pXu — Personal X Units

The automation ownership problem: who owns productive AI?

Artificial intelligence and robotics can raise productivity while reducing the amount of human labour required for many tasks. Policy discussions therefore increasingly focus on technological unemployment, wage pressure, inequality and how the gains from automation should be redistributed.

Universal basic income, robot taxes and wage subsidies mostly intervene after productive ownership has already been established. pXu asks an earlier question: can ownership of productive technological capacity itself be distributed?

If AI and robotics become a major form of productive capital, who should own that productive capacity?

The alternative path explored by pXu is: automation → distributed productive ownership → productive participation → active income.

What is pXu?

pXu means Personal X Unit. The X is deliberately a variable: it identifies the kind of productive technological capacity associated with the person. The framework is therefore not tied to one device, one model, one robot, one vendor or even one technological form.

In policy and technical contexts, an individual unit can be referred to as a pXu. A pXu may be digital, physical or hybrid. Its defining characteristic is economic and legal: it represents portable productive technological capital capable of performing authorized work with, for or on behalf of its owner.

One framework, two primary forms: pAIu and pRu

The variable X allows the framework to distinguish between cognitive and physical automation without creating two unrelated policy concepts.

A hybrid deployment may combine a pAIu and a pRu—for example, personal AI capacity that plans or supervises work performed through robotic hardware. This makes pXu technology-neutral: the policy principle concerns who owns productive automation, not which technology implements it.

Seven principles of productive sovereignty

Active income rather than passive redistribution alone

A universal basic income creates an income floor through cash transfers. pXu investigates a different layer of the automation economy: ownership of productive capacity. The approaches are not necessarily mutually exclusive, but they address different questions.

The framework explores remuneration linked to the economic function performed by a pAIu or pRu and indexed over time, rather than allowing compensation to collapse toward the marginal cost of software execution or machine time. The precise formula would require sector-specific economic modelling.

Qualifying non-negligent technical failures should not automatically eliminate income. A regulated framework could treat such failures analogously to temporary incapacity while assigning repair and operational duties to the deploying organization.

Distributed ownership still requires solidarity

Distributed productive ownership does not remove the need for social protection. People differ in circumstances, caregiving responsibilities and ability to participate in markets. A Digital Solidarity Levy could allocate part of the value generated by highly automated production to people whose productive participation is limited or impossible.

The objective is to prevent distributed productive capital from creating a new divide between owners of highly demanded pXus and structurally disadvantaged citizens. pXu therefore treats productive ownership and social solidarity as complementary layers rather than substitutes.

Why decentralized AI and interoperable robotics matter

Personal productive sovereignty is weak if every pXu depends on one cloud platform, one proprietary model or one vendor. pXu therefore points toward decentralized, private, interoperable and verifiable infrastructure, with local execution where practical.

Possible mechanisms include encrypted sandboxing between personal and employer environments, a national or federated work ledger for authorized productive activity and automatic payment, emergency peer-to-peer capacity rental, cryptographic provenance, and independent physical and logical stop mechanisms for safety-critical deployments.

This is where pXu intersects with the broader Glass Box AI philosophy: critical AI infrastructure should make ownership, authorization, provenance and responsibility inspectable rather than burying them inside opaque service boundaries.

What would a pXu legal framework need to define?

pXu is a policy framework proposal, not enacted law. A real implementation would require jurisdiction-specific analysis of labour law, social security, taxation, data protection, product liability, competition law, intellectual property, insolvency, consumer law and AI regulation.

• Status of the pXu: what exactly is owned, what remains licensed and how productive rights are represented.
• pAIu and pRu classification: how digital, robotic and hybrid units are treated.
• Employer obligations: financing thresholds, maintenance, updates, cybersecurity and timely repair.
• Portability: how a pAIu or pRu remains portable after dismissal, resignation, retirement or insolvency.
• Remuneration: how productive contribution is measured and compensated.
• Liability: allocation of responsibility among individual owner, employer or deployer, manufacturer, model provider and operator.
• Data protection: separation of personal data, employer data and operational telemetry.
• Competition: safeguards against vendor lock-in and excessive concentration of productive infrastructure.
• Safety: supervision, repair obligations and independent stop mechanisms for high-risk systems.

The proposal also considers a productive-concentration threshold. Above a defined level, highly automated firms could face obligations to incorporate worker-associated pXus rather than replacing labour entirely with productive capital owned only by the firm. Thresholds, exemptions and sector-specific rules would require empirical testing.

Open questions: pXu is meant to be challenged

How should productive value be measured when many AI agents and robots cooperate? Can portable pXus remain competitive without recreating platform dependence? How should intellectual property be separated from ownership of productive capacity? Could mandatory financing discourage hiring? Which solidarity mechanisms preserve inclusion without distorting incentives? How should hybrid pAIu/pRu systems be classified? Which parts of the framework belong in labour law, social law, corporate law or AI regulation?

These questions are the research agenda, not defects to hide. pXu is intended to be criticized, modelled, simulated and refined by economists, lawyers, engineers, workers, companies and policymakers.