{"slug": "ai-agents-that-live-inside-a-dreamed-up-world", "title": "AI Agents That Live Inside a Dreamed-Up World", "summary": "Cisco AI engineer Shridhar Shah built a tiny multi-agent system where an agent learns to hallucinate future frames of a predator-prey game and then practices inside its own dream. Shah found that a dream model only stays accurate if it captures how players react to each other: a single-player dream fell apart after one frame, while a multiplayer dream remained 100% accurate 15 frames ahead. The work builds on the World Models approach from Ha & Schmidhuber (2018) and shows that for multi-agent dreams, modeling reactions is essential to avoid drift.", "body_md": "*An agent watches a game, learns to hallucinate the next frame, then plays inside its own dream — but only the model that knows players react to each other stays true.*\n\n**TL;DR:** The hottest idea in agents right now: don't feed them the real world — let them **dream** it. An agent watches some footage, learns to hallucinate the next frame, and practices inside its own head. I built a tiny one, and found the catch: a dream only stays true if it knows the players **react to each other.** Runs on a laptop.\n\nTwo players in an 11-cell corridor: a predator steps toward the prey, the prey steps away. Every move is a *reaction*. An agent watches random games, then closes its eyes and **dreams 15 frames ahead**, feeding each prediction back in as the next input. I built two dreamers from the exact same footage:\n\nTraining an agent inside its own learned dream goes back to Ha & Schmidhuber's [ World Models](https://arxiv.org/abs/1803.10122) (2018); the open frontier is making that dream\n\n% of the dream still matching reality, this many frames ahead:\n\n| frames ahead | 1 | 3 | 5 | 10 | 15 |\n|---|---|---|---|---|---|\n| single-player dream | 11 | 0 | 0 | 9 | 0 |\nmultiplayer dream |\n100 |\n100 |\n100 |\n100 |\n100 |\n\nThe single-player dream falls apart almost immediately. The multiplayer dream stays locked to reality the whole way.\n\n```\nreal = dream = start\nfor _ in range(15):\n    real  = real_next(*real)   # what actually happens\n    dream = model(*dream)      # feed the dream its OWN last frame\n    match += (dream == real)\n```\n\nSame footage, same loop. The only difference: whether the model predicts the two players **jointly** or independently.\n\nThe prey moves *because* the predator moved. A model that looks at the prey alone can't see that — so its tiny errors compound each frame until the dream is pure fiction. The multiplayer model conditions on both, so it captures the reaction and keeps re-predicting the real game.\n\nA dream you can act in is a superpower — an agent can practice a thousand risky moves for free. But a dream that forgets everyone else reacts to you isn't practice. It's a delusion.\n\nDreamed worlds let agents rehearse infinitely, safely, at zero real-world cost. The lesson from this toy: for **multi-agent** dreams, model the reactions or the whole thing drifts. Get it right and agents can plan against each other entirely in imagination.\n\n```\ngit clone https://github.com/Shridhar-2205/secret-lives-of-agents\ncd secret-lives-of-agents/03-dreamed-world && python demo.py\n```\n\n**Shridhar Shah** — Senior Software Engineer on the AI team at Cisco. [GitHub](https://github.com/Shridhar-2205) · [LinkedIn](https://www.linkedin.com/in/shridhar-shah-220b1721b/)\n\nSources & further reading:Ha & Schmidhuber,[(2018) — the \"train inside a dream\" idea · Hafner et al.,]World Models[(DreamerV3, 2023) · Bruce et al.,]Mastering Diverse Domains through World Models[(2024).]Genie: Generative Interactive Environments", "url": "https://wpnews.pro/news/ai-agents-that-live-inside-a-dreamed-up-world", "canonical_source": "https://dev.to/shridhar_shah2297/ai-agents-that-live-inside-a-dreamed-up-world-3n4j", "published_at": "2026-07-21 05:51:55+00:00", "updated_at": "2026-07-21 05:59:50.591211+00:00", "lang": "en", "topics": ["artificial-intelligence", "machine-learning", "ai-agents", "neural-networks", "ai-research"], "entities": ["Shridhar Shah", "Cisco", "Ha", "Schmidhuber", "Hafner", "Bruce", "Genie"], "alternates": {"html": "https://wpnews.pro/news/ai-agents-that-live-inside-a-dreamed-up-world", "markdown": "https://wpnews.pro/news/ai-agents-that-live-inside-a-dreamed-up-world.md", "text": "https://wpnews.pro/news/ai-agents-that-live-inside-a-dreamed-up-world.txt", "jsonld": "https://wpnews.pro/news/ai-agents-that-live-inside-a-dreamed-up-world.jsonld"}}