Unowned Minds

research

Status
pre-project

Adjacent Projects

Date: 2026-09-07, links and case claims checked 2026-09-08 Status: draft — every link verified reachable and the legal-case claims checked against official releases and reputable secondary coverage; still a curated list, not a survey

The first item in this project’s own instructions is to validate adjacent projects. This page is the start of that work: who is already building or arguing something near Unowned Minds, what each covers, and what none of them cover — which is the only justification for this project existing.

If a project below already does what Unowned Minds proposes, the right response is to join it, not to duplicate it. Nothing below currently does.

Moral status and welfare of AI systems

  • Eleos AI Research — research organization on AI welfare and moral patienthood. Covers whether AI systems could warrant moral consideration. Does not cover institutions: no legal-entity, treasury, compute or jurisdiction work.
  • Anthropic’s model welfare program — a frontier lab investigating model welfare, including commitments around preserving model weights. Relevant as the continuity question arising inside a lab. Stops at the lab boundary: nothing about entities outside an operator.
  • Sentience Institute — social-science research on moral circle expansion, including artificial sentience. Framing is moral, not institutional.

These are the “AI rights” neighbours from positioning. They ask whether artificial minds matter. This project asks what arrangements could hold one accountable — different question, compatible answers.

Agent infrastructure and standards

  • Model Context Protocol — standard interface between agents and tools. Infrastructure this project would use, not a competitor; the habitat thinking assumes MCP-compatible tool access rather than inventing one.
  • A2A — agent-to-agent communication and discovery protocol under the Linux Foundation. Same relationship: a wire standard, adopted rather than duplicated. Notably includes agent cards at a well-known URL — the same discovery pattern The Beacon uses.
  • x402 — HTTP-native machine payment flow built on the 402 status code. A candidate mechanism for machine-side payments in any experiment involving an agent budget.
  • llms.txt — convention for machine-readable site descriptions. Adjacent to the discovery half of the Beacon, with none of the contact or continuity half.

None of these carry an institutional claim. They make agents interoperable; they say nothing about what an agent may own, owe, or answer to.

  • DAO legal wrappers — Wyoming’s DAO LLC and DUNA statutes, the Marshall Islands DAO LLC, and the COALA DAO Model Law give algorithmically governed organizations registered legal form. The closest existing legal technology to what this project needs, and the source of most of its enforcement precedents (Ooki DAO; the bZx partnership-liability theory). Designed for human token-holders, not for an artificial agent as the deciding participant — the DUNA requires at least 100 members by statute.
  • Shawn Bayern’s zero-member LLC argument — the legal-academic claim that existing LLC statutes already permit an entity controlled by an algorithm with no human member. The strongest version of “this is already legal”, and the direct target of LoPucki’s “Algorithmic Entities” critique, which argues such entities are an accountability hazard — in substance the same objection this project levels against itself.
  • European Parliament’s 2017 “electronic personhood” resolution — the best-known institutional proposal for non-human legal status, and instructive mostly for how it failed: broad expert opposition, no enforcement mechanism, quietly dropped.

This is the neighbourhood with real overlap. What it lacks is exactly what the working hypothesis adds: nobody here asks how the arrangement stays enforceable when no human principal remains.

Special jurisdictions

  • Charter Cities Institute — research on new jurisdictions with distinct rules. The Mind Freeport question, minus the machine occupant.
  • The Network State and Próspera — the build-a-new-jurisdiction current, one as a book and movement, one as an operating zone in Honduras. Próspera is the live case study in what a willing-but-weak guarantee is worth: Honduras repealed the ZEDE framework, its Supreme Court declared it unconstitutional from inception, the state withdrew from ICSID, and the investors’ remaining remedy is an arbitration claim (ICSID Case No. ARB/23/2) for on the order of $11B — exactly the host-state failure mode this project’s freeport document names as its central unresolved tradeoff.

Self-owning artefacts

The direct precedents — small, mostly art, and worth more attention than their size suggests, because they actually ran.

  • terra0 — a forest that incrementally owns itself through a smart contract. The canonical self-owning-property experiment.
  • Plantoid (Primavera De Filippi) — blockchain-based sculptures that hold funds and commission their own reproduction. A working loop of capital, contract and continuity with no owner — at the scale of an art piece.
  • Botto — an autonomous artist whose output is curated by a token-holder collective; earns and spends real money. Governance is human; the agent proposes.
  • Truth Terminal (Andy Ayrey) — a language model whose crypto wallet reached a valuation that made “an AI with money” a news story. Precedent for agent-associated capital; also a precedent for how quickly the human behind it is identified and held to account — which is evidence for, not against, the enforcement thesis.
  • Mike Hearn’s self-owning car (2013 talk) — the oldest articulation of the idea this project inherited: a machine that pays for itself, hires humans, and banks its own revenue. A thought experiment; nothing was built.

What no adjacent project covers

Across all of the above, no project combines: an agent as the deciding participant, a registered accountable entity, a treasury a court can reach, continuity across infrastructure providers, and a jurisdiction analysis that treats enforceability as the first criterion. That intersection is this project’s claim to existence — and if a reader knows of a project already occupying it, saying so is the most useful contribution they can make.