I think the first AGI lab is already running. It just doesn't call itself an AI company.

In 2025, Jane Street made $39.6 billion with 3,500 people. JPMorgan, with 317,000, made about the same. The top four quant firms — Jane Street, Citadel Securities, Hudson River Trading, XTX — pulled in over $65 billion last year. OpenAI, Anthropic, and xAI together did about $20 billion.

Quant trading is the most successful AI deployment in history. Nobody calls it that. But that's what it is.

I want to make a sharper claim than "markets are a good AGI benchmark." The architecture for AGI already exists. It has been running in Long Island and Manhattan since 1988. Renaissance built it. DE Shaw built it. Two Sigma extended it. They never called it AGI architecture because they had no reason to. They were too busy compounding.

The first AGI lab is whoever ports that architecture to a fast machine meta-learner.

I spent eight years running a quant firm. The most important thing that experience teaches you is that strategies are disposable. Every working strategy you ever build will stop working, on a timeline somewhere between a year and a decade. Regimes change. Other people reverse-engineer your edge. Microstructure shifts under you. The strategies you have now will be dead in three years.

The firm survives only if it produces new strategies faster than the old ones decay.

What every quant firm that has ever mattered actually built is not alpha. It's the system that generates alpha, retires it, generates new alpha, and improves the alpha-generation process itself. The strategies are outputs. The machine is the asset.

Renaissance has been doing this since 1988. Their reported equity curve compounds at roughly 66% a year before fees. Across thirty years. Across every major market regime. The numbers aren't independently audited and Renaissance hasn't officially confirmed them, but they're widely cited in academic finance. Brad Cornell at UCLA wrote in 2020 that the returns "stretched explanation to the limit." To Institutional Investor he was blunter: like the sun rising in the west.

What does it mean to compound at 66% for thirty years across regimes your strategies were never trained on?

It means the firm isn't pattern-matching. It's reasoning. The strategies in 1995 were different from 1988. The strategies in 2008 were different from 1995. The strategies now are different again. The thing producing them has been compounding capability against an environment that fights back.

Renaissance solved a version of the AGI problem. They solved it for one cognitive primitive — forecasting under uncertainty against adversaries — on one substrate, financial markets, using human researchers as the meta-learner.

The architecture is the answer. The substrate is the answer. Only the meta-learner is wrong.

The architecture has a name. Recursive self-improvement. A system that improves the system that produces capability, faster than the capability it produces saturates.

I.J. Good described it in 1965. Yudkowsky formalized it. Aschenbrenner forecast it in Situational Awareness. The CEOs of Anthropic and DeepMind talked about it openly at Davos this year. Anthropic has reportedly discussed spending over a billion dollars on RL environments to enable it. ICLR 2026 hosted a workshop on it.

The frontier labs are racing toward recursive self-improvement using synthetic environments built inside research labs. Coding gyms. Math problem sets. Hand-built RL setups where verifier models grade verifier models. These environments have a fundamental problem: they're constructed by humans who have to decide, in advance, what the system should optimize against. When the system gets good enough, it games the construction. The proxy decouples from reality. You don't know when this happens until it does.

Markets don't have this problem. The reward signal isn't constructed. It's generated by every other participant pricing reality in real time. The signal can't be Goodharted at the substrate level — it can be gamed at the margins, but that's a different problem — because the participant set adapts faster than any single optimizer can stabilize an exploit. The substrate has been adversarially robust for centuries. It will keep being adversarially robust regardless of what we train on it.

The frontier labs are building the meta-learner. They're missing the substrate. The quant firms have the substrate. They're missing the meta-learner.

The first AGI lab is whoever assembles both.

Hassabis already knew.

In 2016, according to Sebastian Mallaby's The Infinity Machine, Hassabis assembled a secret hedge fund operation inside DeepMind. Twenty researchers. High-frequency trading algorithms. An exploratory partnership with BlackRock. Mallaby reports, sourced to a person familiar, that Hassabis wanted to compete with Renaissance.

Google killed it.

Hassabis had spent a decade picking environments and letting agents extract everything they could teach. AlphaGo. AlphaZero. AlphaStar. By 2016 those environments were exhausted. The agents had won at superhuman levels. There was nowhere obvious left to go.

He picked markets. He did this while running the world's leading AI lab. That tells you what he thought came next.

DeepSeek already shipped.

In January 2025, DeepSeek released an open-weight LLM that broke the AI industry's pricing assumptions and triggered a multi-trillion-dollar selloff in US tech stocks.

DeepSeek's parent company is High-Flyer. High-Flyer is a Hangzhou quant fund. Their first deep-learning trades went live in 2016. By 2017 most of their trading was AI-driven. They built about $14 billion in AUM training models on tick data. In 2023 they pointed the same compute stack at "pursuing AGI".

This is the prototype, running in the wild. A quant firm that already knew how to run big GPU clusters, clean adversarial data, and close tight feedback loops at scale — pointing that capability at next-token prediction instead of next-price prediction.

The institutional machine is converging from both sides. Sam Bankman-Fried, a Jane Street alum, provided some of Anthropic's initial funding. Jane Street has participated in later rounds. Scale AI's founder came from Hudson River Trading. Surge AI's founder came from a Thiel hedge fund. OpenAI's research head spent time as a quant. HRT's AI head came from DeepMind. xAI publicly tested its frontier model on live stock markets in 2025; Musk tweeted about using it to "pay for all those GPUs".

FT Alphaville described quants and AI labs in January as converging on the same institutional machine: large-scale learning systems attached to balance sheets.

The convergence is real. The endpoint is whoever gets there first.

The natural objection: if Renaissance already has the architecture, why don't they have AGI?

Because the architecture is necessary, not sufficient. Renaissance's meta-learner is human researchers. Humans are slow. They generalize narrowly. They can only hold so many regimes in attention at once. Renaissance compounds capability against markets at the rate human research teams compound — fast enough to produce the most successful trading firm in history, nowhere near fast enough to produce general intelligence.

The other objection: Renaissance's strategies are narrow. Statistical edge-stacking, not reasoning. This is true and not load-bearing. Narrow strategies are what their meta-learner produces because their meta-learner is optimizing for capital efficiency on liquid markets. A different meta-learner, optimizing for capability gain instead of capital efficiency, would produce different strategies on the same substrate.

What changes when the meta-learner is a machine?

Three things.

It runs continuously instead of weekly. Decisions humans make in research meetings happen in milliseconds. Hypothesis generation parallelizes across thousands of streams.

It generalizes more broadly. A human researcher who learns something about Indian banking stat-arb has to consciously port that lesson to other domains. A model whose representations encode the underlying patterns ports automatically.

It improves itself. Renaissance's researchers got better over thirty years through hiring, training, and culture. A model-based meta-learner gets better through gradient updates against the same substrate that produced the alpha. The improvement loop is closed.

This is the missing piece. Renaissance has been running the architecture with a slow meta-learner. The frontier labs have been building fast meta-learners against the wrong substrate. The first lab to put a fast meta-learner inside the right substrate is the first AGI lab.

Frontier AI is trained for chat. The training objective shapes the actor. A model optimized for human approval on chatbot leaderboards is, by construction, the wrong shape for an adversarial environment with delayed feedback and skin in the game.

This is fixable. What's missing is the training scaffold: the environments, the reward signals, the data, the meta-learning loops that turn a chat-trained model into an alpha-generation system.

That's engineering, not basic research. The 0.42 Sharpe gap between frontier models and human experts on Indian equity data is not a fundamental limit. It's a capability that hasn't been trained yet. The lab that trains it first is the lab that wins.

Picture the first AGI lab.

100 humans. 10,000 agents. A vertical equity curve. No public-facing product. No API. Few or no papers. Compute bills measured in nine figures and growing. Payroll measured in eight figures and stable. The company isn't raising capital because it doesn't need outside money. It's closer in shape to Renaissance than to OpenAI.

The agents do most of the work. They generate hypotheses, design experiments, run simulations, deploy strategies, monitor live performance, retire decaying edges, and propose architectural changes to themselves. The humans pick problems, design the meta-environment, audit decisions the agents flag as uncertain, and intervene when something looks wrong.

The agents trade against the world's deepest adversarial substrate. The substrate generates the reward signal. The reward signal funds more compute. More compute trains more capable agents. More capable agents extract more reward.

The loop closes.

This is recursive self-improvement on the cleanest substrate that exists, with an economic flywheel attached. Every other AI lab is funding-bottlenecked by external capital. This lab is funded by the substrate it's training against.

It doesn't look like an AI lab from the outside. It looks like a quant firm.

There are three candidates for who builds it.

A frontier AI lab. OpenAI, Anthropic, Google DeepMind, xAI. They have the meta-learner. They lack the substrate, the trading scaffolding, and the corporate freedom to pursue this seriously. Hassabis's 2016 attempt is the cautionary tale. The problem is permission.

An existing quant firm. Renaissance, Citadel, Two Sigma, DE Shaw. They have the substrate and the scaffolding. They lack the meta-learner architecture and the research culture to produce it. They could license a frontier model off the shelf, but it would be optimized for chat, not for their substrate. The problem is research culture.

A new entity built specifically for this. No models, no scaffolding to start. Structural freedom to build both. No corporate parent to veto. No legacy research culture to fight. The problem is that nobody has done it yet, which means the playbook is unwritten.

I'd bet on the third. The first is the least likely, despite having the most resources, for the same reason Hassabis got stopped in 2016.

Here is the prediction.

By 2030, the highest-revenue AI company in the world won't call itself an AI company.

This is already partly true. Jane Street alone earned more in 2025 than every frontier AI lab combined. The top four quant firms together earned more than the entire AI industry. The gap isn't closing.

If I'm wrong, it will look like this: the most valuable AI company in 2030 is OpenAI or Anthropic or DeepMind, and their primary revenue still comes from licensing models or API access. That's the version of the future I'm betting against.

The capacity wall is real. Medallion is capped at $10–15 billion because beyond that, its own size moves the market against itself. The first AGI lab will hit a similar wall. But the wall is higher than it sounds. A fund at Medallion's capacity, compounding at 30% net, produces enough capital to fund any research agenda you want. You don't need to buy the world. You need enough to never need outside money again.

The team that solves general reasoning is the team that can deploy it in the highest-leverage environment available. The highest-leverage environment is markets. The architecture for compounding capability against that environment was invented in 1988 in Long Island and is still running.

Whoever does it first won't tell you. They'll just compound.

When we look back, I suspect the moment AGI arrived won't have been marked by a paper.

It'll have been a pattern in the tape. A silent, vertical equity curve.