Network
Sharing abundance
If advanced AI multiplies production, the question that remains is who receives the result. This page sets out the evidence on distribution, the mechanisms that have been proposed, and the design the project puts forward.
The promise, link by link
The case for abundance is a chain of five steps. The first three are engineering. The last two are ownership and policy.
| Link | Claim | Kind of question |
|---|---|---|
| 1 | AI does every mental task | Technical |
| 2 | Robots do physical work | Technical |
| 3 | Production multiplies | Technical |
| 4 | Essentials become cheap | Economic |
| 5 | Everyone receives a part | Ownership and policy |
Most public discussion covers the first three. This project is concerned with the last two.
Evidence on distribution
Progress is real. The number of people in extreme poverty fell from 2.3 billion in 1990 to 831 million in 2025.
Ownership is concentrated. The richest 1% hold 37% of global wealth.
Production alone is not enough. The world produces enough food for everyone, and hundreds of millions of people were still hungry in 2024. The constraint is access, not output.
Economists differ on size. Estimates of AI's effect on output over ten years range from about 1% to several percent in bank forecasts, and to orders of magnitude in the most optimistic public statements. The difference is one of kind: a productivity tool, or a replacement for human labour.
Mechanisms that have been proposed
| Mechanism | How it works | Open question |
|---|---|---|
| Universal basic income | Unconditional cash to everyone | Who funds it, and at what level |
| Universal basic compute | A share of AI capacity for each person, to use, sell or donate | How capacity is allocated and valued |
| Public or national funds | Citizens hold shares in a fund that owns AI assets | Governance, and coverage beyond one country |
| Taxes on automation | Revenue from AI is collected and redistributed | Definition and enforcement |
| Member-owned networks | Members govern a shared resource directly | Participation, and legal standing |
What trials show. In the largest cash trial to date, 1,000 people in the United States received $1,000 a month for three years. Recipients did not stop working, and working hours fell slightly. Spending rose mostly on food, rent and transport. The trial was small and temporary, and it was funded from outside. It does not show how a permanent, universal payment would behave.
Proposed design
The project proposes a member-governed network with three properties.
- Rules before output. The rule for sharing is published and adopted before there is anything to share.
- Open records. Membership, votes and any distribution are recorded on a public ledger that anyone can inspect.
- Member control. Changes to the rules pass through a vote with a published threshold and a time delay.
Status. This is a design. No distribution mechanism is operating, and none is promised. The undecided points are listed in Project status.
What a network cannot do
- It cannot replace tax, law or public institutions.
- It cannot create value that does not exist.
- It cannot guarantee any outcome for its members.
Its purpose is narrower: a transparent mechanism whose rules can be read, tested and changed by the people it serves.