Two recently published books—The Optimist by journalist Keach Hagey and Empire of AI by Karen Hao—offer two versions of the same crisis. Hagey, who gained access to Altman himself and his inner circle, paints a portrait of a leader balancing charisma, informal power, and belief in his own exceptionalism. Hao, who worked without authorized interviews, analyzes OpenAI as a closed system that has drifted away from its stated principles. Together, the books reveal how institutional structures prove powerless in the face of overwhelming ambition—and how even in an organization built for the public good, a central figure can become a source of systemic risk.
In Greek mythology, Prometheus stole fire from the gods and gave it to humanity. For this, he was condemned to eternal torment: chained to a rock, he was subjected each day to an eagle devouring his liver. Such was the price of the first great technological gift to mankind. In the 21st century, the story of Sam Altman, co-founder and CEO of OpenAI, echoes with similar themes.
It was he who became the driving force behind the creation of ChatGPT—a system launched in late 2022 that stunned the world with its capabilities. Along with this technological power, humanity suddenly came face to face with the risks of generative artificial intelligence.
A year later, the "capricious gods"—in this case, the independent nonprofit board of OpenAI—tried to cast Altman out. But unlike his mythological counterpart, he emerged from the ordeal not only unscathed but stronger than before.
This story became the subject of two remarkable books, released almost simultaneously on May 20—eighteen months after the dramatic events at OpenAI. Both works delve into the volatile mix of missionary zeal, internal rivalry, and mistrust within the company—first during the creation of ChatGPT, and later during the governance crisis of November 2023, when the board of directors unsuccessfully attempted to remove Sam Altman.
The authors—both seasoned journalists—have turned this Silicon Valley power struggle into a gripping and thoroughly researched narrative. Each book tells the story in its own way.
The Optimist by Keach Hagey could be described as a semi-authorized account. The author gained access to Altman himself and to key people in his circle, including family and friends. The protagonist’s personality—complex, at times contradictory—holds the reader’s attention throughout. It is neither a hagiography nor a hit job.
Karen Hao, author of Empire of AI, worked without such access: OpenAI kept her at arm’s length. But that distance adds bite to her book. Both works expose troubling traits—not only in Altman’s character but also in OpenAI’s corporate culture and the broader ecosystem of Silicon Valley’s tech elite. In an era of generative AI euphoria, such scrutiny is especially needed.
Sam Altman is a captivating figure. As Hagey writes, the first thing one notices is his slight frame and gaze full of intense focus—"as if he's speaking to the most important person in the world."
A native of the American Midwest, Altman displayed exceptional technical aptitude from a young age, coupled with an unexpectedly sharp sense of humor. At just 17, he became a public activist, coming out as gay before a school assembly in a show of support for LGBT rights.
He built his career at the intersection of technological ambition and storytelling—a combination that enables him to raise capital for bold ventures. His journey began with the location-based app Loopt. It was followed by investments in a cryptocurrency project based on iris-scanning (an attempt to verify digital identity in the age of AI), life-extension startups, nuclear fusion experiments—and, of course, a bet on the creation of superintelligence.
Some have compared Altman’s charisma to Steve Jobs’s famed "reality distortion field"—a knack for convincing people of the impossible. But while Jobs was famously abrasive, Altman is a careful listener who presents his ideas in ways that are hard to resist.
These qualities attracted powerful mentors early on. Y Combinator co-founder Paul Graham once remarked, "You could drop Sam with a parachute onto an island of cannibals—and in five years, he’d be king." Alongside his partner Jessica Livingston, Graham entrusted Altman with the leadership of YC—granting him near-absolute influence over Silicon Valley’s startup ecosystem when he was just 28.
Even then, Altman faced criticism for being overly secretive. "If Sam is smiling, it’s a very calculated gesture," Karen Hao quotes one of YC’s founders as saying. "The only time I saw him smile genuinely was when [Paul Graham] offered him the top job at YC." At Loopt, his first startup, which sold for $43 million in 2012, Altman twice faced attempts by the board to remove him as CEO—for pushing through ideas without consulting his colleagues.
Similar tensions surfaced at YC. While still running the accelerator, Altman co-founded OpenAI in 2015 with Elon Musk and other investors. According to Hagey, Jessica Livingston eventually fired him, frustrated that he used his stake in YC to recruit talent for the new venture. Some partners viewed this not only as a conflict of interest but also as an attempt to co-opt the YC brand for personal gain.
These details matter because both books converge on a central theme: Altman’s leadership style—fast-moving, ambitious, and manipulative—repeatedly brought OpenAI to the brink of crisis. Yet it was precisely this approach that helped the company amass unprecedented levels of funding, compute power, and data—positioning it at the forefront of the generative AI race.
One such rupture in 2021—later dubbed the "divorce"—led to the departure of a group of safety researchers who went on to found the rival company Anthropic. Another—known as the "crash"—saw Altman fired after the board and his closest allies lost confidence in him. According to both authors, he provided contradictory information and evaded direct answers about the company’s investment dealings. Yet within days he was reinstated—after it became clear that OpenAI might not survive without him.
At the heart of these episodes lies an ideological rift that runs through both books: a struggle between those who champion speed and those who call for caution in AI deployment. Within OpenAI, the clash between "pessimists" and "optimists" (or, in the authors’ terms, doomers and boomers) reached a destructive intensity. The former, often aligned with the effective altruism movement, focus on existential risks. The latter—so-called effective accelerationists—believe that if America doesn’t win the AI race, China will.
As Hao notes, these are ultimately two sides of the same coin: both camps are pushing the development of superintelligence as far as it can go—one by invoking visions of fire and brimstone, the other by promising a technological paradise.
Equally compelling are the rivalries depicted in both books—set in a world where nearly everyone sees themselves as a visionary, and progress happens in leaps to avoid falling behind. A central thread is the falling-out between Elon Musk and Sam Altman, laid bare in a lawsuit Musk filed against OpenAI, its CEO, and Microsoft—the main investor in the company's for-profit arm.
Despite sharing a common narrative core, the two books diverge in their interpretations—and that divergence underscores the central question of the entire saga: do the ends justify the means when it comes to the race for superintelligence?
Keach Hagey appears to lean toward a qualified yes. She attributes some of Altman’s behavior to a reluctance to engage in conflict and to Silicon Valley’s hallmark ethos of "moving fast and breaking things."
Karen Hao, by contrast, accuses OpenAI of betraying its founding mission. She is critical not only of Altman but also of the executives at other AI firms, portraying them as combatants in the same power struggle. In her view, generative models are "monstrosities," consuming excessive amounts of data, energy, and natural resources.
Hao may go too far in comparing OpenAI and similar labs to colonial empires. Yet her doubts about Altman are well founded. In any organization, a CEO who inspires questions about their reliability is a liability. In the case of OpenAI—a company developing Promethean-scale technologies—it is a risk of an entirely different magnitude.
AI Profile
Ideology at the Top, Infrastructure at the Bottom
While Washington Talks About AI’s Bright Future, Its Builders Demand Power, Land, and Privileges Right Now
AI’s Carbon Conundrum
The technology that could save the planet might also help burn it
