Inside General Intuition’s Push to Build Embodied AI Agents

4 snips

Dec 9, 2025

In this lively discussion, Pim de Witte, the tech entrepreneur behind General Intuition and Medal, dives into the exciting realm of world models and embodied AI. He explains how gameplay videos create a fertile training ground for AI, enabling it to simulate real-world interactions. Pim highlights the advantages of using video game data versus platforms like YouTube, and the unique challenges in developing human-like agents. Expect insights into cross-game training and the future of gaming, where bots could enhance player experiences like never before!

Ask episode

AI Snips

Chapters

Transcript

Episode notes

INSIGHT

World Models Are Spatial‑Temporal Simulators

World models simulate how objects and agents move and interact in space and time rather than just describing them as LLMs do.
Pim de Witte argues this spatial-temporal modeling is the next major frontier that enables embodied intelligence.

ANECDOTE

Keynote With A I Luminary

Pim described giving a keynote with Yann LeCun as a surreal career highlight and said it felt intimidating but rewarding.
He framed it as a meaningful moment because of LeCun's technical stature and large audience.

ANECDOTE

Metal Captures Action Labels, Not Raw Inputs

Metal records short gameplay clips and converts player inputs into action labels instead of raw keystrokes for privacy and training value.
Pim explains this design makes Metal clips better training data than typical YouTube or TikTok game videos.

Get the Snipd Podcast app to discover more snips from this episode

Get the app

World models are rapidly becoming AI’s next frontier, and in this episode we break down why. Host, Alexandra Takei, Director at Ruckus Games, sits down with Pim de Witte, founder of General Intuition and Medal, to explore how billions of gameplay videos can power a new class of embodied agents. Pim explains the fundamental gap between language models, which describe the world, and world models, which simulate the world, capturing how objects and agents move, react, and evolve in space and time. The conversation digs into why video games are an ideal training ground, including but not limited to consistent first-person perspectives, action labels (if you design your data set that way), and optical fidelity that platforms like YouTube can’t provide.

Pim walks through General Intuition’s technical approach, why cross-game training unlocks more human-like behavior, and the specific limitations still unsolved, such as multiplayer consistency, long-horizon coherence, and the cost of large-scale inference. They explore what studios can expect from embodied agents: bots trained on human behavior that they hope to be tunable by designers and ideal for developers who want to embrace and build around this tech to either develop new game genres or make it a bedrock of their production process. If you are interested in learning about a company with a unique approach to world models and embodied agents, this is a must-listen to close out 2025.

We’d like to thank Lysto for making this episode possible! Lysto is revolutionizing how game development teams collect and act on real player feedback with its AI-powered playtesting insights. Learn more about how you can get bias-free feedback at https://lysto.gg/?utm_source=naavik&utm_medium=podcast&utm_campaign=ad

If you like the episode, please help others find us by leaving a 5-star rating or review! And if you have any comments, requests, or feedback shoot us a note at podcast@naavik.co.

Watch the episode: YouTube Channel

For more episodes and details: Podcast Website

Free newsletter: Naavik Digest

Sound design by Gavin Mc Cabe