The claim: for two centuries we've distributed income mainly by paying people for work. If AI takes over enough cognitive work, that mechanism stops carrying the load — and ownership, not wages, becomes the thing that decides who gets what.
The core of it
- Labor's share of income falls. The slice of the economy that reaches people as wages shrinks as machines do more of the work.
- Capital becomes the binding constraint. If income flows to whoever owns the systems, then concentration of ownership is the inequality story.
- New plumbing is required. Universal basic income, universal basic equity (giving people shares rather than cheques), sovereign wealth funds, AI dividends.
- This wave is different. Earlier automation hit blue-collar work. This one hits both at once, which removes the usual escape route of retraining upward.
The loops that make it self-reinforcing
The most useful part of David Shapiro's version is that it's structural rather than moral. He maps it as feedback loops, and the asymmetry is the point.
Two that destabilize, and run on their own:
- The deflationary spiral. Machines replace workers → wages erode → people buy less → company revenue falls → margin pressure → more automation. Round and round.
- The inequality amplifier. Substitution concentrates income with owners → owners spend a smaller share of it → demand sags → more substitution.
Two that stabilize, and have to be built on purpose:
- The capital stabilizer. Spread ownership wide enough and capital income reaches households, so demand holds up.
- Citizen leverage. New technical levers give people bargaining power, which produces responsive policy, which broadens ownership further.
And one slow one underneath: leverage erosion. As AI reduces what the state needs from its citizens, while citizens grow more dependent on transfers from the state, the power balance tilts.
The insight worth keeping: the destabilizing loops are automatic. The stabilizing ones are deliberate. Nothing builds them by default, so the race is between how fast substitution happens and how fast institutions adapt.
Bilateral dependence
The sharpest and most speculative piece of the argument. The social contract has always rested on mutual need: states need production and service from citizens; citizens need protection and functioning markets from states.
AI and robotics erode the first half. If autonomous systems supply the labor and the military capacity, citizens become, in Shapiro's phrase, financial outpatients of the state. Historically, ordinary people's leverage came from the ability to withhold work: strikes, unions, stoppages. That lever needs the state to need the labor.
His proposed replacement is "algorithmic rights" — the ability to interrupt financial flows, force transparency, control one's data — building leverage into the same technology that removed the old kind. It's the least-proven part of the framework. There's no real institutional precedent, with data-protection law as at best a partial parallel.
Who argues it
- Sam Altman — "Moore's Law for Everything" and universal basic equity: give people ownership rather than payments.
- David Shapiro — the most architecturally complete version. A "full stack" of capital broadening: sovereign wealth funds, state and local funds, employee ownership, baby bonds.
- Yanis Varoufakis — argues capitalism has already been replaced by "techno-feudalism," where platforms collect rent like fiefdoms and users supply unpaid labour. His remedy is collective ownership of that cloud capital rather than regulating around it.
- Daniel Susskind — A World Without Work, the accessible book-length treatment.
- Nick Bostrom — Deep Utopia (2024) takes the question past economics: what is a life worth living when nothing requires effort?
- Vitalik Buterin — decentralized mechanisms for distribution and governance, including prediction markets as allocation tools.
Different lens
- Pope Leo XIV — agrees wage labor is under strain, differs on the remedy. Dignity of work and mediating institutions rather than a basic income.
- Shoshana Zuboff — the transition is already underway, but as behavioral extraction replacing labor extraction. Users have been working unpaid for years.
- Kate Crawford — "post-labor" hides the labor that remains: annotation, moderation, mining. Post-labor for some, brutal labor for others.
- Peter Thiel — rejects redistribution as treating symptoms. Build new industries instead of dividing the existing pie.
- Peter Diamandis — post-labor is a waypoint to post-scarcity, where the arithmetic changes anyway.
Who pushes back
- Daron Acemoglu — the displacement is overstated. Only a small share of tasks will be profitably automated this decade, and institutions can steer the technology toward helping workers.
- Tyler Cowen — diffusion is slow and total replacement unlikely. Human judgment gets scarcer and therefore more valuable.
- Erik Brynjolfsson — rejects the labor-share-to-zero premise outright. See the augmentation thesis.
- Marc Andreessen — rejects the framing entirely; markets handle distribution by creating new categories of work.
- David Autor — about 60% of current employment is in jobs that didn't exist in 1940. New tasks keep appearing.
The evidence
Thin on the central claim, which is worth saying plainly. What exists are proof-of-concept fragments rather than a tested model:
- The Alaska Permanent Fund, Norway's sovereign wealth fund and New Mexico's land grant fund all show that broad public ownership can pay households directly at scale.
- Employee ownership research suggests shares build wealth differently than transfers do.
- Household spending is more than 70% of GDP, which is the empirical anchor under the demand-collapse argument: if wages fall broadly, there isn't an obvious other buyer.
Open questions
- Does labor's share actually fall, or do new tasks keep appearing fast enough?
- Is large-scale redistribution politically achievable anywhere, on any timeline?
- Who ends up owning the AI: a few firms, or widely?
- Can "algorithmic rights" be built into real institutions, or is it a thought experiment?
- Does a meaning economy — care, teaching, craft — absorb enough displaced people to matter, or is human attention the hard ceiling?
All theories · clickbridge.com · Rich Price · last reviewed September 2026 · People linked here have their own pages at people.clickbridge.com