Alexandr Wang

Who they are

Alexandr Wang is founder of Scale AI and now Chief AI Officer at Meta — dropped out of MIT in 2016 to build Scale AI through Y Combinator, then sold Meta a 49% stake for $14.3 billion in June 2025.

Person

Wang grew up in Los Alamos, New Mexico, winning the UNM-PNM Math Contest in high school before enrolling at MIT — then dropped out after his first year in 2016 to found Scale AI through Y Combinator at 19. Before Scale, he'd interned at Quora, done a stint at Addepar, and worked at Hudson River Trading, giving him early exposure to consumer product, fintech, and high-frequency trading infrastructure. He co-founded Scale AI with Lucy Guo, building it into the dominant data labeling and LLM evaluation platform for AI — serving OpenAI, Google, Microsoft, and Meta, and backed by Founders Fund, Accel, Index, Coatue, Y Combinator, and angel investors including Greg Brockman, Kevin Systrom, Mike Krieger, Adam D'Angelo, Drew Houston, Justin Kan, and Nat Friedman. In June 2025, Meta acquired a 49% nonvoting stake for $14.3 billion at a $29 billion valuation, and Wang moved to Meta to lead its superintelligence efforts as Chief AI Officer, remaining on Scale AI's board. The through-line is an early, intense bet on AI infrastructure as the foundational layer of the AI economy — pursued with unusual policy engagement: he testified before the House Armed Services Subcommittee in 2023 on AI policy, wrote letters to President Trump on US AI dominance, and has spoken at CSIS, Davos, TED, and the a16z podcast on national security AI and US-China competition. He posts on LinkedIn about AI leadership, data infrastructure, and AI regulation — practical and geopolitically charged, not theoretical.

Company

The most recent development at Scale AI is the March 2026 launch of Scale Labs, an expanded research division building on its Safety, Evaluation, and Alignment Lab (SEAL) from 2023. That followed the seismic June 2025 restructuring: Meta took a 49% nonvoting stake for $14.3 billion, valuing Scale AI at $29 billion, and Wang departed the CEO role to lead Meta's superintelligence efforts — leaving Scale AI under new leadership with Wang remaining on the board. The aftermath has been turbulent: some major clients including Google and OpenAI reduced or cut ties over competitive concerns, and Scale AI's CFO stated in November 2025 that the company has $1 billion on the balance sheet and is not looking to raise more funds soon, describing a focus on recruiting and new strategic directions. Scale AI also launched new products in 2026 — Scale Launch (AI data management), Scale Validate (model testing), and Scale Rapid (data annotation automation) — and is actively expanding into government and enterprise AI, including a US Department of Defense hub in St. Louis and a five-year partnership with the Qatari government signed in February 2025.

Market

Scale AI sits at the center of the AI data infrastructure market — data labeling, RLHF services, and LLM evaluation — competing against Labelbox, SuperAnnotate, V7, Snorkel AI, and Encord, while also facing commoditization pressure as hyperscalers build internal capabilities. The US-China AI rivalry is a defining industry dynamic: Wang has publicly positioned Scale as a national security asset, and Meta's investment has drawn antitrust scrutiny and accelerated client diversification away from Scale toward competitors. AI regulatory fragmentation and escalating compute costs add further complexity to the market Scale operates in.

Network

Wang co-founded Scale AI with Lucy Guo and built a cap table that reads like an early-2010s SF founder network: Greg Brockman, Kevin Systrom, Mike Krieger, Adam D'Angelo, Drew Houston, Justin Kan, and Nat Friedman are all investors. His June 2025 move to Meta puts him directly under Mark Zuckerberg, whom he joined to lead Meta Superintelligence Labs.

Lucy Guo· Co-founder, Scale AI
Mark Zuckerberg· CEO, Meta
Greg Brockman· Investor, Scale AI
Kevin Systrom· Investor, Scale AI
Adam D'Angelo· Investor, Scale AI
Nat Friedman· Investor, Scale AI
Justin Kan· Investor, Scale AI

How they likely show up

Founded Scale AI at 19 and ran it as CEO for nearly a decade → operates with high conviction early, comfortable holding a position before the market confirms it.
Testified before Congress, wrote open letters to the President, and spoke at Davos and CSIS on national security AI → treats policy and geopolitics as a core operating lever, not a PR function.
Dropped out of MIT after one year to found Scale → prioritizes speed of execution over credentialing; likely impatient with institutional timelines.
Built Scale AI from data labeling through LLM evaluation to a $29 billion company → thinks in infrastructure layers, not point solutions.
Posts on LinkedIn about AI leadership, regulation, and US-China competition — not product announcements → public voice is strategic and thesis-driven, not performative.

Conversation tips

→ He has strong, publicly stated views on US-China AI competition — engage with the substance of that thesis, not just acknowledge it exists.
→ His congressional testimony and CSIS appearances show he takes the policy dimension seriously; questions that treat regulation as a technical afterthought will land flat.
→ He wrote a blog post titled 'DO TOO MUCH: How to be a leader' — if you reference it, be specific about the argument, not just the title.
→ He's now inside Meta running superintelligence, not running Scale day-to-day — frame questions around what he's building at Meta, not what Scale is doing post-deal.
→ His math competition background and early Hudson River Trading exposure suggest he responds well to precision and quantitative framing over hand-waving.

Toolbox

Openers

Open on Scale Labs — Scale AI launched an expanded research division in March 2026 building on its SEAL work; asking what problems Scale Labs is scoped to solve that Scale's core product couldn't is a direct window into where he thinks AI evaluation breaks down.
Reference his 2023 House Armed Services testimony — he argued AI is a national security asset before most CEOs were doing that publicly; it's a point of genuine conviction, not a talking point, and the specifics reward a prepared opener.
Bring up the post-Meta client departures — Google and OpenAI reportedly reduced ties after Meta's June 2025 investment; how Scale AI rebuilds that client base (or doesn't need to) is the live strategic question on the table.

Discovery questions

You've argued the US has to win the AI war — now that you're inside Meta running superintelligence, what does winning actually look like in practice versus in policy testimony?
Scale Labs builds on SEAL — where does model evaluation currently fail that a dedicated research division can fix, and what does that mean for how frontier labs actually use Scale?
The $14.3 billion Meta deal restructured Scale AI's competitive position overnight, with some major clients pulling back — what's the theory of the company now, post-deal?

Avoid

Don't treat his policy and national security positions as talking points — he has testified before Congress, briefed the White House, and written directly to the President on these topics, so surface-level agreement without substantive engagement will read as unprepared.

Make it yours