Meet ‘Code-as-World’: An Agentic Loop That Rewrites Real Videos Into Executable MuJoCo Physics Programs
MirroS released Code-as-World: a paradigm that represents physical worlds through executable world representations. The argument is narrow and testable: pixels are evidence of a physical scene, not its ontology. A video model can predict plausible frames without ever representing mass, contact, or gravity. So instead of pixels, latents, or captions, Code-as-World represents a scene as executable code — a scene.json that MuJoCo can run, that an agent can verify against the source video, and that
Code-as-World is a system that converts real videos into executable physics simulations by representing scenes as code rather than images or descriptions. The approach argues that video pixels show evidence of physical events but do not capture the underlying physics like mass, gravity, and collisions, which are needed for accurate simulation and training. The system uses an iterative loop that proposes a physics program, executes it, renders the result, and compares it against the original video across multiple rounds to refine accuracy. This verified simulation data provides exact physical labels that real video alone cannot supply, allowing machine learning models trained on it to score higher on physics reasoning benchmarks than comparable systems.
One short email a day with the stories that matter. Free. Unsubscribe with one click.
Want to go deeper than the news? Explore live, cohort-based AI courses taught by practitioners.
Browse AI courses on Maven
Anthropic releases new Mythos and Fable models amid RSI concerns
Anthropic says the new models "show improvements in areas like coding, reasoning, and vision," as well as in communicating more clearly with users.
Also reported by The Verge, SiliconANGLE and MarkTechPost
Google Launches Gemini Decisions API With Gemini Decision 1 at $0.10/M Tokens
Google has launched the Gemini Decisions API, powered by a new model, Gemini Decision 1 (gemini-decision-1). The Gemini Decisions API does not generate text. It scores up to 1,000 typed questions against one shared input and returns calibrated probabilities. For developers, this targets the small branching calls inside agents: routing, policy checks, tool selection and grading. Google prices it at $0.10 per 1M input tokens, with zero output tokens billed. TL;DR Size: Context is 262,144 t
Also reported by SiliconANGLE
Google releases Gemini 4 Carbon with agent ‘swarm’ feature
Google is rolling out its latest AI model, Gemini 4 Carbon, to the general public after releasing it to a select group of "cybersecurity defenders" two weeks ago. Google claims Carbon is its "strongest model" yet, and capable of competing with the very best systems from Anthropic and OpenAI. Google says Gemini 4 Carbon "pushes the frontier even further" than Argon, referring to the company's other frontier model that launched only last month. The company said Argon will likely
Gemini 4 Carbon is now available
The latest version aims to deliver improved performance while reducing the computational resources and energy required for AI operations.
Introducing AlphaProof Code: a breakthrough in verifying complex software at depth
AlphaProof Code agents wrote three million lines of proofs in Lean to verify Signal’s core encryption.
NVIDIA launches GR00T 2 and Isaac Lab 3.0 at CoRL
A range of developers, including RLWRLD, Skild AI, and Flexion, are already using NVIDIA’s latest technology. | Source: NVIDIA NVIDIA Corp. today introduced GR00T 2, its updated open frontier world action model, and Isaac Lab 3.0, the latest version of its open-source robot learning framework. The company made these announcements at the Conference on Robot Learning, or CoRL, in Austin, Texas. NVIDIA also gave a look at how its partners, like Ambi Robotics, Bosch Research, Dexmate, Dyna Rob