Sam Altman was stepped down as CEO of ChatGPT maker OpenAI over fears he was ignoring the dangers of artificial intelligence, a new report has found.
The concerns, raised by a fellow OpenAI board member, culminated in the 38-year-old’s sudden resignation Friday, which sent the tech world into a frenzy and set off a firestorm of speculation.
OpenAI, founded by Altman and ten others in 2015, rolled out ChatGPT a year ago — and its ability to mimic human writing has since proven polarizing.
In addition to raising concerns among the public, this uncertainty was apparently also felt within OpenAI’s six-member board, according to The New York Times report – in particular by one of Altman’s co-founders, Ilya Sutskever.
In a series of interviews, people familiar with the matter described how the AI researcher had concerns about OpenAI’s nascent technology before the firing, and shared a belief that his boss was not paying enough attention to the risks.
These concerns came as many warned that AI could one day result in a dystopian machine-controlled landscape, as was the case in James Cameron’s famous Terminator series.
Sam Altman, 38, was stepped down as CEO of ChatGPT maker OpenAI on Friday over fears he was ignoring the dangers of artificial intelligence, according to a new report
In a series of interviews, insiders described how one of Altman’s co-founders was concerned about OpenAI’s nascent technology before the firing. These concerns came as experts warned that AI could result in a machine-run society, as was the case in James Cameron’s Terminator series
A dramatic statement last May warned of such a reality, saying “bad actors” will use the burgeoning technologies to harm others before potentially ushering in an apocalypse.
The statement, signed by experts from companies like Google DeepMind and vets like Google alum Geoffrey Hinton, who resigned over the growing dangers of AI, was also pushed by Altman but was ultimately ignored, the Times said.
According to their report, prominent computer scientist Sutskever, 37, led the charge only to have him withdraw, successfully getting three other board members to do so.
Those executives, according to the Times, were Adam D’Angelo, CEO of the question-and-answer site Quora; Tasha McCauley of Rand Corporation and Helen Toner, the Georgetown University Center for Security and Emerging Technology.
McCauley, a scientist at the Santa Monica Think Tank, and Toner, director of strategy and basic research grants at the university, both have ties to movements that have repeatedly expressed the belief that AI could one day destroy humanity.
Sutskever, who was born in the Soviet Union, had become increasingly aligned with these beliefs over the past year, three insiders said, and reportedly came to them with his concerns.
The Times said neither Toner nor McCauley responded Saturday to requests for comment about the claims.
However, both women, along with D’Angelo, would vote to oust Altman — much to the chagrin of fellow board members Jakob Pachocki and Greg Brockman, who both resigned late Friday night in protest of the alleged ambush.
OpenAI, founded by Altman and ten others in 2015, rolled out ChatGPT a year ago — and its ability to mimic human writing has since proven polarizing
Aside from raising public concerns, this uncertainty was also apparently felt within OpenAI’s six-member board, according to The New York Times report — specifically by one of Altman’s co-founders, Ilya Sutskever.
According to the report, Sutskever led the charge to withdraw Altman – after successfully persuading fellow board members Tasha McCauley (left) and Helen Toner (right)
Both women have ties to movements that have repeatedly expressed the belief that AI could one day destroy humanity and create a society effectively run by machines.
Both women moved to oust Altman — much to the chagrin of fellow board members Jakob Pachocki (left) and Greg Brockman, who both resigned in protest of the alleged ambush late Friday night.
The Times report also sheds some light on the still-shrouded meeting between the board members, in which Altman — hours after making a public appearance on behalf of his company at Thursday’s APEC summit with Chinese President Xi Jinping — participated via video.
There, an insider said, Sutskever read from a script that closely resembled a message the now $90 million company wrote after the election — a message that blamed their boss’s sudden firing on his lack of transparency with the rest of the board.
That blog post, while vague, alleged that Altman “was not consistently forthcoming in his communications with the board, hindering the board’s ability to carry out its responsibilities.”
Neither Sutskever — a University of Toronto student who helped create a breakthrough in AI called neural networks — nor Altman immediately responded to the Times’ requests for comment.
Two insiders who spoke to the paper added that Sutskever, in his bid to force his co-founder out of the company, also objected to what he saw as a diminished role within the company amid its somewhat sudden success.
In a series of tweets late Friday and early Saturday, both Altman and Brockman — another co-founder and the company’s now former president — commented on the ouster, with the former speaking about it somewhat cryptically.
He wrote in two separate messages: “I love you all. (T)oday was a strange experience in many ways. (B)ut one unexpected thing is that it’s a bit like reading your own eulogy while you’re still alive.
The fire appeared to catch Altman off guard, who did not elaborate on what might have led to his departure
“The outpouring of love is amazing. One message: go tell your friends how much you love them.’
In the other, written a few hours earlier and shortly after he was fired, Altman added: I enjoyed my time at Openai. (I)t was transformative for me personally, and hopefully the world a little bit. What I enjoyed most was working with such talented people.
“I’ll have more to say about the future later.”
Brockman, meanwhile, said in his own X-post that while he was the chairman, he was not part of the board meeting where Altman was ousted. He explained how he and others were as surprised by the maneuver as onlookers when he stepped down to express his disgust.
OpenAI, meanwhile, released ChatGPT last November and the technology has taken off since then.
While many remain skeptical – or downright outraged – about the new technology, OpenAI executives expressed confidence in its future earlier this year, saying they both “believe in the responsible creation and use of these AI systems.” .
Brockman, the ex-president of OpenAi, said in his own post that while he was the chairman, he was not part of the meeting where Altman was ousted. He explained how he and others were as surprised by the maneuver as onlookers when he stepped down to express his disgust.
More recently, Microsoft provided OpenAI Global LLC with a $10 billion investment, on top of another $1 billion distributed to the company in 2019.
Employees last month tried to sell some of their stock at a $90 billion valuation, after valuing it at about $29 billion in a stock sale a few months ago.
Forrester analyst Rowan Curran speculated that Altman’s departure, while sudden, likely did not reflect deeper business problems.
“This appears to be a case of a management transition that was about issues with the individual in question, rather than the underlying technology or business,” Curran said.