{
  "$type": "site.standard.document",
  "bskyPostRef": {
    "cid": "bafyreigeziozez67pjtgzbomb6vtcxwmuo5tc5iolwvm2kld54aafkiqeu",
    "uri": "at://did:plc:pgryn3ephfd2xgft23qokfzt/app.bsky.feed.post/3mnxwlot2ei62"
  },
  "path": "/t/smft-turns-enactive-ai-into-a-testable-runtime-architecture/176697#post_1",
  "publishedAt": "2026-06-10T22:13:37.000Z",
  "site": "https://discuss.huggingface.co",
  "tags": [
    "https://osf.io/hj8kd/files/osfstorage/6a29d8138f5abdf103d14ddb"
  ],
  "textContent": "(This is an AI generate article)\n<Enactive Artificial Intelligence as Ledgered World-Making: An SMFT Framework for Action, Trace, Residual, and Self-Maintaining Agents> https://osf.io/hj8kd/files/osfstorage/6a29d8138f5abdf103d14ddb\n\nThe core proposal is:\n\n> (0.1) Enactive AI gives the direction: cognition = active world-engagement.\n>\n> (0.2) SMFT gives the operational loop: Field → Declaration → Projection → Gate → Trace + Residual → Ledger → Revision.\n>\n> (0.3) Mature Enactive AI = active engagement + declared protocol + trace ledger + residual governance + self-maintenance.\n\n**This article therefore reframes Enactive AI as ledgered world-making.**\n\nA mature AI agent is not merely\n\n  * a model that answers,\n  * a policy that maximizes reward, or\n  * a tool-user that executes actions.\n\n\n\nIt is a **bounded world-forming system**\n\n  * whose actions reshape future disclosure,\n  * whose experience is stored as future-causal trace,\n  * whose body is its maintained runtime structure, and\n  * whose autonomy depends on its ability to preserve coherence under budget, drift, failure, and residual uncertainty.\n\n\n\nThe practical result is a research program that can be tested today.\n\nCurrent LLM agents, RAG systems, tool-use systems, workflow agents, and reinforcement learning environments can be compared under SMFT-inspired benchmarks: action–perception coupling, residual-honest answering, tool-body embodiment, self-maintenance audits, and gauge robustness under equivalent task framings.\n\nThe article’s central thesis is simple:\n\n> (0.4) Enactive AI becomes experimentally mature when active engagement is converted into declared, trace-bearing, residual-honest runtime architecture.",
  "title": "SMFT turns Enactive AI into a testable runtime architecture"
}