AI Summary
5 min readWorld Labs co-founder Justin Johnson argues that language models gave AI a way to work with words, but the next frontier is models that can understand and simulate the physical world. On the a16z Show, he introduced Atlas, which the company calls the first multimodal world model—a system designed not just to generate video, but to generate, reconstruct, and simulate 3D environments with precise camera control. The core thesis is that just as language models became a horizontal engine for processing discrete tokens, world models can become a horizontal engine for processing visual and spatial data, applicable across gaming, VFX, architecture, and robotics.
What a World Model Actually Does
Continue reading the full summary in the app — free to try.
Read Full Summary →Free • No credit card required
Never miss an episode of a16z Show
Get every new episode summarized in your inbox — free, ~5 minutes to read.
No spam. Unsubscribe anytime.
What you'll learn
- 1 (00:00) **The Thesis: World Models as a New Horizontal Category** - The core argument that language models are general for discrete tokens, while world models are a new category for visual and physical understanding.
- 2 (01:05) **Introducing Atlas: The First Multimodal World Model** - Justin Johnson introduces Atlas, the multimodal world model announced today, and its three core capabilities.
- 3 (03:09) **The Central Thesis and the Problem of Long-Horizon Generation** - Justin explains how Atlas maintains spatial control to avoid the jumbled outputs typical of video models over long durations.
- 4 (05:12) **Two Futures: Explicit 3D vs. Direct Frame Generation** - The discussion contrasts generating explicit 3D assets (like Gaussian splats) for use in game engines versus having the model directly output 2D frames.
- 5 (07:04) **Atlas vs. Marble: Breaking the Gaussian Dependency** - Justin explains how Atlas represents a totally rebuilt stack that decouples 2D and 3D outputs for better scalability and quality.
- 6 (09:30) **The Path to Dynamic World Generation and Robotics** - The conversation explores how Atlas could eventually power experiences like Microsoft Flight Simulator and the specific value of "real-to-sim" for robotics.
- 7 (12:08) **Real-to-Sim-to-Real: A Path to General Robots** - Justin describes how Atlas could enable rapid robot onboarding by reconstructing a specific space from a few photos and then fine-tuning a general robot within that simulation.
+ Full timestamped outline available in the app
Guests on this episode
Show Notes
World Labs co-founder Justin Johnson joins MTS hosts Theo Jaffee and Sofia Puccini to discuss Atlas, World Labs’ latest world model, and the broader case for AI systems that understand and interact with the physical world.
Justin explains how Atlas approaches three core tasks: generating new worlds, reconstructing real environments from images, and simulating how objects or robots might behave within them. Underlying it is a bigger thesis: just as language models became general-purpose engines for working with text, world models could become a horizontal layer for visual and physical intelligence across industries from entertainment and gaming to construction and robotics.
They also explore how world models could change video games and creative tools, why precise spatial control matters, and the potential for “real-to-sim-to-real” robotics, where a few photos of a physical environment could eventually be enough to build a simulation and adapt a robot to that specific space.
Resources:
Follow Justin Johnson on X: https://x.com/jcjohnss
Follow Theo Jaffee on X: https://x.com/theojaffee
Follow Sofia on X: https://x.com/schisofrenia
Follow MTS on X: https://x.com/mtslive
Stay Updated:
Find a16z on YouTube: YouTube
Find a16z on X
Find a16z on LinkedIn
Listen to the a16z Show on Spotify
Listen to the a16z Show on Apple Podcasts
Follow our host: https://twitter.com/eriktorenberg
Please note that the content here is for informational purposes only; should NOT be taken as legal, business, tax, or investment advice or be used to evaluate any investment or security; and is not directed at any investors or potential investors in any a16z fund. a16z and its affiliates may maintain investments in the companies discussed. For more details please see a16z.com/disclosures.
Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.