NextFin news, On October 16, 2025, General Intuition PBC, an AI startup specializing in spatial-temporal reasoning and world models, announced the closing of a $133.7 million seed funding round. The investment was led by prominent venture capital firms Khosla Ventures and General Catalyst, with participation from Raine Ventures. The company, headquartered in the United States, was recently spun out from Medal B.V., a popular video game clip sharing platform that generates approximately 2 billion video uploads annually from 10 million monthly active users across tens of thousands of games.
General Intuition’s core mission is to develop advanced AI models capable of understanding and interacting with three-dimensional environments through spatial-temporal reasoning. By utilizing Medal’s extensive dataset of curated gameplay footage, the startup trains AI agents to perceive and navigate virtual worlds in a manner akin to human players, relying solely on visual inputs and controller-like commands. The company’s initial focus includes enhancing AI-driven features in video games—such as creating adaptive non-player characters (NPCs) that scale difficulty dynamically—and powering autonomous search-and-rescue drones that can operate in unfamiliar physical environments without GPS.
The startup’s CEO, Pim de Witte, highlighted the unique value of Medal’s dataset, noting that gamers tend to upload clips featuring extreme or edge-case scenarios, which are particularly valuable for training robust AI models. This data moat reportedly attracted acquisition interest from OpenAI last year, which offered $500 million for Medal’s datasets, underscoring the strategic importance of high-quality spatial-temporal training data in AI development.
General Intuition’s approach distinguishes itself from other world model developers like Google DeepMind and World Labs by focusing not on selling world models as standalone products but on applying spatial-temporal reasoning to practical use cases. Their models have demonstrated the ability to generalize to unseen environments and predict plausible actions, a significant technical milestone that leverages the transferability of gaming-derived spatial understanding to real-world robotics and autonomous systems.
Despite the promise, challenges remain in the domain of AI-generated virtual environments, including simulation consistency and the complexity of multi-agent interactions. Current state-of-the-art world models, such as DeepMind’s Genie 3, face limitations in simulating complex, interactive scenarios, which General Intuition aims to address through ongoing research and engineering efforts funded by this round.
The $133.7 million capital infusion will accelerate General Intuition’s hiring of researchers and engineers and support the development of sophisticated data processing pipelines to handle the massive influx of gameplay footage. This infrastructure is critical for filtering noise and duplicates, ensuring high-quality training data that underpins the spatial-temporal reasoning capabilities of their AI agents.
From a broader industry perspective, General Intuition’s funding round reflects a growing investor appetite for AI technologies that extend beyond language models to incorporate embodied intelligence and real-world interaction capabilities. Spatial-temporal reasoning is increasingly recognized as a foundational component for achieving artificial general intelligence (AGI), addressing the limitations of large language models that lack intrinsic understanding of physical space and time.
Looking ahead, General Intuition’s technology has the potential to disrupt multiple sectors. In gaming, adaptive AI agents could revolutionize player engagement and retention by providing personalized, dynamic challenges. In robotics and autonomous systems, spatial-temporal reasoning models could enable safer, more efficient navigation and decision-making in complex, unstructured environments such as disaster zones.
Moreover, the company’s strategic choice to leverage video game data as a proxy for real-world spatial learning may set a precedent for future AI training paradigms, emphasizing the value of curated, high-fidelity datasets over generic large-scale data. This approach could catalyze a new wave of AI startups focusing on embodied cognition and interactive intelligence.
In conclusion, General Intuition’s successful $133.7 million seed round marks a significant milestone in the evolution of AI world models. By harnessing spatial-temporal reasoning and leveraging unique video game datasets, the startup is well-positioned to advance the frontier of AI agents capable of understanding and interacting with the physical world, a critical step toward the realization of AGI and transformative applications across industries.
According to SiliconANGLE, this funding round is one of the largest seed investments in the AI spatial-temporal reasoning space, signaling strong market confidence. The company’s progress and commercialization strategy will be closely watched as it navigates technical challenges and competitive dynamics in the rapidly evolving AI landscape.
Explore more exclusive insights at nextfin.ai.
