A sudden and seismic leadership crisis at OpenAI has led to a revolving door of CEOs at the artificial intelligence company, with tech entrepreneur Emmett Shear becoming the latest to take the helm on Monday.
Just days earlier, OpenAI’s board abruptly ousted then-CEO Sam Altman, installing Chief Technology Officer Mira Murati as interim CEO. As the “Game of Thrones”-style drama played out over the weekend, questions swirled about Altman’s fate. But by Monday morning, Altman had accepted a job at Microsoft, the tech giant with a sizable investment in OpenAI; Shear was named interim CEO; and hundreds of OpenAI employees — including Murati — were calling for the board’s resignation and threatening to follow Altman to Microsoft.
The stunning events of the past 60 hours raise deep questions about the future of OpenAI, the company — with its unusual hybrid structure as a non-profit and for-profit — that first brought the world ChatGPT and jumpstarted a global debate about generative AI’s promise and peril.
Now, picking up the pieces at OpenAI will be Shear, the 40-year-old co-founder of the video-game live-streaming company Twitch. On Monday, Shear announced he had accepted the job of interim CEO because he believes OpenAI “is one of the most important companies currently in existence.”
Whether it can remain so depends on what Shear does next.
After leaving Twitch earlier this year to care for his newborn son, Shear takes the reins of a hollowed-out firm that has lost key co-founders, senior employees and is at risk of losing many more. He will have to deal with a potentially moribund board team that voted to trigger the crisis and that allegedly viewed the hypothetical collapse of OpenAI as a beneficial outcome that would serve the company’s own mission.
As he vows to investigate the events that led to Altman’s firing, Shear must not only refocus a shrunken team and salvage the company’s position as a leading AI developer — in an industry that changes radically with Altman’s departure — but also redefine what OpenAI stands for in a sprawling global debate about the risks and benefits of artificial intelligence and how to regulate it.
“When the board shared the situation and asked me to take the role, I did not make the decision lightly,” Shear said in a post on X, the platform formerly known as Twitter. “Ultimately I felt that I had a duty to help if I could.”
Despite being known for launching a social media company that was acquired by Amazon in 2014 for $970 million, Shear has become an increasingly vocal commentator on AI, and some of his prior writings and remarks offer a window into his management style and philosophy on artificial intelligence risks.
Shear’s first company, a calendar startup named Kiko, graduated from Y Combinator’s program in 2005, part of an initial batch of companies that also included Reddit and Loopt, a location-based social network founded by Altman.
The Yale-educated computer scientist, entrepreneur and investor describes himself as a habitual advisor. He’s spent years guiding fledgling tech companies as a part-time partner at Y Combinator, the startup accelerator once led by Altman. He has a penchant for dispensing nuggets of business wisdom on X, sprinkled throughout other posts about video games and science fiction books.
In a 2021 thread reflecting on the 10th anniversary of Twitch’s launch, Shear posted 23 tweets distilling what he’s learned into pithy management lessons such as that “for internet companies, growth is more important than profit.” He also wrote “there are only five growth strategies that exist, and your product probably only fits one.”
In more recent podcast appearances, Shear has combined his fondness for high-level, abstract thinking with a tendency toward colorful analogies to chess, Star Trek and early human evolution to articulate his views on AI — particularly on artificial general intelligence, an advanced technology that remains years away but that many researchers in the AI field believe could be the ultimate outcome of their work.
Shear has said he resembles many of his Silicon Valley peers in broadly favoring limited regulation of technology, or regulations that can better unlock the promises of innovation. But he has also argued that in the specific case of AI, future improvements in the technology are likely to happen so quickly, and eventually independent of any human intervention, that it could easily overwhelm its creators.
“You’ll be able to point the thing we’ve built back at itself … that loop will get tighter and tighter and tighter, and faster and faster and faster, until it can fully self-improve,” Shear said in June, outlining his concerns. “That kind of intelligence is just an intrinsically very dangerous thing, because intelligence is power.”
A native of Seattle, Washington, Shear was one of four co-founders of Justin.tv, an early precursor to Twitch that was built to livestream the life of Justin Kan, Shear’s childhood friend and classmate.
Though Kan’s experiment with live-streaming his own life to the internet was short-lived, Justin.tv expanded to allow virtually anyone to broadcast real-time video to the web. It organized its content by verticals such as sports and entertainment, but Shear and his co-founders soon saw immense potential in video game content.
“People hooked up their Xboxes and started broadcasting. We had never even thought of that,” Shear told Forbes in 2013.
In 2011, Justin.tv’s gaming vertical was spun off to become Twitch, the platform now owned by Amazon and favored by tens of millions to watch — and share — real-time video. Gaming is a central focus of the site, but it also promotes podcasts and has categories for art, cooking and other pursuits. Some politicians have used Twitch to connect with voters, particularly during the pandemic.
But in taking the reins of OpenAI, Shear is now helping to lead an industry with even more potential economic and political impact. And in September, Shear laid out his stance on a charged debate in AI circles: Whether AI development should “pause” due to its potentially existential risks to human society.
“The way you make it safely through a dangerous jungle at night is not to sprint forward at full speed, nor to refuse to proceed forward. You poke your way forward, carefully,” he said, adding: “If we’re at a speed of 10 right now, a pause is reducing to 0. I think we should aim for a 1-2 instead.”
Even if artificial intelligence does not become smarter than human intelligence, Shear has argued, it could still wreak havoc in the same way that people can.
“Imagine 100,000 of the smartest people you know, all running at 100x real-time speed, and able to communicate with each other instantaneously via, like, telepathy,” he said in September. “Those 100,000 people can credibly take over the world. They don’t have to be smarter than a human.”
Those views seem to dovetail with concerns about AI safety that reportedly may have been a factor in the OpenAI board’s firing of Altman, though Shear on Sunday disputed that Altman’s dismissal was “over any specific disagreement on safety” and that “their reasoning was completely different from that.”
Still, Shear’s perspectives set the stage for OpenAI to adopt a more cautious approach in its post-Altman future as Altman heads to Microsoft. And that raises its own set of questions about how Shear may manage OpenAI’s relationship with Microsoft. Each has reiterated its commitment to the other, as part of a deal that has seen OpenAI’s technology baked into the Bing search engine and Microsoft investing billions into OpenAI.
But with Altman and his allies working in-house at Microsoft, Shear — and whomever succeeds him as OpenAI’s permanent CEO — may forever be overshadowed.