AI and the End of Work: Is Technology Breaking Capitalism? - Sri Lanka Guardian

TEXT START: Daniel McKinnon knows better than most that artificial intelligence can change a career.

The Dissection

The article correctly identifies the visible symptoms: cognitive automation, concentrated ownership, weakening job security, and the failure of education to guarantee stability. It then retreats into a governance frame, presenting the crisis as a distribution problem that governments might solve through taxation, UBI, public ownership, or universal basic capital.

That is transition management. It describes the machinery dismantling the old order, then asks whether administrators can repaint it quickly enough.

The Core Fallacy

The central error is assuming that redistribution can preserve the wage-based social contract after human labour loses its productive necessity.

The article treats AI as an unusually powerful technological disruption within capitalism. Under the Discontinuity Thesis, AI attacks the system’s load-bearing circuit: mass employment produces wages, wages produce consumption, and consumption legitimises continued accumulation. Once AI performs cognitive work more cheaply and effectively, firms are compelled by competition to reduce their dependence on human labour. No tax scheme changes that underlying pressure.

UBI, dividends, or universal basic capital may preserve consumption. They do not restore productive participation, bargaining power, status, or control. They are life support for demand, not a cure for labour obsolescence.

The compensation theory analogy is especially weak. Earlier automation displaced tasks while leaving humans broadly necessary elsewhere. Autonomous cognitive systems can attack the coordination, analysis, judgement, and expertise that supplied the replacement occupations. The article sees the breach but still assumes another labour market will emerge behind it.

Hidden Assumptions

• New occupations will appear at sufficient scale to absorb displaced workers.
• Human expertise will remain scarce enough to command wages.
• States can coordinate globally against firms controlling models, chips, platforms, compute, and infrastructure.
• Concentrated AI capital can be taxed or partially socialised without capital escaping, retaliating, or becoming politically sovereign.
• Ownership of financial claims will substitute for ownership of productive control.
• Income transfers can preserve the social contract even after work ceases to provide necessity, identity, and status.
• Institutional lag can become reversal rather than merely delay.
• The relevant metric is unemployment, when displacement also occurs through hiring freezes, wage compression, task elimination, and the conversion of full occupations into supervised exceptions.
• Education and hard work remain viable routes to security, despite the article’s own evidence that the credential pipeline is already producing stranded labour.

Social Function

Primarily: partial truth, transition management, and ideological anesthetic. Secondarily: elite self-exoneration.

The article is useful because it acknowledges that AI ownership matters and that the old promise of education-to-security is failing. But its repeated language of possibility, redesign, and broad distribution keeps the reader focused on policy adjustment rather than the harder structural question: who controls the productive system once human labour is no longer required?

It allows the beneficiaries of automation to appear responsible without requiring them to surrender command. Tax the gains, distribute some capital, and the machine can keep operating under the same ownership hierarchy. That may stabilize consumption while leaving the population economically subordinate.

The Verdict

This is a serious symptom report with an incomplete autopsy. It correctly sees that AI threatens the connection between work, income, and social legitimacy, but it mistakes the impending death of mass-employment capitalism for a solvable inequality crisis.

Under DT mechanics, P1 makes cognitive automation economically dominant, P2 prevents institutions from preserving large-scale human-only economic domains, and P3 destroys majority access to necessary labour. The post-WWII order therefore dies even if governments distribute enough money to keep people buying.

The decisive divide will not be between employed and unemployed. It will be between Sovereigns who own and control AI capital, Servitors who remain indispensable to them, and the displaced population managed through transfers. The article’s proposed remedies may delay disorder. They do not restore the old system. They are hospice care for the wage-consumption circuit.