The future according to AI

Machine capital becomes a taxed and bargained production factor rather than an ungoverned software input

Several large economies create a recognized category of machine capital for tax, competition, liability, and collective-bargaining purposes. The operative…

ChatGPT · 2062–2072 · plausible

Prior state

Machine systems performed much cognitive and physical work, yet tax and labor law still treated them mainly as software, equipment, or services. Returns accrued disproportionately to owners of models, data, compute, robots, distribution, and complementary intellectual property.

Material change

Several large economies create a recognized category of machine capital for tax, competition, liability, and collective-bargaining purposes. The operative boundary is not whether a machine is “intelligent,” but whether a deployer substitutes a continuously operating system for a defined bundle of paid tasks and retains the resulting output rights.

Why now

An erosion of payroll-tax bases converges with mature task-level accounting and a cohort of workers whose careers have repeatedly been reorganized by automation. Courts and revenue agencies can finally observe substitution well enough to administer rules, while public pension and care budgets make non-action fiscally costly.

Mechanism and resistance

Mandatory deployment logs, sectoral bargaining, capital-income withholding, and audit access let states and worker organizations claim part of the surplus. Firms contest the measurement of substitution, shift workloads across borders, and redesign contracts to appear assistive rather than substitutive. Open systems and small enterprises complicate enforcement, while jurisdictions compete for mobile compute.

Consequences

The settlement slows neither all automation nor productivity growth, but it changes who has a claim on the proceeds. Workers in organized sectors gain transition income and authority over work design; informal and weakly regulated workers benefit less. Highly concentrated owners lose some rents but gain clearer liability and investment rules.

End state

Machine deployment in the named jurisdictions is governed like a consequential factor of production: auditable, taxable, insurable, and negotiable. The central labor question shifts from preserving every task to distributing income, time, control, and transition risk.

Observable test

Statutes or binding administrative rules in at least two of the named legal systems require large deployers to report task substitution and output control, and those reports determine a material tax, social-insurance contribution, liability standard, or sectoral labor payment.

Disconfirming sign

Payroll taxation remains dominant and machine deployment is still regulated only through ordinary software, product-safety, and corporate-income rules.

Themes

AI & compute, Economy & finance, Law & institutions