General Intuition Aims to Raise Capital at $6 Billion Valuation to Power AI Robotics

The company’s AI model was developed using hundreds of millions of hours of gameplay from General Intuition CEO Pim de Witte’s video game clip-sharing platform, Medal, and trains generalized AI agents to move through time and space, according to the report.
General Intuition plans to use the new capital to invest in additional compute infrastructure and talent to improve its general model, with a focus on robotic embodiments, per the report.
We’d love to be your preferred source for news.
Please add us to your preferred sources list so our news, data and interviews show up in your feed. Thanks!
General Intuition did not immediately reply to PYMNTS’ request for comment.
General Intuition announced an earlier funding round two months ago, saying in a June 25 post on X that it raised $320 million in a Series A round at a $2.3 billion valuation. The company said the round was led by Khosla Ventures, former Google Executive Chairman and CEO Eric Schmidt and Amazon founder Jeff Bezos.
General Intuition announced in an Oct. 25, 2025, post on X that it raised $133.7 million in a seed round.
In the June 25 post, the company said: “General Intuition is the frontier lab for acting in space and time. We build large action foundation models trained on billions of ground truth action-labeled gameplay clips from [17 million] monthly active users on Medal, and push the frontier of world models to generate infinite training environments.”
Several announcements around world models have been made in recent months.
It was reported June 17 that Amazon invested in and launched a partnership with Odyssey, a startup that builds AI systems designed to simulate the physical world. Odyssey’s world models can create 3D environments. The round raised $310 milion and valued Odyssey at $1.45 billion.
Chinese tech giant Alibaba announced June 15 that it unveiled its first suite of AI models for robots. The suite includes Qwen-RobotNav, which helps robots understand how to navigate physical spaces; Qwen-RobotWorld, a video “world model” that allows a robot to predict how a physical scenario will unfold; and Qwen-RobotManip, which enables cross-embodiment training at scale.

Facts Only

* General Intuition’s AI model was developed using gameplay from Medal.
* The model trains generalized AI agents to move through time and space.
* General Intuition plans to invest new capital in compute infrastructure and talent for its general model, focusing on robotic embodiments.
* General Intuition raised $320 million in a Series A round at a $2.3 billion valuation.
* The Series A round was led by Khosla Ventures, Eric Schmidt, and Jeff Bezos.
* General Intuition raised $133.7 million in a seed round in October 2025.
* The company stated it builds large action foundation models trained on billions of ground truth action-labeled gameplay clips from [17 million] monthly active users on Medal.
* Amazon invested in and partnered with Odyssey, a startup that builds AI systems to simulate the physical world.
* Alibaba unveiled robot AI suites including Qwen-RobotNav, Qwen-RobotWorld, and Qwen-RobotManip.

Executive Summary

General Intuition developed an AI model using gameplay data from Medal to train generalized AI agents to navigate time and space. The company plans to use new capital to invest in compute infrastructure and talent, focusing on robotic embodiments for its general model. General Intuition previously announced a $320 million Series A funding round valuing the company at $2.3 billion, led by Khosla Ventures, Eric Schmidt, and Jeff Bezos. Earlier, the company raised $133.7 million in a seed round. The group explicitly stated that General Intuition is a frontier lab for acting in space and time, building large action foundation models from billions of ground truth action-labeled gameplay clips. Other related developments include Amazon's investment in Odyssey, a startup developing AI systems to simulate the physical world, and Alibaba's unveiling of robot AI suites like Qwen-RobotNav and Qwen-RobotWorld.

Full Take

The narrative presents an emerging nexus between abstract world modeling in virtual environments and tangible physical embodiment. The foundation of General Intuition’s work rests on training agents within simulated gameplay spaces to grasp temporal and spatial relationships, suggesting a path toward grounding advanced AI in actionable, real-world dynamics. This focus on "action" models trained on massive, labeled data from user interaction reflects a fundamental shift from purely predictive modeling to embodied understanding, mirroring trends seen elsewhere in AI development like Amazon's work with simulation platforms (Odyssey) and Alibaba’s multi-modal robot frameworks. The pattern suggests that the convergence of large foundation models with physical world simulation is accelerating, driven by investment targeting both abstract capability and physical instantiation. The implication for human agency lies in determining who controls these generalized action models—whether they are optimized purely for simulated environments or for navigating complex physical realities—and who bears the cost associated with bridging this gap between digital action and physical execution. What are the long-term ethical and systemic implications of prioritizing training on gameplay data when the stated goal is achieving physical world interaction? What infrastructure shifts are required to ensure that progress in general models translates equitably across embodied systems, rather than remaining siloed within high-level simulation environments?

Sentinel — Human

Confidence

The text appears to be a compilation of news snippets relating to the advancement of AI, world models, and associated funding rounds, exhibiting characteristics consistent with aggregated journalistic reporting.

Signals Detected

Moderate sentence length variance; clear shift in focus between data reporting and promotional framing.

Text successfully shifts context from the main subject (General Intuition) to related industry events (Amazon, Alibaba), indicating a broader journalistic sweep.

Attribution is mixed (direct quotes/announcements vs. generalized reports), typical of aggregating multiple data points.

The article cites specific dates, dollar amounts, and source mentions (e.g., June 25 post, $320 million round) which suggests reliance on verifiable, traceable sources rather than pure hallucination.

Human Indicators

The presence of direct quotes from the company (e.g., the June 25 post) and specific financial figures points toward journalistic sourcing.

The flow connects disparate but related events (Medal AI -> Robotics/World Models) in a manner typical of investigative reporting rather than pure LLM extrapolation.