Sam Altman is, by most accounts, the single most consequential figure in the commercialization of artificial intelligence. As CEO of OpenAI, he oversaw the launch of ChatGPT — the fastest-growing consumer application in history — and has since steered the company from a quiet research lab into a juggernaut valued at over $300 billion, projecting revenues of $280 billion by 2030. He is the face of the AI revolution, and depending on who you ask, either its greatest visionary or its most skilled salesman.

Born in Chicago in 1985 and raised in St. Louis, Missouri, Altman received his first Apple Macintosh at the age of eight and immediately began learning to code and disassembling hardware. He attended the prestigious John Burroughs School before enrolling at Stanford University to study computer science. He dropped out after two years — later remarking that he learned more playing poker with classmates than attending lectures.

The Y Combinator Years

In 2005, the 19-year-old Altman co-founded Loopt, a location-based social networking app that became one of the first companies funded by Y Combinator. Though Loopt never achieved mass adoption — it was eventually sold to Green Dot for $43 million in 2012 — it earned Altman a seat at the table of Silicon Valley's startup elite.

In 2011, Paul Graham invited Altman to join Y Combinator as a partner, and by 2014 he had risen to president. Under his leadership, YC cemented its reputation as the premier launchpad for tech startups, having helped companies like Airbnb, Stripe, DoorDash, Reddit, and Twitch. He expanded its ambitions, aiming to fund 1,000 new companies per year and investing in "hard technology" beyond typical software plays.

Altman left Y Combinator in 2019 to focus full-time on OpenAI, though the transition was not entirely smooth — reports later emerged of tensions around his departure and self-appointment as chairman.

Career Timeline

• 1985Born in Chicago; raised in St. Louis, Missouri
• 2003–2005Studied computer science at Stanford; dropped out to found Loopt
• 2005Co-founded Loopt, one of the first Y Combinator–backed companies
• 2011–2019Partner, then president of Y Combinator; oversaw ~1,900 companies
• 2015Co-founded OpenAI with Elon Musk and others as a nonprofit research lab
• 2019Left YC to become full-time CEO of OpenAI; introduced the "capped-profit" model
• 2022Launched ChatGPT — reached 100 million users in two months
• Nov 2023Briefly ousted by OpenAI's board; reinstated within five days
• 2025OpenAI's annualized revenue crossed $20 billion; published "The Gentle Singularity"
• 2026Announced enterprise pivot as top priority; OpenAI begins testing ads in ChatGPT

Building OpenAI

OpenAI was founded in 2015 as a nonprofit, with $1 billion in backing from Altman, Elon Musk, Peter Thiel, and others. The mission was explicit: develop artificial general intelligence for the benefit of all humanity. Musk departed in 2018 over potential conflicts with Tesla's AI work, leaving Altman to carry the project forward.

Recognizing that AI research required staggering resources, Altman introduced a "capped-profit" model in 2019 — a novel corporate structure in which profits are limited in order to keep the mission front and center. Microsoft subsequently invested billions, becoming OpenAI's primary compute partner and commercial ally.

The November 2022 launch of ChatGPT — originally conceived as a demo built on GPT-3.5 — changed everything. It reached 100 million users in roughly two months, a speed record no consumer product had previously achieved. Suddenly, Altman was not just a startup CEO. He was the public face of a technological revolution.

The Firing and Return

In November 2023, Altman was abruptly fired by OpenAI's board of directors, who cited a loss of confidence in his candor. The dramatic episode — reportedly triggered in part by a 52-page memo from co-founder Ilya Sutskever — sent shockwaves through the tech world. Nearly the entire staff threatened to resign in solidarity with Altman. Within five days, he was reinstated, and the board was restructured. The crisis made Altman appear, paradoxically, more indispensable than ever.

Key Contributions

The Hype Question

Altman is both celebrated and scrutinized for his role as AI's chief evangelist. MIT Technology Review has described him as the field's "ultimate hype man," noting that his claims about AI's potential often arrive well before the evidence. He has compared OpenAI's work to the Manhattan Project and predicted that AI will surpass human intelligence in virtually every domain by 2030.

Critics argue that this rhetoric conveniently doubles as a fundraising pitch — each vision of world-changing AGI is also, implicitly, a case for more capital and friendlier regulation. Supporters counter that Altman's predictions have, so far, been directionally right more often than not.

In his June 2025 essay "The Gentle Singularity," Altman laid out a 15-year vision in which AI produces novel scientific discoveries, transforms the labor market, and reshapes the social contract. He predicted that 2026 would bring AI systems capable of generating genuinely original insights — a claim that remains to be tested.

Looking Forward

At 40, Altman is steering OpenAI through its most consequential chapter. The company has crossed $20 billion in annualized revenue, begun testing advertising in ChatGPT, and signaled that enterprise will be its top priority in 2026. It projects revenue exceeding $280 billion by 2030 — a figure that, if achieved, would make it one of the most valuable companies in history.

But the road is not without obstacles. Competition from Anthropic, Google, and others is intensifying. Questions about OpenAI's governance, conflicts of interest, and the gap between benchmarks and real-world reliability persist. And the broader question — whether the AI revolution will deliver on its extraordinary promises or become the most expensive disappointment in tech history — remains open.

What is beyond dispute is that Sam Altman, more than any other individual, has shaped the terms of the debate. Whether history judges him as a prophet or a promoter, the age of AI is, in large part, the age of Altman.

Ilya Sutskever is the rare figure in artificial intelligence who can credibly claim to have shaped the field's most important breakthroughs over the past decade — and then walked away from the empire he helped build. As co-founder and chief scientist of OpenAI, he was instrumental in developing the research that led to GPT and ChatGPT. He established the "scaling ethos" that defined an era. And then, in a move that stunned the industry, he voted to fire Sam Altman, expressed regret, disappeared from public view, and ultimately left to start something entirely new.

Born in 1986 in Nizhny Novgorod (then Gorky) in the Soviet Union, Sutskever immigrated to Israel at age five with his Jewish family, growing up in Jerusalem. He later moved to Canada, where he would earn his bachelor's in mathematics and his master's and PhD in computer science at the University of Toronto — all under the supervision of Geoffrey Hinton, the godfather of deep learning.

The Foundations of Deep Learning

In 2012, Sutskever, along with Hinton and Alex Krizhevsky, created AlexNet — the convolutional neural network that won the ImageNet competition by a dramatic margin and ignited the modern deep learning revolution. It was a watershed moment: the paper demonstrated that neural networks, given enough data and compute, could outperform hand-engineered systems at visual recognition.

After a brief postdoc with Andrew Ng at Stanford, Sutskever joined Google Brain, where he contributed to foundational work including the sequence-to-sequence learning algorithm, TensorFlow, and was among the co-authors on the AlphaGo paper. His time at Google cemented his reputation as one of the most productive and influential researchers in machine learning.

At the end of 2015, Sutskever left Google to co-found OpenAI, where he served as chief scientist for over eight years. During that period, he was credited with establishing the lab's core conviction that scaling — making models bigger with more data and compute — was the path to artificial general intelligence.

Career Timeline

• 1986Born in Nizhny Novgorod, Soviet Union; emigrated to Israel at age 5
• 2005–2013B.Sc. in Mathematics, M.Sc. and PhD in Computer Science at University of Toronto under Geoffrey Hinton
• 2012Co-created AlexNet with Hinton and Krizhevsky — igniting the deep learning revolution
• 2013–2015Research scientist at Google Brain; co-developed sequence-to-sequence learning, contributed to TensorFlow and AlphaGo
• 2015Co-founded OpenAI as chief scientist
• Nov 2023Voted with the board to fire Sam Altman; expressed regret days later
• May 2024Departed OpenAI; announced new venture
• June 2024Co-founded Safe Superintelligence Inc. (SSI) with Daniel Gross and Daniel Levy
• Apr 2025SSI raised $2 billion at a $32 billion valuation — with no product
• Jul 2025Became CEO of SSI after co-founder Daniel Gross departed for Meta
• Nov 2025Gave rare interview declaring the "age of scaling" is ending

The Altman Crisis

In November 2023, Sutskever authored a 52-page memo — drawing heavily on information from then-CTO Mira Murati — that accused Altman of dishonesty, manipulation, and fostering internal division. He submitted the memo to the board and joined in voting to terminate Altman's position as CEO. In an all-hands meeting shortly after, he called it "the board doing its duty."

Within days, however, the backlash was overwhelming. Nearly all of OpenAI's staff threatened to leave. Altman was reinstated, the board was reconstituted, and Sutskever was left in an ambiguous position — reportedly absent from the office and cut off from his team's work. By May 2024, he quietly announced his departure, saying only that he was pursuing something "very personally meaningful."

Safe Superintelligence Inc.

In June 2024, Sutskever launched Safe Superintelligence Inc. alongside Daniel Gross, Apple's former AI lead, and Daniel Levy, an OpenAI researcher. The company's website was little more than a mission statement: one goal, one product — a safe superintelligence.

The fundraising was extraordinary. SSI raised $1 billion by September 2024, then another $2 billion in April 2025, reaching a $32 billion valuation — despite having no product, no revenue, and no public demonstrations. Investors included Greenoaks, Andreessen Horowitz, Lightspeed Venture Partners, DST Global, Alphabet, and Nvidia. The company operates between Palo Alto and Tel Aviv, and is notably one of Google Cloud's largest external customers for tensor processing units.

When Gross departed to join Meta in July 2025 — following a rejected acquisition offer from Mark Zuckerberg — Sutskever stepped into the CEO role, an unusual position for someone who had spent his career as a pure researcher.

Key Contributions

The End of Scaling

In a rare November 2025 interview with Dwarkesh Patel, Sutskever made his most striking public argument yet: the "age of scaling" — the era from roughly 2020 to 2025 when throwing more compute at bigger models reliably produced better results — is ending. The returns from simply adding parameters have diminished. The next breakthroughs, he argued, will come from new training methods and fundamental research insights, not larger GPU clusters.

He pointed to a problem he calls "jaggedness" — the observation that today's models can ace benchmarks but fail at trivially simple tasks in deployment. They oscillate between errors in ways that suggest something deeply incomplete about their understanding. For Sutskever, this gap between benchmark performance and real-world reliability is the central challenge of the field.

His vision for SSI is accordingly different from the competition. Rather than releasing incremental products to fund research — as OpenAI and Anthropic do — SSI focuses entirely on the research problem itself. The product is the intelligence. Safety is not a compliance afterthought but a training philosophy, baked in from the first line of code.

Looking Forward

Sutskever occupies a singular position in AI. He is the scientist who proved that scaling works, and then declared the age of scaling over. He helped build one of the most powerful organizations in technology, then tried to tear it down from within. He raised $3 billion on the promise of a product that doesn't yet exist, and whose timeline he openly admits is uncertain.

In 2025, he received an honorary doctorate from the University of Toronto. In early 2026, he was awarded the National Academy of Sciences Award for the Industrial Application of Science. His former mentor, Geoffrey Hinton, has publicly supported his stance on AI safety.

Whether SSI delivers on its extraordinary promises remains an open question. But Sutskever's bet is clear: the future of intelligence belongs not to whoever has the most compute, but to whoever makes the next fundamental discovery. In a field defined by scale, he is wagering everything on ideas.

Andrej Karpathy is one of those rare figures in technology who moves fluidly between the worlds of cutting-edge research, large-scale product engineering, and public education. Born in Bratislava, Czechoslovakia, in 1986, he emigrated with his family to Toronto at the age of fifteen. What began as a quiet immigrant story would eventually lead him to the founding team of OpenAI, the helm of Tesla's AI division, and a new venture that aims to reinvent how the world learns.

His early internet presence was unassuming: a YouTube channel called badmephisto, where he posted Rubik's cube tutorials that racked up millions of views and were used by speedcubers around the world. It was a hint of what would become his defining trait — an instinct for making complex things accessible.

Academic Foundations

Karpathy earned dual bachelor's degrees in Computer Science and Physics at the University of Toronto in 2009, followed by a master's at the University of British Columbia, where he worked on physically simulated figures under Michiel van de Panne. He then moved to Stanford for his PhD, working under Fei-Fei Li at the Stanford Vision Lab, with additional mentorship from luminaries like Andrew Ng, Daphne Koller, and Sebastian Thrun.

During his PhD, his research focused on convolutional and recurrent neural networks applied at the intersection of computer vision and natural language processing — work that contributed to early breakthroughs in image captioning. Along the way, he squeezed in internships at Google Brain, Google Research, and DeepMind.

Perhaps his most lasting academic contribution was pedagogical. In 2015, he designed and became the primary instructor of Stanford's CS 231n: Convolutional Neural Networks for Visual Recognition — widely regarded as the course that introduced deep learning to an entire generation of practitioners. Enrollment exploded from 150 students in its first year to 750 by 2017.

Career Timeline

• 2006Launched YouTube channel badmephisto with Rubik's cube tutorials (9M+ views)
• 2009–2011B.Sc. at University of Toronto; M.Sc. at University of British Columbia
• 2011–2015PhD at Stanford under Fei-Fei Li; internships at Google Brain, Google Research, DeepMind
• 2015Created CS 231n at Stanford — became one of the university's most popular classes
• 2016Co-founded OpenAI as a founding member
• 2017–2022Senior Director of AI at Tesla, leading the Autopilot Vision team
• 2023Briefly returned to OpenAI; launched "Neural Networks: Zero to Hero" series
• 2024Founded Eureka Labs, an AI-native education company
• 2025Coined "vibe coding"; published influential LLM Year in Review

OpenAI & Tesla

In 2016, Karpathy became a founding member of OpenAI. He left the following year to join Tesla as Senior Director of AI, where he led the computer vision team responsible for Autopilot. At Tesla, his work centered on processing massive amounts of real-world visual data in real time — a challenge at the intersection of deep learning research and industrial-scale deployment.

He departed Tesla in 2022, briefly returning to OpenAI in 2023, before ultimately striking out on his own.

The Educator

If Karpathy is known for any single quality, it is his ability to explain. His YouTube channel has become a canonical resource for understanding large language models from the ground up. His "Neural Networks: Zero to Hero" series walks viewers through building neural networks from scratch.

This commitment to education culminated in the founding of Eureka Labs in July 2024 — an "AI-native school" that pairs human expertise with AI teaching assistants. The company's first product, LLM101n, is an undergraduate-level course designed to teach students how to train their own AI model.

Key Contributions

A Sober Voice in 2025

In a widely shared October 2025 interview with Dwarkesh Patel, Karpathy argued that AGI remains at least a decade away and cautioned that many companies are overstating AI agent reliability. His year-end review for 2025 became a landmark document, tracing the rise of RLVR and introducing his memorable framing of LLMs as "summoned ghosts" rather than "evolved animals."

Looking Forward

At 39, Karpathy sits at a unique crossroads. He has helped build two of the most consequential AI organizations in history, trained thousands of students, and now runs a startup aimed at democratizing education through AI. His insistence on building things from scratch — embodied in his mantra "if I can't build it, I don't understand it" — continues to set him apart.

Whether through his YouTube lectures, his sharp commentary on X, or the courses emerging from Eureka Labs, Andrej Karpathy remains what he has always been: a teacher first, and one of the most trusted voices navigating the uncertain terrain between where AI is and where it's going.