Sam Altman and the Race to AGI

In the spring of 2023, Sam Altman sat before the United States Senate and said something that no technology executive had ever said quite so plainly: his company might be building one of the most dangerous things in human history, and he thought the government should regulate it. Then he flew home to San Francisco and kept building.

That tension, between the warning and the acceleration, is the defining characteristic of Sam Altman's public life. He is simultaneously the most prominent advocate for slowing down artificial intelligence development and the person most responsible for speeding it up. He has called AGI, artificial general intelligence, potentially the most transformative and dangerous technology ever created. He has also staked his career, his company, and a significant portion of the world's computing resources on being the first to build it.

"The most important technology in human history is being built by a nonprofit that became a capped-profit that is now becoming something else entirely."

Understanding Sam Altman requires sitting with that contradiction rather than resolving it. He is not a hypocrite in the conventional sense. He genuinely believes both things. He believes AGI is coming regardless of what any single company does, and he believes it is better for the world if OpenAI, with its stated commitment to safety and broad benefit, gets there first. Whether that logic holds is one of the central debates of our technological moment.

From Y Combinator to the End of the World

Before OpenAI, Altman was best known as the president of Y Combinator, the startup accelerator that has produced Airbnb, Stripe, Dropbox, and hundreds of other companies. He took the role at 28, making him one of the most influential figures in Silicon Valley before most people outside the industry had heard his name. His reputation was for pattern recognition: the ability to identify which founders and ideas had the potential to become something significant.

In 2015, Altman co-founded OpenAI alongside Elon Musk, Greg Brockman, Ilya Sutskever, and others. The founding premise was unusual: a nonprofit AI research lab whose explicit mission was to ensure that artificial general intelligence benefited all of humanity. The structure was designed to insulate the organization from the profit pressures that, the founders believed, would otherwise push AI development in dangerous directions.

That structure did not survive contact with the capital requirements of frontier AI research. By 2019, OpenAI had created a "capped-profit" subsidiary to attract investment. Microsoft eventually committed more than $13 billion. The nonprofit mission remained on paper, but the organization had become something considerably more complex.

ChatGPT and the Moment Everything Changed

In November 2022, OpenAI released ChatGPT as what the company described as a research preview. Within five days, it had one million users. Within two months, it had one hundred million, making it the fastest-growing consumer application in history. The product had not been designed as a mass-market launch. It became one anyway.

The release changed everything for OpenAI, for the technology industry, and for Altman personally. He became, almost overnight, the most recognizable face of the AI revolution. He was invited to meet with heads of state. He testified before Congress. He appeared on magazine covers. He became the person journalists called when they needed someone to explain what was happening to the world.

"Within five days of launch, ChatGPT had one million users. Within two months, one hundred million. The fastest-growing consumer application in history."

The Firing and the Return

In November 2023, OpenAI's board fired Sam Altman. The announcement was abrupt, the stated reason vague. The board said he had not been "consistently candid" with them, and the fallout was immediate. Nearly every OpenAI employee signed a letter threatening to resign if Altman was not reinstated. Microsoft, which had just committed billions to the company, made clear it would hire Altman and his team if necessary. Within five days, Altman was back as CEO.

The episode revealed something important about the organization Altman had built. OpenAI's employees were not simply working for a company. They were working for a mission, and in their view, that mission was inseparable from its CEO. The board that fired him was reconstituted. The governance structure that had made the firing possible was revised. Altman emerged from the crisis with more authority than he had entered it with.

The AGI Question

Artificial general intelligence, a system capable of performing any intellectual task that a human can, has been the theoretical goal of AI research since the field's founding in the 1950s. For most of that time, it remained a distant aspiration. The question of when, or whether, it would arrive was treated as philosophical rather than practical.

Altman believes that question is now practical. In early 2025, he published an essay titled "The Intelligence Age," in which he argued that AGI was likely to arrive within the next few years and that its effects on human civilization would be profound. He described a future in which AI systems could compress decades of scientific progress into years, solve diseases that have resisted human effort for generations, and drive economic growth at a scale that would lift billions of people out of poverty.

He also acknowledged that the same technology could be catastrophically misused. He has spoken about the risk of AI being used to create biological weapons, to concentrate power in the hands of a small number of actors, and to undermine democratic institutions. He has said that OpenAI's own success could become a danger if the company itself accumulated too much power.

"He has said that OpenAI's own success could become a danger if the company itself accumulated too much power."

The Infrastructure of Intelligence

In 2025, Altman announced the Stargate Project, a joint venture with SoftBank, Oracle, and others to invest up to $500 billion in AI infrastructure in the United States over four years. The first $100 billion was committed immediately. The project represented one of the largest private infrastructure investments in American history, and it was organized around a single premise: that the future of AI would be determined, in large part, by who controlled the computing power required to build it.

The scale of the investment reflects how Altman thinks about the competitive landscape. He is not primarily worried about which company has the best researchers or the most elegant algorithms. He is worried about compute: the raw processing power that determines how large a model can be trained and how quickly it can improve. In that framing, the race to AGI is as much a logistics problem as a scientific one.

What Comes Next

OpenAI's product lineup has expanded dramatically since ChatGPT's launch. The company now offers a suite of models under the GPT-4o and o-series names, a voice interface, an image generator, a video generator, and an operator framework that allows AI agents to take actions on behalf of users. Each release has pushed the boundary of what the technology can do and accelerated the timeline for what comes next.

Altman has said he expects OpenAI to achieve AGI, by the company's own internal definition, within the next few years. He has been careful to note that the definition matters: AGI as OpenAI uses the term refers to a system that can perform the work of a highly capable knowledge worker across a wide range of domains, not a system with human-level consciousness or general reasoning in every conceivable context.

Whether that definition is the right one, and whether reaching it will feel like the transformative moment Altman has described, remains to be seen. What is not in doubt is that the decisions being made at OpenAI right now, about safety, about deployment, about governance, about who gets access to the most powerful systems, will shape the trajectory of a technology that may prove to be the most consequential in human history.

Sam Altman knows this. He has said so, repeatedly and publicly, in terms that leave little room for ambiguity. The question is whether knowing it is enough.