He recently got bolder, there's a new board of directors that you assume he chose personally, and no one else appointed them, he has all the power, all the authority, [and] Microsoft is behind him.»
Altman's return doesn't undo all the damage caused by the failed coup.
Connor Leahy, chief executive of security research firm Conjecture, says the saga shows that companies like OpenAI can't be trusted with their own technology and they need stricter regulation.
«This is the clearest demonstration yet that even the most ambitious and well-intentioned 'self-government' is woefully inadequate [compared to] properly governing industries and people with such enormous power,» he says.
Some ChatGPT clients have also remained unhappy.
“OpenAI is in free fall until the governance problem is solved,” says Barney Hussey-Yeo, chief executive of AI budgeting app Cleo and an OpenAI customer.
“We're spending hundreds of thousands a month from [AI] vendors and shifting volumes to [competitor] Anthropic until there's clarity. It is unacceptable to have a CEO who is not on the board of directors and is in open war with another board member.”
Altman must prove to investors and customers that OpenAI is not the circus it has been for the last few years . days. However, after a tumultuous week, this may seem like a relatively simple task compared to what Oppenheimer just went through in the field of artificial intelligence.
Свежие комментарии