Origin Lab raises $8M to bridge video game data with AI world-model builders

BitcoinWorld

Origin Lab raises $8M to bridge video game data with AI world-model builders

As artificial intelligence expands beyond text and images into the physical world, a new class of AI systems known as world models is emerging. These models, designed to understand and simulate physical environments, require vast amounts of data that capture how objects move, interact, and behave in three-dimensional space. Unlike large language models, which can draw from the open web, there is no readily available dataset for physical-world training data. That gap is now attracting startups willing to build the infrastructure from scratch.

A marketplace for synthetic physical data

Origin Lab, a startup founded by Anne-Margot Rodde and her team, has identified an unlikely but abundant source of this the video game industry. The company announced an $8 million seed funding round led by Lightspeed Ventures, with participation from SV Angel, Eniac, Seven Stars, and FPV. Angel investors include Twitch co-founder Kevin Lin and Cruise founder Kyle Vogt. The startup aims to create a marketplace where AI labs building world models can purchase high-quality, licensed data extracted from video game assets.

“The AI systems that are being built now need to understand how the physical world works and how things move,” Rodde told Bitcoin World. “That data essentially lives in video games.” The process involves converting video game assets—such as 3D environments, object physics, and character animations—into structured training data. This can range from simple rendering runs to complex automation that generates hours of walkthrough footage.

Why video game data matters for AI

World models, like those being developed by Yann LeCun’s AMI Labs or Fei-Fei Li’s World Labs, require data that simulates real-world physics, object permanence, and spatial relationships. Video games, particularly those with realistic physics engines and open-world environments, already contain this information in a structured, repeatable format. Licensing these assets, however, has historically been a barrier. In December 2024, OpenAI faced scrutiny when its Sora video-generation model appeared to reproduce footage from popular video games and Twitch streams, raising questions about training data provenance.

Origin Lab positions itself as a bridge between two industries that have not traditionally collaborated. “It became clear that the video game industry was sitting on some incredibly valuable data, but there was no real way or infrastructure to basically connect AI labs and the video game industry,” Rodde said. “So essentially, we built that bridge.”

Investor confidence in data infrastructure

The funding round signals growing investor interest in the data supply chain for AI. Faraz Fatemi, a partner at Lightspeed who led the investment, drew parallels to the success of companies like Scale AI, which provides data labeling and management services to major AI labs. “We’ve seen how sharp the revenue scaling can be for data vendors that are serving the major labs,” Fatemi told Bitcoin World. “These are very well-capitalized businesses, and the bottleneck for all of them is data.”

Origin Lab’s approach offers video game companies a new revenue stream from existing digital assets, while AI labs gain access to clean, licensed data that reduces legal and ethical risks. The startup’s ability to secure backing from prominent investors suggests that the market for physical-world training data is maturing rapidly.

Conclusion

Origin Lab’s $8 million seed round highlights a critical inflection point in AI development: the shift from text-based models to systems that must understand and interact with the physical world. By connecting the video game industry with AI labs, the startup is addressing a fundamental data bottleneck. As world models become central to robotics, autonomous systems, and simulation, the demand for high-quality, licensed training data will likely intensify. Origin Lab’s marketplace model could become a template for how specialized data is sourced in the AI era.

FAQs

Q1: What are world models in AI?
World models are AI systems designed to understand and simulate physical environments, including how objects move, interact, and behave in three-dimensional space. They are used in robotics, autonomous vehicles, and simulation training.

Q2: Why is video game data valuable for training world models?
Video games contain structured, physics-based environments with realistic object interactions, spatial relationships, and movement patterns. This data is difficult and expensive to collect from the real world but is readily available in game engines.

Q3: How does Origin Lab make money?
Origin Lab operates as a marketplace, charging fees for converting video game assets into training data and facilitating transactions between game companies and AI labs. The company handles licensing, data formatting, and quality assurance.

This post Origin Lab raises $8M to bridge video game data with AI world-model builders first appeared on BitcoinWorld.