Sam Altman (left) and Ilya Sutskever (right) may have had differences on how to balance AI innovation with regulation.
JACK GUEZ
Fears over the speed of AI development at OpenAI may have sowed discord within its top ranks.Some OpenAI board members may have disagreed with Altman’s approach to AI innovation.OpenAI cofounder Ilya Sutskever has long called for caution due to concerns AI could harm humanity.
Sam Altman’s possibly brief dismissal from OpenAI has exposed a schism between the company’s top leaders over how to curb the existential threat posed by AI.
While Altman has spoken about the dangers of “potentially scary” AI and the “critical” need to regulate it, OpenAI’s former CEO has been chiefly a poster child for rapid AI innovation.
Altman was known to aggressively position OpenAI to stay ahead of the pack in the AI arms race, chasing large amounts of funding and rapid development.
In September, Altman tried to win $1 billion in funding from Softbank for a hardware device to run tools like ChatGPT, for instance. The Japanese conglomerate’s CEO, Mayoshi San, has previously said he uses ChatGPT every day and is a believer in the power of AI.
Ilya Sutskever, OpenAI’s cofounder, chief scientist, and a board member who played a role in Altman’s dismissal, meanwhile, preferred to tread more carefully given the potential threats AI poses to society.
Sutskever created a “Super Alignment” team within the company to ensure that future versions of GPT-4, the technology behind ChatGPT, wouldn’t be harmful to humanity, according to The New York Times.
Two other OpenAI board members — Helen Toner, who’s a director at Georgetown University’s Center for Security and Emerging Technology, and technology entrepreneur Tasha McCauley — also have ties to the so-called effective altruism movement, which works to ensure advances in artificial intelligence align with the best interest of humans.
If their concern over Altman’s commitment to the effective altruism movement was behind his ouster, it wouldn’t be the first time such disagreements over AI’s dangers have pushed people out of the company.
In 2021, Dario Amodei and several other OpenAI employees left the company to found Anthropic, an OpenAI rival that’s made building a safer AI central to its mission.
Even Elon Musk, who left OpenAI’s board in 2018 citing a conflict of interest with Tesla, was concerned with how much (or how little) the company prioritized safety, according to Wired.
OpenAI’s board has not offered any more details on what led them to fire Altman beyond noting that they had “lost confidence” in him and that he was not “consistently candid” in communications.
The fallout over the abrupt dismissal, however, has been swift. Multiple executives have quit in protest, and now many of OpenAI’s employees are calling for Altman’s return.