The Longest Vacation
On August 5, 2024, Greg Brockman posted a message that sent tremors through Silicon Valley's most watched company. "I'm taking a sabbatical through end of year," he wrote on X. "First time to relax since co-founding OpenAI 9 years ago."
The timing was conspicuously terrible. OpenAI was hemorrhaging talent at an alarming rate. Co-founder John Schulman had just announced his departure for rival Anthropic. Ilya Sutskever, the chief scientist who had helped orchestrate the previous year's dramatic boardroom coup, had left months earlier to launch Safe Superintelligence. CTO Mira Murati would depart weeks later. Of the eleven original founders who had gathered in Brockman's San Francisco living room in December 2015, only three remained: Sam Altman, Wojciech Zaremba, and Brockman himself—now stepping away indefinitely.
"Mission far from complete," Brockman added, cryptically.
Three months later, on November 12, 2024, he returned. "Longest vacation of my life," he posted. "Back to building OpenAI."
What Brockman came back to was a company transformed—and a role that had evolved beyond anything imagined during those early days when the world's most ambitious AI lab operated out of his apartment. OpenAI's valuation had soared past $157 billion. Weekly active ChatGPT users had climbed to 500 million. And the company was pivoting from its identity as a research lab to something unprecedented: an infrastructure colossus planning to deploy $1.4 trillion in compute capacity over the coming years.
The sabbatical, it turned out, was not a retreat. It was a prelude. Greg Brockman, the engineer's engineer who had always preferred building systems to building his profile, was about to become the most important operator in the history of artificial intelligence.
The Science Prodigy from Thompson, North Dakota
Gregory Brockman was born on November 29, 1987, in Thompson, North Dakota—a town of fewer than 1,000 people situated along the banks of the Red River, roughly 15 miles north of Grand Forks. His parents, Ronald Brockman and Ellen Feldman, were physicians who worked at Altru Health System, the regional healthcare provider. It was an unlikely origin story for someone who would one day help build one of the most transformative technologies in human history.
But Thompson, for all its isolation, produced something remarkable in young Greg Brockman: an insatiable intellectual curiosity uncontaminated by the conformist pressures of coastal tech culture. At Red River High School, he excelled in mathematics, chemistry, and computer science with a precocity that marked him as exceptional even among gifted students.
In 2006, Brockman won a silver medal at the International Chemistry Olympiad. That same year, he became the first finalist from North Dakota to participate in the Intel Science Talent Search since 1973—a gap of thirty-three years. The achievement was so unusual for the state that it made local news, a small-town prodigy who had somehow taught himself to compete at the highest levels of scientific competition without access to the elite resources available to students in Boston or San Francisco.
His summers were spent at Canada/USA Mathcamp, an intensive program for mathematically gifted high schoolers, in 2003, 2005, and 2007. These experiences exposed him to a community of peers who shared his obsessive intellectual drive—and to the realization that his future lay far from the Red River Valley.
It was during a gap year after high school that Brockman first encountered the idea that would eventually define his life. He read Alan Turing's "Computing Machinery and Intelligence," the foundational 1950 paper that asked whether machines could think. Something about Turing's vision captivated him: the notion that you could write code capable of understanding things that you, as the code's author, did not understand yourself. It was an idea that would remain with him through two incomplete university educations, a transformative stint at a payments startup, and ultimately to the co-founding of the most influential AI company in the world.
The Dropout's Path
In 2008, Brockman enrolled at Harvard University, intending to double-major in mathematics and computer science. He threw himself into the Harvard Computer Society, administering and building systems that served the undergraduate community. But the traditional academic path chafed against his restless need to build things that mattered.
"I'd started out at Harvard and transferred to MIT, trying to constantly surround myself by people who I could learn from and build something useful with," Brockman later wrote on his blog. At MIT, he worked on popular projects including XVM, Linerva, and scripts.mit.edu—infrastructure that fellow students actually used.
But even MIT couldn't hold him. Within months of transferring, Brockman dropped out entirely. It was 2010, and he had found something more compelling than a diploma: an invitation to join a tiny payments startup called Stripe, founded by his MIT classmate Patrick Collison and his brother John.
The decision to leave MIT was not made lightly. Brockman came from a family of physicians who understood the value of credentials. But he had recognized something that would prove prophetic: in the emerging world of technology, what you could build mattered infinitely more than what degree hung on your wall.
Stripe, when Brockman joined, had roughly four employees. The Collison brothers had a vision of making online payments as simple as a few lines of code—a radical proposition in an era dominated by clunky legacy systems. Brockman was brought on as an engineer, one of the company's earliest technical hires.
The Stripe Years—Building at Scale
At Stripe, Brockman discovered his calling. He was not just a good engineer; he was the kind of engineer who understood that building great systems required more than elegant code. It required organizational architecture, cultural design, and the ability to scale both technology and teams simultaneously.
In the early days, Brockman focused on backend infrastructure—designing server architecture, creating the credit card vault for secure storage, building foundational APIs for payment processing. These were not glamorous tasks, but they were essential. A single vulnerability in the payment infrastructure could destroy the company. A poorly designed API could cripple growth. Brockman approached these challenges with the rigor of someone who understood that infrastructure failures don't just create technical problems—they destroy trust.
As Stripe grew, so did Brockman's responsibilities. He found himself spending less time writing code and more time doing what he would later call "full-time early employee" work: writing cultural guides, onboarding new hires, running the recruiting program. By 2013, the company had promoted him to its first Chief Technology Officer.
The CTO title, however, came without a clear job description. Brockman would later write extensively about the eighteen months it took him to define what being CTO at Stripe actually meant. His conclusion was characteristically practical: "Technical leadership should call the shots while still doing hands-on technical work."
Under his leadership, Stripe grew from 5 to 205 employees. The company became a model for how to scale technical organizations without sacrificing engineering excellence—a rare achievement in an industry littered with companies that had grown fast and broken everything.
But by early 2015, Brockman was growing restless. Stripe, now valued at $3.5 billion, was no longer a scrappy startup. The systems he had built were humming along with the reliability of a utility. And a dinner conversation was about to change everything.
The Genesis of OpenAI
In August 2015, a small group gathered for dinner in Silicon Valley. The attendees included Elon Musk, Sam Altman (then president of Y Combinator), Ilya Sutskever (a top researcher who had studied under AI pioneer Geoffrey Hinton at Google), and Greg Brockman. The topic was artificial intelligence—and specifically, the growing concern that the technology's development was becoming dangerously concentrated.
Google had acquired DeepMind in 2014 for $500 million. Facebook was building its own AI research lab. The most advanced work in machine learning was increasingly locked inside corporate research divisions with no obligation to share their findings—or their safety protocols—with the broader world.
Musk and Altman proposed something audacious: a nonprofit AI research laboratory that would develop artificial general intelligence (AGI) safely and ensure its benefits were distributed broadly. The name, Musk suggested, would be "Open AI Institute"—or simply, OpenAI.
Brockman was hooked immediately. This was the Turing vision he had encountered as a teenager in North Dakota, now rendered concrete and urgent. He committed on the spot and immediately began planning logistics.
"Sam and Brockman started rallying a team to turn this idea into reality," one account of the founding recalls. "They were missing a core ingredient: they needed an AI technical visionary. Ilya Sutskever was clearly the best person in the world for this."
Recruiting Sutskever from Google was a coup. He was one of the most talented machine learning researchers in the world, a key architect of the neural network revolution that was transforming AI. Brockman and Altman convinced him to leave a well-compensated position at Google for the uncertainty of a nonprofit with grand ambitions and no track record.
In December 2015, OpenAI was formally announced. The founding team included eleven people: Sam Altman, Elon Musk, Greg Brockman, Ilya Sutskever, Trevor Blackwell, Vicki Cheung, Andrej Karpathy, Durk Kingma, John Schulman, Pamela Vagata, and Wojciech Zaremba. The press release trumpeted $1 billion in committed funding from private investors including Musk, Altman, and PayPal co-founder Peter Thiel.
The reality was more modest. As internal emails later revealed during Musk's lawsuit against OpenAI, only a small fraction of that billion dollars actually materialized in the early years. When Brockman and Altman had initially planned to raise $100 million, Musk had insisted on announcing a bigger number "to avoid sounding hopeless." The gap between the announced billion and actual capital would shape the company's evolution in ways no one anticipated.
OpenAI's first headquarters was Greg Brockman's living room. The company that would one day deploy trillions of dollars in infrastructure started with founders working from couches and kitchen tables, fueled by conviction that they were building something that would reshape human civilization.
The First CTO
At OpenAI, Brockman served as Chief Technology Officer—a role that leveraged everything he had learned at Stripe about building organizations that could operate at the frontier of technical possibility. His job was to create the systems and culture that would allow some of the world's most brilliant researchers to do their best work.
The early projects were eclectic. Brockman led the development of OpenAI Gym, a toolkit for developing and comparing reinforcement learning algorithms that became a standard in the field. He oversaw OpenAI Five, an AI system that learned to play the complex strategy game Dota 2 well enough to compete against professional human players.
These projects served multiple purposes. They attracted talent by offering researchers the opportunity to work on challenging, publicly visible problems. They generated data about how AI systems learned and scaled. And they built the organizational capabilities that would prove essential when OpenAI shifted its focus to language models.
The transition to GPT (Generative Pre-trained Transformer) began in 2018. The first GPT model demonstrated that training large neural networks on vast amounts of text data could produce systems with remarkable linguistic capabilities. GPT-2, released in 2019, was so convincing at generating text that OpenAI initially withheld the full model over concerns about potential misuse.
By 2020, GPT-3 had arrived—and with it, a fundamental shift in what seemed possible. The model could write essays, generate code, answer questions, and perform tasks that seemed to require genuine understanding. Silicon Valley began to realize that OpenAI was onto something transformative.
The Pivot to Commercial Reality
But transformation requires capital, and OpenAI was running into the limits of nonprofit economics. Training advanced AI models demanded enormous computing resources. GPT-3 cost an estimated $4.6 million just for a single training run. The next generation would be orders of magnitude more expensive.
In 2019, OpenAI restructured. The nonprofit became a parent organization overseeing a new "capped-profit" subsidiary. Investors could now put money into OpenAI and receive returns—but those returns were capped at 100x their investment. Anything beyond that would flow back to the nonprofit's mission.
The restructuring was controversial. Elon Musk had already departed the board in 2018, citing conflicts of interest with Tesla's own AI work. He would later sue OpenAI, alleging that the company had abandoned its founding principles by pursuing profit. Internal emails revealed that Brockman, Sutskever, and Altman had all favored some form of commercial structure as early as 2017, while Musk had resisted.
The compromise—capped-profit—allowed OpenAI to raise the billions it needed while ostensibly preserving some connection to its original mission. Microsoft invested $1 billion in 2019 and would eventually commit over $13 billion, becoming OpenAI's most important partner and largest shareholder outside the nonprofit parent.
For Brockman, the restructuring represented a pragmatic acceptance of reality. Building AGI would cost more money than any nonprofit could sustainably raise. The choice was between compromising on the business model or compromising on the mission. He and Altman chose to preserve the mission.
ChatGPT and the World Stage
On November 30, 2022, OpenAI released ChatGPT. The company expected modest interest—perhaps a million users within a few months. They got a million users in five days.
ChatGPT was not a fundamentally new technology. It was built on GPT-3.5, an iteration of the language model architecture OpenAI had been developing for years. But the interface—a simple chat window where anyone could converse with an AI—made the technology accessible in a way that previous releases had not.
Students used it to write essays. Programmers used it to debug code. Journalists used it to draft articles. The public, which had been hearing about AI breakthroughs for years without experiencing them directly, suddenly understood what all the fuss was about.
For Brockman, ChatGPT's success validated years of infrastructure investment. The systems he had built—for training, for deployment, for scaling—were now serving hundreds of millions of users. The technical challenges were immense: maintaining reliability, managing costs, preventing misuse. But the systems held.
On March 14, 2023, Brockman took the stage to unveil GPT-4 in a live video demo. The fourth-generation model represented a significant leap in capability, able to reason about images, pass professional licensing exams, and perform complex multi-step tasks that GPT-3 had struggled with.
A month later, at TED2023 in Vancouver, Brockman delivered what would become one of the most watched tech presentations of the year. With head of TED Chris Anderson looking on, he demonstrated unreleased ChatGPT plugins that allowed the AI to browse the web, execute code, and interact with external services—all live, without a safety net.
"A computer is a bicycle for the mind," Brockman said, quoting Steve Jobs. "I think AI is almost like a rocket ship for the mind. We are able to go places, solve problems that are just totally impossible."
The demos were electrifying. ChatGPT created a recipe, generated an image of the finished dish, drafted a tweet about it, and built a grocery list in Instacart—all without leaving the chat interface. It fact-checked its own work with citations. It interpreted complex spreadsheets with vague instructions.
But Brockman also addressed the risks. "I think this technology is so desirable, it will be built with or without us," he said. "But what we can do, what we can all do, is help steer it, is help decide how we want this to integrate with society."
It was a philosophy of "responsible acceleration"—the belief that AI must advance quickly enough to deliver benefits, but not recklessly enough to lose control. The approach would be tested sooner than anyone expected.
Five Days in November
On the evening of Thursday, November 16, 2023, Sam Altman received a text message from Ilya Sutskever, OpenAI's chief scientist and one of its co-founders. Sutskever wanted to talk the next day at noon.
When Altman joined the Google Meet call at the scheduled time, the entire board was present—except for Greg Brockman, who was simultaneously receiving his own notification that he was being removed as board chairman. Sutskever delivered the news: Altman was being fired, effective immediately. The announcement would go out within minutes.
The board's statement cited a "deliberative review process" that had concluded Altman was "not consistently candid in his communications." No specific allegations were provided. No warning had been given.
"Sam and I are shocked and saddened by what the board did today," Brockman wrote on X later that evening. "We too are still trying to figure out exactly what happened."
Then he added: "based on today's news, i quit."
The next 96 hours would become the most dramatic corporate governance crisis in Silicon Valley history. What had begun as a board coup transformed into a full-scale employee rebellion.
The Rebellion
Within hours of the announcement, OpenAI employees began organizing. By Sunday, more than 500 of the company's 700 employees had signed a letter threatening to resign unless the board reversed its decision and resigned itself. The letter warned that Microsoft had offered to hire any OpenAI employee who wanted to leave.
The pressure was overwhelming. The board had apparently not anticipated that firing a popular CEO would trigger an existential crisis. Their position collapsed almost immediately.
Most remarkably, Ilya Sutskever himself publicly reversed course. On Monday, he posted: "I deeply regret my participation in the board's actions. I never intended to harm OpenAI. I love everything we've built together and I will do everything I can to reunite the company."
According to later reporting by The Wall Street Journal, Sutskever's change of heart came after an emotionally charged conversation with Anna Brockman—Greg's wife, whom Sutskever had married to Greg in a ceremony at OpenAI's offices in 2019, with a robotic hand serving as ring bearer. Anna cried and pleaded with Sutskever to reconsider. The personal history between the families made the betrayal unbearable for all involved.
By Tuesday, November 21, a deal was struck. Altman would return as CEO. Brockman would return as president. The existing board—Helen Toner, Tasha McCauley, and Adam D'Angelo—would be replaced by a new board chaired by Bret Taylor, the former Salesforce co-CEO, with D'Angelo remaining as the sole holdover.
An independent investigation by law firm WilmerHale would later conclude that Altman's behavior "did not mandate removal." The board, it turned out, had fired the CEO of the world's most important AI company based on disagreements about the pace of commercialization—disagreements that could have been resolved through normal governance processes.
Aftermath
The November crisis left scars that would never fully heal. Sutskever, despite his public reversal, left OpenAI six months later to found Safe Superintelligence, a startup focused on building AI safely without commercial pressures. His departure was framed as amicable, but colleagues noted that his relationship with Brockman had become distant. When asked during later legal proceedings about his last communication with Brockman, Sutskever said it had been "maybe a year and a quarter ago."
The crisis also accelerated a broader exodus. John Schulman, another co-founder, left for Anthropic in August 2024. Mira Murati, the CTO who had served as interim CEO during the five-day chaos, announced her departure in September. Research chief Bob McGrew and VP Barret Zoph followed her out the door.
Of the eleven people who had gathered in Brockman's living room in 2015, only three remained: Altman, Zaremba, and Brockman himself. The company they had built together had survived the crisis. The founding team had not.
The Sabbatical
When Brockman announced his sabbatical in August 2024, observers wondered whether he, too, was preparing to leave. The timing—amid cascading departures and just nine months after the most traumatic experience of his professional life—suggested a man at a breaking point.
But Brockman's announcement carried a different tone. "First time to relax since co-founding OpenAI 9 years ago," he wrote. Not a goodbye, but a pause. Not a retreat, but a recalibration.
The three months away gave Brockman something he had not had since leaving MIT in 2010: space to think without the pressure of daily operations. OpenAI was changing—from a research lab that occasionally shipped products to an infrastructure company that would need to deploy unprecedented amounts of capital. The skills that had made Brockman successful as CTO might not be the skills required for what came next.
Or perhaps they were exactly the skills required. Brockman had built Stripe's infrastructure from scratch, scaling systems that processed billions of dollars in payments. He had built OpenAI's research infrastructure, creating the organizational architecture that produced GPT and ChatGPT. Now the company needed someone who could build at a scale that dwarfed everything that came before.
On November 12, 2024, Brockman returned. "Back to building OpenAI," he posted. His new role, an OpenAI spokesperson confirmed, would focus on "key technical challenges" for the company.
Those challenges were about to become the largest infrastructure project in the history of technology.
The Master Builder
On January 21, 2025, President Donald Trump stood in the White House alongside Sam Altman, SoftBank CEO Masayoshi Son, and Oracle chairman Larry Ellison to announce the Stargate Project: a joint venture that would invest up to $500 billion in AI infrastructure in the United States over the next four years.
The scale was staggering. Stargate would build data centers capable of powering AI systems that did not yet exist, in anticipation of compute demands that were growing faster than anyone had predicted. The initial funding came from SoftBank and OpenAI ($19 billion each), with Oracle and MGX (an Abu Dhabi investment fund) contributing $7 billion apiece.
But $500 billion was just the beginning. Altman soon clarified that Stargate, combined with OpenAI's other infrastructure commitments, totaled approximately $1.4 trillion in planned compute deployment. To put this in perspective: the entire Manhattan Project, adjusted for inflation, cost roughly $30 billion. The Interstate Highway System cost around $500 billion. OpenAI was planning to spend nearly three times the highway system's cost on AI infrastructure alone.
And Greg Brockman was the man tasked with making it happen.
The Builder-in-Chief
While Sam Altman crisscrosses the globe meeting heads of state and closing deals, Brockman has become what Fortune magazine calls OpenAI's "builder-in-chief"—the executive responsible for translating trillion-dollar ambitions into operational reality.
The job requires navigating a thicket of challenges that would overwhelm most operators. OpenAI needs to:
Secure chips. AI models run on specialized hardware, primarily NVIDIA GPUs, which are in chronic short supply. In October 2025, Brockman and AMD CEO Lisa Su announced a multiyear partnership worth tens of billions of dollars that will see OpenAI deploy hundreds of thousands of AMD chips across Stargate data centers—reducing dependence on NVIDIA and diversifying supply chains.
Build data centers. As of September 2025, Stargate has announced sites in Shackelford County, Texas; Doña Ana County, New Mexico; Lordstown, Ohio; Milam County, Texas; and additional midwestern locations. Combined capacity: nearly 7 gigawatts, with plans to reach 10 gigawatts by year's end. For comparison, a typical large data center uses about 100 megawatts. OpenAI is building at 100x that scale.
Manage power. AI data centers are extraordinarily energy-intensive. The Stargate facilities will require more electricity than many small countries. OpenAI is exploring partnerships with energy providers, nuclear facilities, and renewable developers to ensure sufficient power without overwhelming local grids.
Operate globally. Stargate UAE launched in May 2025, with 1 gigawatt of planned capacity in the United Arab Emirates. Stargate Norway followed in July. Stargate Argentina, announced in October, will develop a facility in Patagonia powered by renewable energy. The infrastructure is no longer just American—it's planetary.
Lisa Su, who has worked closely with Brockman on the AMD partnership, describes his approach with something like awe. "What I love the most about working with Greg is he's just so clear in his vision that compute is the currency of intelligence, and his just maniacal focus on ensuring there's enough compute in this world."
She adds: "Failure is not an option. The infrastructure we're building is at a very different scale from how normal people build. We're building gigawatts of compute in a very short amount of time."
The Operator
Brockman's leadership style is the opposite of Altman's. Where Altman is a charismatic futurist comfortable on stages and in magazine profiles, Brockman is methodical, technical, and allergic to the spotlight. Employees describe him as "calm under chaos"—the person who brings rigor to moments of turbulence.
Former colleagues note his ability to hold two contradictory imperatives in mind simultaneously: the urgency of moving fast and the discipline of building things right. At Stripe, he built systems that processed billions of dollars without catastrophic failure. At OpenAI, he built research infrastructure that produced breakthrough after breakthrough while maintaining enough stability for millions of users.
His philosophy, articulated in blog posts and interviews over the years, centers on what he calls the importance of "technical leadership calling the shots while still doing hands-on technical work." He is not a manager who delegates and disappears. He is an architect who draws the blueprints and then picks up a hammer.
Known within Silicon Valley as "the engineer's engineer," Brockman sees OpenAI not primarily as a research organization or a product company, but as a systems problem. "Greg sees OpenAI as a systems problem, not just a research lab," one person close to the company told Fortune. The insight explains everything about his current focus: you cannot deliver intelligence at planetary scale without building planetary-scale infrastructure.
The Road Ahead
OpenAI enters 2026 as the most consequential technology company on Earth—and possibly the most precarious. Its valuation has reached $300 billion following a $40 billion funding round led by SoftBank. Annual recurring revenue has climbed past $13 billion, with projections pointing toward $20 billion by year's end. Weekly active ChatGPT users have surged to 700 million.
But the company is also burning cash at an unprecedented rate—$8 billion in 2025 alone, $1 billion more than forecast at the start of the year. Much of that spending is going toward the infrastructure buildout that Brockman oversees. The Stargate investments are not yet generating returns. They are bets on a future that OpenAI is racing to create.
The company is simultaneously navigating a complex corporate restructuring. OpenAI's goal is to convert its capped-profit subsidiary into a public-benefit corporation, a structure that would give the for-profit arm independence from the nonprofit that currently governs it. The restructuring is a condition of the SoftBank funding round; if it fails, billions of dollars in committed capital could evaporate.
Adding to the pressure: Elon Musk's ongoing lawsuit, which alleges that OpenAI has betrayed its founding principles by pursuing commercial interests. The case has exposed years of internal tensions—including early disagreements among the founders about whether to remain a nonprofit. Whatever the legal outcome, the lawsuit ensures that OpenAI's original mission will remain a subject of public debate.
The Three Remaining Founders
Of the eleven founders who launched OpenAI in December 2015, three remain: Sam Altman, Greg Brockman, and Wojciech Zaremba.
Altman is the visionary—the dealmaker who raises billions, the evangelist who preaches the gospel of AGI to governments and corporations worldwide. His job is to imagine a future that most people cannot yet comprehend and convince them to invest in making it real.
Zaremba, who leads language and code generation research, represents continuity with OpenAI's technical roots. He contributed to core initiatives including GPT, GitHub Copilot, and Codex—the foundation on which everything else is built.
Brockman occupies the space between vision and execution. He translates Altman's trillion-dollar dreams into concrete plans, timelines, and systems. He builds the organizations that build the infrastructure that runs the models that serve the users. Without him, OpenAI would have ambitions but no capacity to realize them.
The partnership between Altman and Brockman is one of the most consequential in technology history. They have known each other since before OpenAI existed, survived the November crisis together, and emerged with their alliance intact. Altman handles the world outside; Brockman builds the world within.
The AGI Question
Everything OpenAI does—the infrastructure buildout, the funding rounds, the corporate restructuring—is in service of a single goal: artificial general intelligence. AGI, in OpenAI's conception, is AI that can perform any intellectual task that a human can. It is the technology that would transform every industry, automate every job, and potentially pose existential risks to human civilization.
OpenAI believes AGI is coming—and coming soon. Internal timelines, reported by various outlets, suggest the company expects to achieve AGI within the next few years. The Stargate infrastructure is being built not for today's models, but for models that do not yet exist, models that will require orders of magnitude more compute than anything currently deployed.
Brockman's philosophy—"responsible acceleration"—reflects the company's approach to this challenge. Move fast enough to capture the benefits and maintain leadership. Move carefully enough to avoid catastrophe. Build the systems that will allow humanity to guide AI development rather than simply react to it.
Whether this balance is achievable remains an open question. The departures of Sutskever, Schulman, and other researchers focused on AI safety suggest that not everyone at OpenAI agreed with how the balance was being struck. The rapid commercialization of ChatGPT, the partnership with Microsoft, the pivot to infrastructure—each decision prioritized acceleration over caution.
But Brockman, who has been there from the beginning, who built the systems and recruited the team and survived the crisis, continues to believe that the mission is achievable. That OpenAI can develop AGI safely and ensure its benefits are broadly distributed. That the trillion-dollar bets he is now overseeing will pay off not just in financial returns, but in genuinely transformative technology that improves human life.
It is the same vision that captivated him as a teenager in Thompson, North Dakota, reading Turing's paper about machines that could think. The same vision that led him to drop out of two elite universities and join a payments startup and then co-found an AI lab in his living room. The same vision that, three decades after that first encounter with "Computing Machinery and Intelligence," has made Greg Brockman the master builder of the most ambitious technological project in human history.
The Engineer's Engineer
In October 2025, Brockman sat for an interview with CNBC alongside Lisa Su at an event announcing the AMD partnership. The contrast with Altman's globe-trotting style was stark: Brockman was understated, precise, focused entirely on the technical and operational challenges of scaling AI infrastructure.
"Steve Jobs had this famous quote that a computer is a bicycle for the mind," he said. "And I think that AI is almost like a rocket ship for the mind, right? I think that we are able to go places, solve problems that are just totally impossible."
The quote captured something essential about Brockman's worldview. He is not primarily interested in AI as a business opportunity or a geopolitical chess piece. He is interested in AI as a tool for expanding human capability—the same interest that drew him to programming as a teenager, to infrastructure engineering at Stripe, to the foundational work of building OpenAI.
His path from Thompson, North Dakota to the helm of a trillion-dollar infrastructure project is as improbable as it is instructive. He did not follow the conventional route through elite institutions and prestigious credentials. He dropped out—twice—when he found something more compelling to build. He joined a four-person startup when he could have pursued a comfortable career at a prestigious tech company. He co-founded an AI lab in his living room when that lab had no revenue, no products, and no certainty of success.
At every turn, Brockman chose building over credential-collecting, execution over planning, systems over status. The choices made him the person uniquely qualified for the role he now occupies: translating the most ambitious vision in technology into operational reality.
The infrastructure he is building will outlast him. The data centers rising in Texas and New Mexico and Norway and Argentina will process AI workloads for decades to come. The organizational systems he has designed—for recruiting, for scaling, for managing complexity—will shape how OpenAI operates long after he has moved on to whatever comes next.
And the mission he has pursued for nearly a decade—building artificial general intelligence safely and ensuring its benefits are broadly distributed—will either succeed or fail based in large part on decisions being made right now, under his oversight, in the largest infrastructure buildout in human history.
Greg Brockman, the science prodigy from a small town on the Red River, the double dropout who built Stripe's technical foundation and co-founded the world's most important AI company, is now the master builder of the age of intelligence. Whether that age brings flourishing or catastrophe depends on whether he and his colleagues can build not just powerful systems, but safe ones.
It is the question that has animated his entire career. The answer will shape the future of humanity.