OpenAI’s Soap Opera Collapse Bodes Ill for AI Benefiting Humanity

Whatever fantasies we may have had about the nonprofit structure of OpenAI have been eviscerated. While it remains a nonprofit, it’s proven entirely beholden to ruthless capitalism

Event photo of Sam Altman in front of a blue background

Sam Altman, seen here at APEC Leader’s Week in San Francisco, California, was fired briefly as CEO of OpenAI in Nov. 2023.

On November 17 Sam Altman was fired as CEO of OpenAI, arguably the most prominent privately held tech company, famed for pioneering ChatGPT and sparking the current AI boom, as well as fears of “risk of extinction” from the technology.

Altman was sacked—briefly—for a lack of “consistent candor” with his nonprofit board. One might wonder how a company could suddenly oust Silicon Valley’s most beloved son, and it’s largely thanks to OpenAI’s convoluted, multitiered corporate structure. A nonprofit entity fully owns its subordinate for-profit entity, which sells ChatGPT—the same subordinate that Microsoft invested $10 billion into in January. This nonprofit structure exists—or existed, as you’ll find out—to focus on building an artificial general intelligence (AGI) that “benefits humanity”—a noble goal, and one that theoretically protected the company from the influence of the tech industry’s growth-at-all-costs “Rot Economy.” The board was at the time made up of Altman, the company’s then president Greg Brockman, its chief scientist Ilya Sutskever, AI safety researcher Helen Toner, Quora CEO Adam D’Angelo and Tasha McCauley, a robotics engineer who was one of the board members with deep ties to the “effective altruism” movement, which fears an AGI could destroy humanity.

The Vichy France–rivaling collapse of this board and its intended nonprofit firewall, all unable to withstand the unfettered force of raw money power, revealed in the Altman saga where the AI revolution is headed—and it’s wherever the most rich and powerful people in tech want it to go.

The firing’s first few days were a confusing whirlwind, where nobody, not even the board, would say exactly why Altman was fired. A day before, he was trumpeting a monumental breakthrough in ChatGPT onstage at the APEC conference in San Francisco. Reuters theorized that might have been Q* (pronounced “Q-star”), an artificial intelligence that can do grade school math (though there are now reports refuting this story), a significant breakthrough that would mean an artificial intelligence can learn rules. An $80-billion valuation was rumored for the company.

The board rushed to place former Twitch CEO Emmett Shear at the helm, which led to hundreds of OpenAI employees threatening to quit if Altman wasn’t put back on top. Meanwhile Microsoft hired both Altman and Brockman, with Microsoft CEO Satya Nadella demanding changes to OpenAI’s nonprofit governance structure.

A day later, Altman returned as CEO of OpenAI with a new, all-male board of directors, replacing technological idealists with a who’s who of Rot Economy capitalists such as Bret Taylor, former co-CEO of Salesforce (a company that burns billions to make millions), D’Angelo (the only returning member), and former U.S. secretary of the treasury and former Harvard University president Larry Summers, who is best known for calling for polluting less-developed nations and for blowing his economic predictions. Worse still, Microsoft now holds a “nonvoting observer seat” on the board, giving the bloated software titan full visibility into every major decision at OpenAI. Vote or not, Nadella will now have direct influence on everything that OpenAI does going forward.

While OpenAI remains a nonprofit organization, it’s proven entirely beholden to the forces of capitalism, forced to oust its entire board based on a decision that the potentates of Silicon Valley—men such as Khosla Ventures founder Vinod Khosladidn’t like.

While the Valley applauds this as a “win” for the “good guys,” it’s important to see this situation clearly: a nonprofit was defanged by a conspiracy of extremely rich people who didn’t like its personnel decisions. This is a resounding victory for “effective accelerationists,” the fervent acolytes of the Rot Economy, who believe we should build technology as fast as humanly possible, no matter the cost. They don’t care that more than 30,000 people have been laid off in the Bay Area alone in the past two years or that many tech companies are deeply unprofitable and heavily reliant on unreliable Faustian bargaining with venture capital. OpenAI itself is on pace to make more than $1 billion of revenue in 2023, for example, and still isn’t clearly profitable. That’s likely because of the vast cloud computing costs that come with running large language models. Microsoft also has a tight hold on OpenAI’s leash—despite “investing” $10 billion in OpenAI last year, OpenAI has only received a fraction of that money, which is both divided into tranches and mostly made up of cloud computing credits.

Whatever fantasies we may have had about the nonprofit structure of OpenAI have been eviscerated by the dread hand of capitalism. Worse still, we don’t actually know why Altman was fired, other than that it was a conflict between the for-profit and nonprofit sides of the company. Regardless, the result is that the largest artificial intelligence company in the world is—corporate structure be damned—controlled by venture capitalists and a multitrillion-dollar public tech company. The nonprofit board is entirely subordinate to Altman, who is subordinate to the financial interests that put him back in power, which can (and will) control his destiny in the same shadowy way that they reinstalled him.

And that’s fundamentally worrying. While the execution of the coup was messy, it’s obvious that Altman’s mindset is locked intimately with the growth-at-all-costs Valley mindset—to the point that he was trying to raise money from sovereign wealth funds in the Middle East for a future AI chipset company. The new OpenAI is one that is laser-focused on growth—in revenue, in users and in capabilities—with no concern for whether its products are actually improving as they sell into critical infrastructure like the life sciences.

The OpenAI debacle is a dark day for the Valley, where the rich and powerful used their might to crush those who won’t aggressively and recklessly pursue technological progress at all costs. Putting aside the drama and intrigue, this is a cautionary tale, and proof that the Valley’s elite only wants you to “Think different” if you’re thinking exactly like they are.

This is an opinion and analysis article, and the views expressed by the author or authors are not necessarily those of Scientific American.