Alexandr Wang

Los Alamos, New Mexico is where the United States built its nuclear weapons program. The town exists specifically because of what happens when you concentrate the country's best physicists in one place and give them an impossible problem. Alexandr Wang grew up there. His parents - Chinese immigrants - were among those physicists. The lesson of Los Alamos wasn't just science. It was that concentrated talent, pointed at a hard problem, changes history.

Wang absorbed this early. By 13, he was competing nationally in math olympiads. By 15, he placed 5th in the USA Mathematical Talent Search. He qualified for the Mathematical Olympiad Summer Program, made the USA Physics Team, and was a USA Computing Olympiad finalist. These aren't participation trophies. They're the competitions where roughly 0.01% of participants advance.

Then came Quora. As a teenager - while most kids were grinding homework - Wang was writing code professionally for one of Silicon Valley's most technically demanding startups. He went on to intern at Hudson River Trading, the quantitative trading firm known for its brutally selective hiring. He enrolled at MIT. Then he left.

In 2016, the AI world was obsessed with architectures, algorithms, and compute. The race to build smarter neural networks was loud. Wang looked at the same landscape and noticed something quieter: every model, no matter how clever, needed labeled data to learn from. And the supply of good labeled data was shockingly thin.

He co-founded Scale AI with Lucy Guo through Y Combinator. The pitch was simple and specific: be the infrastructure layer that turns raw information into the structured training data AI companies needed to actually work. Not glamorous. Not a consumer app. Not a chatbot. The unsexy plumbing beneath the AI revolution.

It turned out to be the most important plumbing in the industry.

By 2019, Founders Fund - Peter Thiel's firm - led Scale AI's $100M Series B. The company hit unicorn status. By 2021, at a $7.3B valuation, Wang became what Forbes certified as the world's youngest self-made billionaire. He was 24. Most people his age were still doing first-round interviews.

The client list that Scale AI assembled is essentially a roll call of the entities that matter in AI: OpenAI, Google, Microsoft, Meta, General Motors, and the U.S. Department of Defense. Not a niche player. Not one category. All of them.

Scale AI is easy to misunderstand from the outside. "Data labeling company" undersells it by roughly $28 billion. By 2025, Scale had built the infrastructure through which over 15 billion human decisions flowed into AI model training - every click, annotation, and judgment that taught the models what "correct" looks like. It paid over $1 billion to human contributors worldwide in the process.

The defense portfolio is where Scale gets genuinely unusual. "Donovan" - Scale's AI decision support platform - became the first commercial AI system deployed on a U.S. Army classified network in May 2023. Wang was 26. The platform helps military analysts process intelligence and evaluate options faster than traditional methods allow. The Pentagon's Chief Digital and Artificial Intelligence Office selected Scale to test large language models for military applications. A $250M federal contract came in January 2022. The Thunderforge contract in March 2025 extended Scale's reach into AI-driven military asset planning.

When Wang testified before the House Armed Services subcommittee in July 2023, he wasn't a tech founder explaining his product to politicians. He was a stakeholder at the intersection of commercial AI and national security, arguing for a faster, more aggressive American posture on AI development.

Wang is not shy about where he stands. In June 2024, when much of corporate America was still managing DEI programs carefully, Scale AI publicly replaced its diversity hiring framework with "MEI" - Merit, Excellence, Intelligence. It was a deliberate, high-profile move that came months before similar pivots became fashionable in Silicon Valley. Wang framed it plainly: he wants the most capable people, and he's willing to say so out loud.

His China position is more pointed. Wang has argued consistently that the United States must outcompete China in AI - not coexist with it, not collaborate with it, beat it. He attended Trump's second inauguration in January 2025 and published an open letter addressed to the new administration. The thesis: America must win the AI war, or the strategic consequences will be severe. At the February 2025 Web Summit, he made the same case to an international audience that largely disagreed. He made it anyway.

The diplomatic access Wang commands is remarkable for a 28-year-old tech founder. In February 2025, in a single month, he met with UK Prime Minister Keir Starmer, Indian Prime Minister Narendra Modi, French President Emmanuel Macron, and House Speaker Mike Johnson. Those conversations weren't about product partnerships. They were about AI strategy at the national level. The category Wang occupies is somewhere between industry executive and geopolitical actor.

His willingness to stake public positions - on hiring, on China, on defense - sets him apart from peers who prefer ambiguity. Whether you agree or not, Wang is legible. You know what he thinks.

In June 2025, Mark Zuckerberg made a move that looked less like an acquisition and more like a recruitment. Meta invested $14.8 billion in Scale AI for a 49% non-voting stake - one of the largest deals in AI history. As part of the arrangement, Wang stepped down as Scale AI CEO and took on a new role: Chief AI Officer at Meta Platforms. Jason Droege, formerly of Uber, became Scale AI's new CEO.

The structure is telling. Meta didn't absorb Scale. Wang retained his equity and Scale retained its independence. What Zuckerberg bought wasn't a company - he bought Wang's continued involvement in both enterprises, and Scale's position as the data and training infrastructure underpinning Meta's AI efforts. At Meta, Wang now leads the Superintelligence Labs, the division building toward AGI.

It's a strange trajectory to plot: Los Alamos kid, math olympiad competitor, teenage Quora programmer, MIT dropout, data company founder, youngest self-made billionaire, congressional witness, diplomatic counterpart to heads of state, Meta's chief AI officer. But each step follows logically from the last if you know the through-line: Wang has always been drawn to the problem that determines the outcome. In 2016, that was training data. In 2025, it's intelligence itself.

His parents built weapons at Los Alamos because nuclear physics was the leverage point of their era. Wang is doing something structurally similar - and he'd probably appreciate the comparison.