AI Startup General Intuition Bets Video Game Data Can Unlock Human-Like Machine Intuition, Raises $320m at $2.3bn Valuation

The race to build artificial intelligence systems capable of understanding and interacting with the physical world has produced a growing number of contenders. Yet few are pursuing a strategy as unconventional as General Intuition, a New York-based startup that believes the key to creating more capable AI agents lies not in warehouses of robots or fleets of autonomous vehicles, but in hundreds of millions of hours of video game footage.

That vision has attracted some of the biggest names in technology and venture capital. General Intuition announced a $320 million funding round that values the company at $2.3 billion, bringing its total disclosed funding to $454 million after a $134 million raise at launch in October last year.

The round was led by Khosla Ventures and included participation from General Catalyst, Amazon founder Jeff Bezos, former Google chief Eric Schmidt, former Formula One champion Nico Rosberg, as well as researchers from Google DeepMind and MIT.

At the center of the company is 31-year-old co-founder and chief executive Pim de Witte, who argues that today’s AI systems remain fundamentally limited because they lack an intuitive understanding of how actions affect the world around them.

According to TechCrunch, that philosophy was on display at the company’s research facility, where an AI agent had reportedly been playing a Fortnite-like game continuously for more than 100 hours.

“Our agent has been playing for 100 hours straight,” said chief product officer Kent Rollins.

The demonstration was more than a gaming experiment. According to the company, the same underlying model controlling the virtual character was also guiding a quadrupedal robot navigating the office.

“The same brain powering the agent playing the game is powering the robot,” de Witte said.

The robot relied on a single camera as its primary sensor and roamed through the office in an exploratory mode, occasionally bumping into chairs and trash bins as it learned to understand its surroundings.

According to data analyst Josh Duplantis, the robot required only eight minutes of real-world robotics data for fine-tuning, with the training data collected outdoors rather than in the office environment it was navigating. That ability to transfer knowledge from gaming environments to simulations and then into physical machines forms the core of General Intuition’s long-term strategy.

Unlike many AI developers that focus on text-based large language models, the startup is building what researchers call a “world model” — an AI system designed to understand cause and effect, movement, space, and time.

The company’s roots help explain that approach.

General Intuition emerged from de Witte’s earlier company, Medal, a platform that allows gamers to upload and share gameplay clips. Over the years, Medal accumulated hundreds of millions of hours of gaming footage, creating a unique data asset.

But de Witte argues that the real value lies not in the video itself but in the metadata. Most gameplay clips contain records of every button pressed by players and the exact timing of those actions. That information allows AI systems to observe not only what happened, but why it happened.

“Most competitors are trying to infer actions from video alone,” de Witte argues, while General Intuition has access to actual human decision-making records embedded in the data.

“We view this as just the next stage of future pre-training,” de Witte said. “We have a single model that can respond to Fortnite information on the screen and take action, but also to real-world dynamics in a way that an LLM could never.”

The company’s internal simulation platform serves as what executives call “the gym” — a training environment where AI models learn how to interact with dynamic worlds.

According to de Witte, exposure to vast amounts of gameplay allows the model to learn physical rules and spatial relationships naturally. The system has demonstrated an understanding that walls are obstacles, ladders can be climbed, and shadows change as the sun moves across a scene.

For General Intuition, however, the simulation itself is not the end product. The company ultimately intends to commercialize the underlying agentic AI model, which it believes can generalize across gaming, robotics, autonomous systems, industrial automation, and other real-world applications.

“We’re not gonna build a self-driving car company,” de Witte said. “We’re gonna make it 10 times easier for the next person to build a self-driving car company.”

Today, the company already has customers across gaming, robotics, and simulation markets. The startup also sees opportunities in industrial automation, digital twins, robotics testing, and hazardous-environment operations.

According to de Witte, the technology can already operate any system controllable through familiar interfaces.

“It works on anything that you can control using a game controller or a keyboard mouse,” he said.

Much of the new capital will be directed toward expanding computing infrastructure.

General Intuition has partnered with CoreWeave and plans to invest heavily in training larger versions of its model. Part of the funding will also support a broader rollout of the company’s API later this year.

For investors, the appeal extends beyond the technology itself.

Vinod Khosla, whose firm led the round, sees the company’s proprietary dataset as a potentially decisive advantage.

“If you look at LLMs, when reasoning emerged, it was a quantum leap,” Khosla said. “In world models, I think the quantum leap is the emergence of intuition in the AI, a human intuition-like capability. The human action data and reaction data you have in games is the key part to the emergence of intuition.”

That unique data asset has reportedly attracted acquisition interest from major AI laboratories, but General Intuition says it has rejected multiple offers.

According to company executives, the goal is not to become an acquisition target but to build foundational AI infrastructure that could support an entire ecosystem of applications.

“At this point, it would be a data acquisition, which is sort of uninteresting,” Khosla said.

The company’s ambitions arrive as technology giants, including OpenAI, Anthropic, Google DeepMind, and Meta, pursue their own efforts to create more capable AI agents that can act autonomously in both digital and physical environments.

Yet General Intuition is also attempting to distinguish itself through its ethical framework. De Witte, who previously spent several years working in humanitarian efforts, including with Doctors Without Borders, says the company will not pursue lethal military applications.

“We don’t want to be an escalatory part of the system,” he said.

“Let’s say I were to come out and say, ‘We’re doing lethal autonomy.’ What do you think would happen in other countries?”

He added that he remains supportive of applications such as search-and-rescue missions.

The company’s culture also reflects its European roots. De Witte, who is Dutch, has recruited staff whose views align with his approach to responsible AI development.

“I don’t know why Silicon Valley does what it does,” he said. “There’s a reason I’m not there.”

The startup is also thinking about the economic consequences of AI. Recognizing concerns about job displacement, General Intuition recently launched a platform called Nerve, a marketplace designed to allow gamers to earn income through data-labeling work, robot teleoperation, and other AI-related tasks.

De Witte believes gaming communities represent one of the populations most exposed to future AI disruption and wants them to benefit from the technology’s growth.

Looking ahead, General Intuition plans to use customer deployments to create a self-reinforcing data flywheel. The company intends to select partners not only for commercial reasons but also for the unique real-world data that those deployments can generate.

“We’ll pick customers where we can diversify the embodiments that this generalized foundation model is serving as the backbone for,” de Witte said.

“So we’re going to prioritize picking customers on whether they can offer real-world data that’s going to be interesting and useful to move the needle on research. And if they’d have an agile internal team where we can be real embedded partners and learn from each other.”

However, the strategy remains unproven. Industry researchers broadly agree that transferring capabilities learned in simulations into the physical world remains one of AI’s most difficult challenges.

Even Khosla acknowledges that whether simulation-to-reality learning can work at scale remains an open question.

Still, with nearly half a billion dollars in funding, access to one of the world’s largest repositories of human gameplay behavior, and growing interest in AI agents capable of understanding the physical world, General Intuition has emerged as one of the more closely watched startups in the next phase of artificial intelligence development.

Its wager is that before machines can truly understand reality, they may first need to learn how humans play.