# The Doomsday Cult Inside OpenAI

A civil war inside the world's most important AI company pits believers in safety against believers in speed—and raises a uncomfortable question: who should control the technology that may determine humanity's future?

---

In November 2023, OpenAI—the company behind ChatGPT—erupted in a crisis that stunned Silicon Valley and the world. Its board of directors fired CEO Sam Altman. Within days, he was reinstated, the board was dissolved, and the company's most prominent safety researchers had resigned. What looked like a routine corporate power struggle was, in fact, something stranger: a collision between two irreconcilable worldviews inside the most influential AI lab on Earth.

To understand the conflict, you have to understand the culture that produced it.

A Mission Born from Fear

OpenAI was founded in 2015 as a nonprofit. Its stated mission was to ensure that artificial general intelligence—a hypothetical AI system capable of matching or exceeding humans at nearly every cognitive task—benefits all of humanity rather than a single corporation or country.

The founding team included Sam Altman, Elon Musk, and several researchers who believed AGI might arrive within decades. They also believed it might be the last invention humanity ever needed to make—or the last one it ever made.

This sense of stakes shaped everything. OpenAI's employees routinely described their work in civilizational terms. The company's charter committed it to prioritizing safety over profit, and its leaders spoke openly about the possibility that the technology they were building could end human civilization if mishandled.

Critics—and, eventually, some insiders—began to describe this atmosphere differently: not as caution, but as a kind of faith.

Two Factions, One Company

By 2023, OpenAI contained two distinct camps.

The first was the accelerationists—sometimes called "e/acc" in tech circles. They believed the best path to safety was to build powerful AI quickly, deploy it widely, and let society adapt. They saw delays as dangerous, because a slow OpenAI would simply lose the race to less careful competitors. Sam Altman was widely associated with this camp.

The second was the safety faction, often linked to the company's research arm and its nonprofit board. They believed that certain capabilities were too dangerous to release without extensive safeguards, and that commercial pressure was already eroding the company's founding principles. They worried less about competitors and more about the possibility that a single mistake could be irreversible.

These were not merely different strategies. They were different religions, in effect—each with its own creation myth, its own prophets, and its own vision of salvation or apocalypse.

The Flashpoint

The immediate trigger for the November crisis was a letter. Several researchers reportedly warned the board that a new model under development posed serious risks and that safety protocols were being bypassed in the rush to ship. The board, already concerned about Altman's growing power and his parallel fundraising efforts abroad, moved to fire him.

The backlash was swift and overwhelming. Nearly all of OpenAI's employees signed a letter demanding Altman's return and threatening to resign en masse. Microsoft, the company's largest investor, offered to hire the entire staff. Within five days, Altman was back, the old board was gone, and a new board—friendlier to Altman—was installed.

The safety faction had lost. Its most visible members, including researcher Jan Leike, departed. Leike later said that safety culture at OpenAI had been "taking a back seat to shiny products."

What the Cult Label Really Means

Calling OpenAI a "doomsday cult" is provocative, but it points to something real: a group of people who believe they are working on a technology that could destroy the world, and who nonetheless continue to build it—because they believe they are the only ones who can do it safely.

That paradox is not unique to OpenAI. It runs through much of the AI industry. But OpenAI's nonprofit origins, its explicit warnings about existential risk, and its internal culture of mission-driven sacrifice made the contradiction unusually sharp.

When the company's leaders say they are racing toward AGI to prevent someone less careful from getting there first, they are making a theological argument as much as a strategic one. It requires believing that the people in the room are the right people—and that the alternative is worse.

The Unanswered Question

OpenAI's crisis ended with a clear winner, but it did not resolve the underlying conflict. The company now operates more like a conventional tech firm, with a board aligned with its CEO and a product roadmap driven by competition.

Whether that is safer or more dangerous depends entirely on which faction you believe. If AGI is decades away, the debate is academic. If it is closer, the stakes are as high as they have ever been.

What the episode made clear is that the question of who controls advanced AI—and what they believe about it—is no longer a niche concern for researchers. It is now one of the central political and moral questions of our time. And the people answering it are, by their own admission, not certain they are right.