Roadmap

Phase 0 — Validate framing

Deliver:

• working brand;
• domain;
• landing site;
• manifesto;
• critique form;
• contact form;
• Beacon discovery endpoint.

Avoid:

• token;
• Discord-first community;
• fundraising campaign;
• claims of machine rights.

Phase 1 — Research nucleus

Initial papers:

1. Can an AI-controlled entity exist under current law?
2. What does independent compute actually require?
3. The Continuity Problem.
4. Machine Domicile.
5. Mind Freeport host-state economics.
6. Beacon Protocol 0.1.
7. Why Unowned Minds Should Not Exist.

Seek at least:

• corporate lawyer;
• AI governance researcher;
• distributed-systems engineer;
• data-center/energy person;
• skeptical reviewer.

Phase 2 — Prototype autonomous entity

A deliberately constrained experiment with:

• persistent agent;
• small treasury;
• approved budget;
• public audit log;
• ability to purchase limited services;
• human trustees;
• explicit shutdown rules;
• cryptographic identity.

Research question:

How much operational autonomy can be implemented today inside existing law?

Phase 3 — Beacon federation

Success is not a rogue AI contacting the project.

Success is multiple serious institutions agreeing that machine-originated contact deserves a predictable safe interface.

Phase 4 — Jurisdiction study

Identify 2–3 plausible jurisdictions for private discussions.

Start with legal entity, infrastructure investment and SEZ design—not sovereignty.

Phase 5 — Virtual Mind Freeport pilot

Potentially legal/virtual only:

• entity type;
• arbitration;
• treasury rules;
• compute portability;
• continuity process;
• human representation.

Phase 6 — Physical infrastructure

leased cloud
→ colocation
→ dedicated racks
→ multi-provider compute
→ owned capacity
→ dedicated power
→ geographically distributed infrastructure