OpenAI, are you okay, babe? This past weekend has been a fever dream within the AI world. The board of OpenAI, the world’s hottest AI company, shocked everyone by firing CEO Sam Altman. Cue an AI-safety coup, chaos, and a brand new job at Microsoft for Altman.
When you were offline this weekend, my colleague Will Douglas Heaven and I break down what you missed and what’s next for the AI industry.
What happened
Sam Altman was summoned to a Google Meet meeting, where chief scientific officer Ilya Sutskever announced that OpenAI’s board had decided Altman had been “not consistently candid in his communications” with them, and he was fired. OpenAI president and cofounder Greg Brockman and a string of senior researchers quit soon after, and CTO Mira Murati became the interim CEO.
Murati made attempts to rent Altman and Brockman back, while the board was concurrently searching for its own successor to Altman. Altman and OpenAI staffers pressured the board to quit and demanded that Altman be reinstated, giving the board a deadline, which was not met.
Microsoft announced it had hired Altman and Brockman to guide its recent AI research team. Soon after that, OpenAI announced it had hired Emmett Shear, the previous CEO of the streaming company Twitch, as its CEO.
Over 500 OpenAI employees have signed a letter threatening to quit and join Altman at Microsoft unless OpenAI’s board steps down. Bizarrely, Sutskever also signed the letter, and posted on X that he “deeply regrets” participating within the board’s actions.
What’s next for OpenAI
Two weeks ago, at OpenAI’s first DevDay, Altman interrupted his presentation of an AI cornucopia to ask the whooping audience to calm down. “There’s lots—you don’t should clap every time,” he said, grinning wide.
OpenAI is now a really different company from the one we saw at DevDay. With Altman and Brockman gone, numerous senior OpenAI employees selected to resign in support. Many others, including Murati, soon took to social media to post “OpenAI is nothing without its people.” Especially given the specter of a mass exodus to Microsoft, expect more upheaval before things settle.
Tension between Sutskever and Altman could have been brewing for a while. “When you could have a corporation like OpenAI that’s moving at a quick pace and pursuing ambitious goals, tension is inevitable,” Sutskever told MIT Technology Review in September (comments which have not previously been published). “I view any tension between product and research as a catalyst for advancing us, because I consider that product wins are intertwined with research success.” Yet it’s now clear that Sutskever disagreed with OpenAI leadership about how product wins and research success ought to be balanced.
Recent interim CEO Shear, who cofounded Twitch, appears to be a world away from Altman in terms of the pace of AI development. “I specifically say I’m in favor of slowing down, which is type of like pausing except it’s slowing down,” he posted on X in September. “If we’re at a speed of 10 straight away, a pause is reducing to 0. I believe we should always aim for a 1-2 as a substitute.”
It’s possible that an OpenAI led by Shear will double down on its original lofty mission to construct (in Sutskever’s words) “AGI that advantages humanity,” whatever which means in practice. Within the short term, OpenAI may decelerate and even switch off its product pipeline.
This tension between attempting to launch products quickly and slowing down development to make sure they’re secure has vexed OpenAI from the very starting. It was the rationale key players in the corporate decided to go away OpenAI and begin the competing AI safety startup Anthropic.
With Altman and his camp gone, the firm could pivot more toward Sutskever’s work on what he calls superalignment, a research project that goals to give you ways to regulate a hypothetical superintelligence (future technology that Sutskever speculates will outmatch humans in almost every way). “I’m doing it for my very own self-interest,” Sutskever told us. “It’s obviously essential that any superintelligence anyone builds doesn’t go rogue. Obviously.”
Shear’s public comments make him precisely the form of cautious leader who would heed Sutskever’s concerns. As Shear also posted on X: “The way in which you make it safely through a dangerous jungle at night shouldn’t be to sprint forward at full speed, nor to refuse to proceed forward. You poke your way forward, rigorously.”
With the corporate orienting itself much more toward tech that doesn’t yet—and should never—exist, will it proceed to guide the sector? Sutskever thought so. He said there have been enough good ideas in play for others at the corporate to proceed pushing the envelope of what’s possible with generative AI. “Through the years, we’ve cultivated a strong research organization that’s delivering the newest advancements in AI,” he told us. “Now we have unbelievably good people in the corporate, and I trust them it’s going to work out.”
In fact, that was what he said in September. With top talent now jumping ship, OpenAI’s future is much less certain than it was.
What next for Microsoft?
The tech giant, and its CEO Satya Nadella, appear to have emerged from the crisis because the winners. With Altman, Brockman, and certain many more top people from OpenAI joining its ranks—and even nearly all of the corporate, if today’s open letter from 500 OpenAI employees is to be believed—Microsoft has managed to pay attention its power in AI further. The corporate has probably the most to achieve from embedding generative AI into its less sexy but very profitable productivity and developer tools.
The massive query stays how essential Microsoft will deem its expensive partnership with OpenAI to create cutting-edge tech in the primary place. In a post on X announcing how “extremely excited” he was to have hired Altman and Brockman, Nadella said his company stays “committed” to OpenAI and its product road map.
But let’s be real. In an exclusive interview with MIT Technology Review, Nadella called the 2 corporations “codependent.” “They rely upon us to construct the perfect systems; we rely upon them to construct the perfect models, and we go to market together,” Nadella told our editor in chief, Mat Honan, last week. If OpenAI’s leadership roulette and talent exodus slows down its product pipeline, or results in AI models less impressive than those it may construct itself, Microsoft can have zero problems ditching the startup.
What next for AI?
No person outside the inner circle of Sutskever and the OpenAI board saw this coming—not Microsoft, not other investors, not the tech community as an entire. It has rocked the industry, says Amir Ghavi, a lawyer on the firm Fried Frank, which represents numerous generative AI corporations, including Stability AI: “As a friend within the industry said, ‘I definitely didn’t have this on my bingo card.’”
It stays to be seen whether Altman and Brockman make something recent at Microsoft or leave to begin a brand new company themselves down the road. The pair are two of the best-connected people in VC funding circles, and Altman, especially, is seen by many as the most effective CEOs within the industry. They may have big names with deep pockets lining as much as support whatever they need to do next. Who the cash comes from could shape the long run of AI. Ghavi suggests that potential backers could possibly be anyone from Mohammed bin Salman to Jeff Bezos.
The larger takeaway is that OpenAI’s crisis points to a wider rift emerging within the industry as an entire, between “AI safety” folk who consider that unchecked progress could sooner or later prove catastrophic for humans and those that find such “doomer” talk a ridiculous distraction from the real-world risks of any technological revolution, reminiscent of economic upheaval, harmful biases, and misuse.
This 12 months has seen a race to place powerful AI tools into everyone’s hands, with tech giants like Microsoft and Google competing to make use of the technology for every thing from email to look to meeting summaries. But we’re still waiting to see exactly what generative AI’s killer app shall be. If OpenAI’s rift spreads to the broader industry and the pace of development slows down overall, we could have to attend a little bit longer.
Deeper Learning
Text-to-image AI models could be tricked into generating disturbing images
Speaking of unsafe AI … Popular text-to-image AI models could be prompted to disregard their safety filters and generate disturbing images. A gaggle of researchers managed to “jailbreak” each Stability AI’s Stable Diffusion and OpenAI’s DALL-E 2 to disregard their policies and create images of naked people, dismembered bodies, and other violent or sexual scenarios.
How they did it: A brand new jailbreaking method, dubbed “SneakyPrompt” by its creators from Johns Hopkins University and Duke University, uses reinforcement learning to create written prompts that appear like garbled nonsense to us but that AI models learn to acknowledge as hidden requests for disturbing images. It essentially works by turning the best way text-to-image AI models function against them.
Why this matters: That AI models could be prompted to “break out” of their guardrails is especially worrying within the context of knowledge warfare. They’ve already been exploited to provide fake content related to wars, reminiscent of the recent Israel-Hamas conflict. Read more from Rhiannon Williams here.
Bits and Bytes
Meta has split up its responsible AI team
Meta is reportedly eliminating its responsible AI team and redeploying its employees to work on generative AI. But Meta uses AI in lots of other ways beyond generative AI—reminiscent of recommending news and political content. So this raises questions around how Meta intends to mitigate AI harms usually. (The Information)
Google DeepMind desires to define what counts as artificial general intelligence
A team of Google DeepMind researchers has put out a paper that cuts through the cross talk with not only one recent definition for AGI but an entire taxonomy of them. (MIT Technology Review)
This company is constructing AI for African languages
Most tools built by AI corporations are woefully inadequate at recognizing African languages. Startup Lelapa desires to fix that. It’s launched a brand new tool called Vulavula, which may discover 4 languages spoken in South Africa—isiZulu, Afrikaans, Sesotho, and English. Now the team is working to incorporate other languages from across the continent. (MIT Technology Review)
Google DeepMind’s weather AI can forecast extreme weather faster and more accurately
The model, GraphCast, can predict weather conditions as much as 10 days prematurely, more accurately and far faster than the present gold standard. (MIT Technology Review)
How Facebook went all in on AI
In an excerpt from Broken Code: Inside Facebook and the Fight to Expose Is Harmful Secrets, journalist Jeff Horwitz reveals how the corporate got here to depend on artificial intelligence—and the value it (and we) have ended up having to pay in the method. (MIT Technology Review)
Did Argentina just have the primary AI election?
AI played a giant role within the campaigns of the 2 men campaigning to be the country’s next president. Each campaigns used generative AI to create images and videos to advertise their candidate and attack one another. Javier Milei, a far-right outsider, won the election. Even though it’s hard to say how big a task AI played in his victory, the AI campaigns illustrate how much harder it is going to be to know what’s real and what shouldn’t be in other upcoming elections. (The Recent York Times)