General Intuition's $2.3B Valuation: AI Agents From Gaming Data

General Intuition just raised another $320 million at a $2.3 billion valuation — less than a year after its initial $134M round — bringing total disclosed funding to $454 million. The company trains AI agents on billions of gameplay clips from Medal’s 10 million monthly users, using embedded action labels (button presses, timing) rather than video alone. Khosla Ventures led the latest round, with participation from Jeff Bezos, Eric Schmidt, and researchers from Google DeepMind and MIT.

I covered their first raise last month. At the time, the pitch felt interesting but speculative. Now there’s a working robot, a jobs platform, and an API arriving this summer. Here’s what actually changed.

Why are top investors pouring money into a company that learns from Fortnite?

When the same people who backed OpenAI and Anthropic write checks for a startup that trains on video game clips, it tells you something about where the industry thinks the next breakthrough lives — and it’s not in bigger language models.

The new round was led by Khosla Ventures, with General Catalyst, Jeff Bezos, Eric Schmidt, Nico Rosberg, and researchers from Google DeepMind and MIT joining in. The vast majority of the capital goes toward scaling compute through a deal with CoreWeave, focused on pre-training the next version of their model. A smaller portion is earmarked for making their API available to outside developers by end of summer.

The investor list matters more than the dollar figure. These aren’t gaming VCs chasing the next battle royale. They’re the people who shaped the foundation model era, and they’re betting that spatial intelligence — models that understand physical environments — is the next frontier.

How do video game clips turn into a walking robot?

During a visit to General Intuition’s New York R&D floor, TechCrunch reporters watched an AI agent play Fortnite for 100 hours straight. Then they watched the same model power a quadrupedal robot that walked around the office using a single camera. It bumped into chairs like a toddler and fine-tuned itself to a new environment in just eight minutes of real-world data.

The core idea: the same brain that learns to navigate a virtual world can transfer that understanding to the physical one. This is what AI agents becoming employees actually looks like — not chatbots drafting emails, but models that grasp space, time, and cause-and-effect well enough to walk a robot through a room.

Most competitors try to infer actions from video alone. General Intuition’s edge is the action labels baked into Medal’s gameplay clips — records of exactly what buttons a player pressed and when. CEO Pim de Witte argues this helps the model distinguish “self” from “environment,” giving it a richer understanding of causality. Watching someone drive versus feeling the steering wheel yourself.

What is Nerve, and why does it matter for gamers?

Nerve is a jobs marketplace General Intuition recently launched. It lets gamers earn money using their existing hardware, starting with data labeling tasks and eventually moving toward robot teleoperation and higher-skilled work.

The logic tracks. Medal’s user base is precisely the generation most exposed to AI-driven displacement. De Witte wants them to have a stake in what comes next. Instead of extracting training data from gamers and building models that might replace them, Nerve creates a pathway for those same gamers to participate in the AI economy.

It’s also a shrewd business move. As the models get more capable, General Intuition will need human oversight for edge cases, safety validation, and real-world data collection. A workforce that already understands spatial environments — because they’ve spent years navigating virtual ones — is a competitive moat that’s hard to replicate.

Is General Intuition drawing ethical lines around military use?

Yes, explicitly. De Witte spent three years in humanitarian work, including time with Doctors Without Borders, and has drawn a firm boundary: no agents built to harm humans. Military applications are prohibited.

This matters because the same technology that navigates a robot through an office could theoretically navigate one through a battlefield. As Silicon Valley grows more comfortable with defense contracts, General Intuition’s stance stands out — especially given that investors like Khosla and Bezos have their own positions on defense tech.

The company’s chief of staff, Brianna Martin, publicly quit Palantir over its work with ICE. De Witte brought her on specifically because of that decision. It signals the kind of company General Intuition wants to be, not just what it wants to build.

Can the API justify a $2.3 billion valuation?

That’s the real question. The API launch this summer will test whether third-party developers can build useful applications on top of General Intuition’s world model — and whether the model holds up across diverse environments, not just a controlled office demo.

The company has a handful of customers in gaming, simulation, and robotics today. But the ecosystem play is what makes the valuation make sense. De Witte’s vision is to be a model provider, not an end-product company. “We’re not gonna build a self-driving car company,” he said. “We’re gonna make it 10 times easier for the next person to build a self-driving car company.” Same playbook OpenAI ran with language, applied to spatial intelligence.

If you’re building with AI tools — whether it’s automating your business or creating content with AI — General Intuition’s approach suggests the next wave of capable AI agents won’t come from scaling language models. They’ll come from models that understand the physical world. The API launch will tell us how close that future actually is.

For anyone watching the AI space, this is one to track. If you want to stay on top of what’s actually working in AI tools, I cover the developments that matter — the ones that ship, not just the ones that make headlines. And if you’re new to AI agents, start here — I break down the concepts without the jargon.