Argos: a stateful simulation engine turns video models into a renderer for a persistent world
mallow610 · x · 2026-09-05
Developer whyarethis demoed Argos, a stateful simulation engine rendered through the H3 video model via Fal Director (with fal credits).
- He texted a simulated character who had to physically go upstairs and find her phone before replying — the video model didn't decide the plot; a persistent world of characters, objects, locations, and systems did
- Each character has its own cognition, memories, and goals, choosing actions autonomously while the simulation determines outcomes
- World state is real: phones have locations, messages have delivery/read states, typing takes time — move the phone before she arrives and it's gone
- Argos converts changing world states into instructions for H3 to depict, and can expose the character's thought stream
The takeaway: video models stop inventing stories and become a rendering layer over a stateful world — a reference architecture for persistent AI-simulated environments.
More from coding & agent
- 105 Real Bugs Tested: GPT-6.1 Sol Over 10x Cheaper Than GPT-5.6 Sol — PawelHuryn · 2026-10-23
- Dev builds SkiFree-style sled game in one prompt, praises DeepSeek Flash — victormustar · 2026-10-23
- Every ditched personal agents for a shared Claude Managed Agents coworker in one week — every · 2026-10-23
- WorkOS shows how to let AI agents safely sign up for your app with auth.md and ID-JAG — AI Engineer · 2026-10-23
- Devs on Reddit swap tactics for stopping prompt regressions in AI agents — SleepFlimsy6963 · 2026-10-23
- Claude Code + ComfyUI MCP workflow for client work: paid models or open weights? — newaccount47 · 2026-10-23