Human labor
Time, skill, judgment and physical effort are scarce and cannot be copied instantly.
1. Labor becomes capitalized
AI labor
Software intelligence can be reproduced across millions of tasks at very low marginal cost.
Physical AI
Robotics converts parts of manual work into machines, compute, energy, sensors and software ownership.
Central idea: automation does not merely eliminate labor. It converts many labor capabilities into reproducible capital. Ownership of models, robots, data centers, energy systems, intellectual property and distribution therefore becomes more important.
2. Why capital may gain power
1
Tasks automatedWork is embedded in software and machines
2
Marginal cost fallsOne system serves many users
3
Scale expandsWinning platforms operate globally
4
Returns follow ownershipProfits accrue to scarce productive assets
Automation research distinguishes a displacement effect, which can reduce labor's share, from a reinstatement effect, in which technology creates new human tasks. The outcome depends on whether new tasks and productivity gains offset displacement.
3. Scarcity moves rather than disappears
| More abundant | Still scarce | Implication |
|---|---|---|
| Routine analysis, translation, basic coding | Compute, chips, electricity, data centers | Value shifts toward infrastructure ownership |
| Standard manufacturing and logistics | Land, minerals, grid capacity, supply chains | Physical bottlenecks gain strategic value |
| Basic personalized education and advice | Trust, accountability, accreditation | Verified human judgment commands a premium |
| Mass customization | Status, exclusivity, authentic experiences, attention | Positional goods remain competitive |
4. The macroeconomic issue may be purchasing power
An economy can become capable of producing almost everything while still failing socially if most households neither own productive capital nor receive enough income to buy what is produced.
Distribution is therefore not only a moral question. It is also a demand-management question. Firms cannot monetize unlimited output if households lack purchasing power.
5. Absolute prosperity and relative dissatisfaction can coexist
Absolute needs
Food, transport, education, health administration, energy management and basic services can become cheaper and more accessible.
Positional needs
Prestigious neighborhoods, elite institutions, influence, attention, rank and exclusive human access remain scarce because their value depends on comparison.
AI can reduce deprivation without eliminating hierarchy. Everyone can have a better home; everyone cannot live in the top 1% neighborhood.
6. The likely outcome: abundance and inequality rise together
| Group | Before | After major AI gains | Meaning |
|---|---|---|---|
| Typical household | Real-income index 100 | 160 | Living standards improve |
| Large capital owner | Real-wealth index 1,000 | 5,000 | Ownership compounds faster |
| Social perception | Moderate gap | Much larger relative gap | People feel richer and more unequal |
7. Progressive income tax may not be enough
Income is taxed annually
Wages and realized income enter the tax system relatively quickly.
Wealth compounds silently
Unrealized gains in businesses, stocks, land and intellectual property may grow for decades.
Ownership captures upside
Redistributing income after production may not correct concentrated ownership.
The policy question may shift from “How much should high earners be taxed?” to “Who owns the AI, robotics, energy, data and infrastructure that produce national income?”
8. A durable policy architecture
Broad capital ownership
Automatic retirement accounts, employee equity, public investment funds, matched savings and citizen dividends.
Purchasing-power support
Negative income taxes, wage supplements, portable benefits or a limited basic income during transition.
Low-cost essentials
Use AI to reduce the cost of health, education, transport, energy and administration.
Competition policy
Open standards, interoperability, infrastructure access and antitrust enforcement.
New human tasks
Responsibility, care, trust, creativity, persuasion, coordination and real-world judgment.
Tax redesign
Broad bases, fewer avoidance channels and careful taxation of economic rents.
9. Capital returns will not necessarily remain unlimited
If AI and robots become widely available and competition remains strong, prices may fall and excess returns may be competed away. Consumers capture much of the gain. If a few firms control models, chips, energy, data and distribution, returns may remain concentrated as monopoly or scarcity rents.
The decisive variables are institutional: competition, ownership, intellectual-property rules, energy supply, capital access, education and taxation.
10. Conclusion
AI is unlikely to abolish scarcity. It is more likely to move scarcity away from routine labor and toward ownership, infrastructure, energy, land, trust, attention and status.
A healthy post-AI economy would spread ownership of productive capital, maintain competition, create new human-centered tasks and use technology to lower essential living costs.
Research references
- IMF, “AI Adoption and Inequality,” 2025. Source.
- ILO, “Generative AI and Jobs: A Refined Global Index,” 2025. Source.
- Acemoglu and Restrepo, “Automation and New Tasks,” NBER. Source.
- OECD, “The Impact of Artificial Intelligence on Productivity, Distribution and Growth,” 2024. Source.
- OECD, “Macroeconomic Productivity Gains from AI in G7 Economies,” 2025. Source.