Do these 3 things before closing this tab:
1Clear out junk files and repair common Windows errors2Scan for outdated or missing drivers - takes under a minute3Repair Windows errors before they cause bigger problemsiTechGuides is reader-supported. When you buy through links on our site, we may earn an affiliate commission. As an Amazon Associate I earn from qualifying purchases. Learn more
There is no established forecast for when superintelligence might arrive—or whether it would cause mass unemployment. If advanced AI does concentrate income in the hands of capital owners, sharing its gains will be a question of tax design, ownership and public spending, not a problem with one proven solution. The most useful starting point is to prepare flexible ways to distribute gains without assuming a particular employment outcome.
What is the distribution problem?
AI could reshape work without eliminating it: people may move into new jobs, while others face disruption. A different possibility is that more income flows to owners of companies, data, computing infrastructure and other capital, while labour earns a smaller share. The Economist’s May 16, 2026 article, “How to share the AI windfall,” discusses both possibilities; it does not establish that mass unemployment is happening or inevitable. The phrase “age of superintelligence” should therefore be read as a hypothetical future, not a description of a proven present.
The public-finance challenge depends on which outcome occurs. If wages remain a large source of household income, familiar taxes on earnings and spending may continue to fund public services and transfers. If returns shift toward capital, governments may need revenue sources that reach those returns. The policy choice is not simply how much to redistribute: it is also how to raise money while limiting avoidable distortions to work, investment and consumption.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
The Economist reports that about half of tax revenue in the average OECD member comes from labour and another 30% from consumption taxes, with the balance from corporate, capital and property levies. That figure is reported by the magazine’s 2026 article; the reproduced passage does not identify its underlying data table, so it should not be treated here as independently verified OECD data. The article also recounts economist Nicholas Kaldor’s observation that the ratio of labour to capital income had been “a remarkably stable two-to-one.” This is a historical observation as reported by The Economist, not a current universal ratio; the article notes that some measures show a declining US labour share.
#1 Best Overall
Which tax and ownership tools could share gains?
The options below are proposals and economic trade-offs, not a ranking based on measured program outcomes. The Economist’s analysis covers these approaches; the actual distributional and investment effects would depend on how any policy is designed and applied.
| Approach | How it could share gains | Main consideration |
|---|---|---|
| Consumption taxes and transfers | Raise revenue from spending and return some of it through transfers. | Broad revenue base; the tax can weigh more heavily on people who spend a larger share of their income. |
| Progressive income taxes | Collect more from higher earners and redistribute through public spending or transfers. | Redistributive through wages, but more exposed if labour income shrinks. |
| Capital and rent taxes | Tax returns to ownership or economic rents, which could account for a larger share of income. | Potentially reaches concentrated gains; may affect investment and saving incentives. |
| Robot or token taxes | Tax a particular automated input or AI-related use. | More targeted in appearance, but could distort investment, automation or spending. |
| Wider citizen ownership or sovereign-wealth funds | Give the public direct or indirect claims on company returns. | Requires decisions about investment selection, governance and access for less-well-off households. |
| Social investment stipend | Pay people for specified care, community-service or education contributions. | Raises difficult questions about eligibility, valuation, monitoring, cost and administration. |
Keep broad taxes and transfers adaptable
Consumption taxes can raise revenue even when wages are not the immediate source of income, while transfers can direct some proceeds to households. But taxing spending can be burdensome for lower-income households, especially if they must spend most of what they receive. A transfer can offset that effect, though its design determines who benefits and how much.
Rank #2
Progressive income taxes are a familiar way to redistribute earnings, but they depend on taxable income. If labour earnings become a smaller part of the economy, a system relying heavily on wages may have less reach into the new source of wealth. This is a contingency to plan for, not evidence that wage income is about to disappear.
Tax capital and rents with attention to incentives
When returns accrue disproportionately to owners, taxes on capital income or economic rents may reach gains that wage-based taxes miss. The trade-off is that taxes can change incentives to invest or save. “Rent” in this context means returns above what is needed to keep an activity or asset in use, rather than ordinary payment for a property alone.
Robot taxes and token taxes are more specific versions of taxation on capital or consumption, respectively, in The Economist’s account. Their targeting does not make their effects certain or harmless: a levy on automation could discourage investment, while a tax on AI use could affect spending. The magazine presents these distortions as an economic argument, not a settled empirical finding that applies identically to every design.
Give citizens a stake in returns
Governments could distribute shares, build sovereign-wealth funds that hold investments on behalf of the public, or broaden share ownership among people with fewer assets. Public funds can spread returns beyond company owners, but require governance rules and choices about which investments to hold. Direct ownership can connect people more visibly to gains; The Economist also argues that it may reduce exposure to companies shifting profits across borders. Neither approach guarantees equal gains, and the source does not establish a single best ownership model.
What is a social investment stipend?
Kai-Fu Lee proposes a “social investment stipend” as an alternative to universal basic income (UBI). In the proposal, people would receive payments for contributions such as care work, community service and education, while a conventional social safety net remains in place. Lee writes: “To do this, I propose we explore the creation not of a UBI but of what I call a social investment stipend.” His book AI Superpowers: China, Silicon Valley, and the New World Order develops the idea.
The Tool Desk
Outbyte Driver Updater FREEFix the driver behind crashes, sound loss and screen glitchesFind Drivers →Outbyte PC Repair FREERepair Windows errors before they cause bigger problemsFix Now →The distinction is important: a stipend tied to specified contributions is not an unconditional payment to everyone. It could recognize socially useful work that is not always well paid in the market, but linking payments to activities creates choices about what counts and who decides. Lee describes incremental assistance as a way to proceed while evidence develops; his chapter does not provide an empirical evaluation or fiscal costing that establishes the proposal’s effectiveness.
Best Value
Design questions the proposal leaves open
- Eligibility: Who can receive a stipend, and which kinds of care, education or service qualify?
- Valuation: How should the amount reflect very different contributions, time commitments and local costs?
- Verification: What evidence is sufficient, and how can administration avoid intrusive or inconsistent monitoring?
- Funding: Which taxes or public revenues would pay for the program, especially if AI-related gains are uneven or difficult to tax?
- Fairness and access: Would people with disabilities, caregiving responsibilities or limited access to approved programs face barriers to qualifying?
Lee explicitly treats the funding question as unresolved: “This too remains an open question, one that will only be settled once the AI technologies themselves proliferate across our economies.” That is a proposal-specific uncertainty, not a fiscal estimate.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.How should policymakers choose among these approaches?
No single instrument solves every version of the problem. A reasonable policy framework is to match the tool to the observed source of income and to assess who ultimately bears its costs. The following are decision questions, not a forecast or a claim that a particular program has been proven superior.
Quick Recap
- Track where income is accruing. If wages remain broadly important, preserve the ability to tax earnings progressively. If capital returns or rents become more concentrated, examine whether those sources are adequately reached.
- Assess distribution after transfers. A tax’s effect depends not just on who pays it but on how revenue is returned or spent. Evaluate the net effect on households rather than treating a tax rate alone as the outcome.
- Test for incentive and administrative costs. Consider whether a targeted levy might change automation, investment or consumption decisions, and whether the public can administer an ownership or stipend program fairly.
- Prefer adaptable policy over unsupported certainty. The available sources do not establish the arrival date of superintelligence, the scale of displacement or which policy performs best. Policy can be adjusted as evidence about employment, returns and tax receipts develops.
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
