General Intuition's $300M Funding Talks Signal $2B Valuation

General Intuition, the New York-based startup at the forefront of developing a foundational AI model designed to teach agents how to navigate and interact within spatial and temporal environments, is reportedly in advanced discussions to secure approximately $300 million in new funding, according to sources familiar with the matter who spoke to TechCrunch.

This significant capital injection comes merely eight months after General Intuition emerged from Medal, a popular platform for uploading and sharing video game clips, having initially raised a $134 million seed round. The anticipated fresh funds are expected to propel the startup's valuation to just over $2 billion, as indicated by these sources.

TechCrunch sources further reveal that General Intuition has already garnered commitments from prominent investors, including Jeff Bezos and Eric Schmidt, alongside continued support from existing backers Khosla Ventures and General Catalyst.

The company is led by Pim de Witte, a co-founder of Medal, who established General Intuition alongside co-founders Eloi Alonso, Adam Jelley, and Vincent Micheli. This founding team brings extensive research expertise in the critical areas of world modeling and simulation.

General Intuition leverages Medal's vast dataset, comprising 2 billion videos annually from 10 million monthly active users, to train its embodied AI and world models. The startup's core proposition highlights the unparalleled value of this dataset, which is uniquely suited for teaching machines profound spatial-temporal reasoning. Its origin in interactive, first-person gameplay allows AI to learn to perceive, anticipate, and interact in real-time within simulated environments.

This distinctive dataset has reportedly captured the interest of major players, including OpenAI, which had previously attempted to acquire Medal. Sources suggest that OpenAI is not the only prominent AI laboratory that has approached General Intuition.

The domain of world models, where General Intuition operates, is experiencing rapid growth and increased competition. Several startups, such as Runway, Decart, and World Labs, have recently unveiled their own world models. Concurrently, Google’s Genie 3 has begun incorporating Google Maps data to enhance its real-world simulation capabilities.

While many companies in this space primarily view gaming and robotics training as immediate commercial applications, General Intuition distinguishes itself with a unique strategy: it develops world models specifically to train its own AI agents, rather than selling the world models themselves. The agents constitute the primary product, and the startup's proprietary dataset provides a crucial pathway to market viability.

A source close to the company indicated that General Intuition plans to allocate the newly raised capital towards significantly expanding its compute capacity. This scaling effort is aimed at enabling the release of a new product by the close of summer or early fall.