Itemoids ★ Christopher Nolan

www.theatlantic.com › newsletters › archive › 2023 › 11 › sam-altman-open-ai-what-happened › 676062

This story seems to be about:

This is an edition of The Atlantic Daily, a newsletter that guides you through the biggest stories of the day, helps you discover new ideas, and recommends the best in culture. Sign up for it here.

I spoke with my colleagues Karen Hao and Charlie Warzel this afternoon about the tensions at the heart of the AI community, and how Sam Altman’s firing may ironically entrench the power of a tech giant.

First, here are three new stories from The Atlantic:

Has anyone noticed that Trump is really old? The other Ozempic revolution No, you shouldn’t “date ’em ’til you hate ’em.”

An Enabling Mantra

For a while earlier this year, Sam Altman was everywhere. As the head of OpenAI, the company that launched ChatGPT, he quickly became an emissary of the future of the technology. He appeared before Congress and foreign heads of state to discuss how AI would reshape society. As recently as last week, he was hyping up the future of his company. Then, suddenly, Altman was fired. Below is a brief timeline of the drama that unfolded:

Friday afternoon: In a blog post, the company said that Altman “was not consistently candid in his communications with the board.” Greg Brockman—the president of OpenAI who, along with Altman, had encouraged the rapid commercialization of the company’s technology—quit in solidarity. Mira Murati, formerly the chief technology officer of the company, was named interim CEO. Over the weekend: By Sunday night, OpenAI had rejected Altman’s bid to return to his job, and Microsoft (a major investor in OpenAI) had hired him to lead an AI-research lab. Emmett Shear, the former CEO of Twitch, stepped into the top role at OpenAI on an interim basis, replacing Murati. Today: Some 700 of OpenAI’s 770 employees signed a letter saying that they may leave the company and join Altman at Microsoft if he and Brockman are not reinstated at OpenAI.

What happens next may be hugely consequential for the future of AI—particularly for the question of whether profits or existential fears will drive its path forward. My colleagues Karen Hao and Charlie Warzel spoke with 10 current and former OpenAI employees, and in an article published last night, they explained how a simmering years-long tension at the company led to Altman’s ouster.

Lora Kelley: I was shocked to see the news on Friday that Sam Altman had been fired. Was this news just as stunning to those who closely watch OpenAI and the industry?

Karen Hao: It was a huge shock to me. OpenAI was at the height of its power. Altman was still doing so many meetings all around the world and hyping up the company.

Charlie Warzel: Sam Altman was essentially the avatar of the generative-AI revolution. You would think he would have a lot of leverage in discussions. If he had just simply left to start his own thing, it would have made some sense to me. It would have still been dramatic, but the fact that it was announced in this cryptic blog post accusing him of not being candid was wild. It’s one of the most shocking tech stories of the past couple of years.

Lora: You wrote in your article about the different factions within OpenAI: Some employees and leaders thought launching products and putting AI into the hands of everyday users was the right path forward, while others were more cautious and thought that stronger safety measures needed to be taken. How did that dynamic emerge over the past few years?

Karen: Sam Altman sent out an email back in 2019 acknowledging that there were different “tribes” at OpenAI. Because of the way that OpenAI was founded—the original story was that Elon Musk and Sam Altman came together and specifically founded OpenAI kind of as an entity to counteract Big Tech—it was always in the crosshairs of a lot of different ideas about AI: What is the purpose of the technology? How should we build it? How should an entity be structured? As the technology got more powerful—specifically with the catalyst of ChatGPT—so did the Game of Thrones mentality of who got to control it. That came to a head with this news this weekend.

Charlie: There is not only a power struggle but also this quasi-religious belief in what is being built or what could potentially be built. You can’t discount the fact that there are these true believers who are both energized by the idea of an all-powerful AI and horrified by it. That adds an unstable dynamic to the conversation.

Lora: You wrote in your article that this whole situation illustrates the fact that a very small group of people is shaping the future of AI. Given that OpenAI is so closely tied to the future of the technology, I’m curious: To what extent do you think of OpenAI as a traditional tech company? Did this weekend change how you see it?

Karen: The board successfully maintained its action to keep Altman out, but the question is whether or not there will still be a company left when everything falls into place. If all 700-plus employees who have signed on to the letter say that they’re going to leave and join Altman and Brockman at Microsoft now, then did firing Altman really make any difference? The whole company would be disintegrated, and OpenAI employees are ultimately going to continue commercializing, just as a branch of Microsoft.

But if, for some reason, a significant number of employees stays at OpenAI, and the company continues to move forward, then that would suggest a different model emerging. The board would have successfully taken action on its nonprofit-driven mission and very dramatically turned the company in a different direction, not on the basis of shareholders or profit optimization.

It’s too early to tell, and it really is up to the employees themselves.

Charlie: I can’t stop thinking that, if OpenAI was founded in opposition to the way that traditional tech companies were trying to develop and commercialize AI, and it was a sanctuary for those who wanted to build this technology safely, then the principled decision by the board to fire Altman, and the chain of events it has set in motion, may drive a bunch of their talent—certainly their CEO and president—into the arms of one of the largest tech companies in the world.

Karen: Ultimately, both the techno-optimists and the other faction have the same endgame: They’re both trying to control the technology. One is using morality as a cover for that, and the other one is using capitalism as its banner. But both are saying This is for the good of humanity, and they’re using that as their enabling mantra for a seizure of power and control.

Charlie: This is a very small group of people with a lot of power. This is fundamentally a power struggle.

Today’s News

The Supreme Court rejected an appeal from the former police officer Derek Chauvin for his conviction in the murder of George Floyd. Javier Milei, a hard-right libertarian who has drawn comparisons to Donald Trump, will be Argentina’s next president. President Joe Biden stated that he believes a deal to release some of the hostages Hamas is holding in Gaza is close at hand.

Evening Read

Alex Webb / Magnum

How the Hillbillies Remade America

By Max Fraser

On April 29, 1954, a cross section of Cincinnati’s municipal bureaucracy—joined by dozens of representatives drawn from local employers, private charities, the religious community, and other corners of the city establishment—gathered at the behest of the mayor’s office to discuss a new problem confronting the city. Or, rather, about 50,000 new problems, give or take. That was roughly the number of Cincinnati residents who had recently migrated to the city from the poorest parts of southern Appalachia. The teachers, police officials, social workers, hiring-department personnel, and others who gathered that day in April had simply run out of ideas about what to do about them.

“Education does not have importance to these people as it does to us,” observed one schoolteacher. “They work for a day or two, and then you see them no more,” grumbled an employer.

Read the full article.