Is AI really taking young people's jobs? | The Spectator Australia

TEXT START: We’ve had three announcements this week that paint a worrying picture for the job prospects of young people.

The Dissection

The article documents a genuine collapse in youth entry points, then performs causal displacement. It accepts that AI may contribute but relocates primary blame onto National Insurance, wages, regulation, and business costs. That produces a comforting diagnosis: reverse government policy and the old hiring ladder returns. Under the Discontinuity Thesis, this is transition management disguised as investigation. It mistakes the most visible accelerant for the engine of the collapse.

The Core Fallacy

The article commits an inverse version of the lazy AI headline. It is right that the supplied data do not prove AI caused every lost job. It is wrong to infer that policy pressure therefore explains the structural decline.

AI often strikes first at the hiring margin. Firms do not need to fire every junior worker; they can stop recruiting juniors, increase senior-worker output with AI, and let attrition erase the entry-level layer. The absence of an AI label in vacancy data proves nothing about the mechanism. Retail and hospitality losses show broader demand and cost stress, but they do not exonerate AI from destroying cognitive entry routes elsewhere.

Government policy can accelerate P3 by making human labor more expensive. It cannot reverse P1. Once AI makes work cheaper, faster, and less dependent on inexperienced humans, cutting payroll taxes does not recreate tasks that no longer require people. The article replaces a simplistic AI monocause with a simplistic policy monocause.

Hidden Assumptions

• If employer costs fall, firms will restore youth hiring rather than direct savings into automation, margins, or fewer workers.
• Entry-level jobs remain economically necessary, rather than being training scaffolding that AI can remove.
• Existing sectors are a reliable measure of AI’s impact, despite AI first changing the composition of work and future hiring.
• Human labor retains durable price and performance competitiveness against AI.
• Policy can preserve large human-only economic domains at scale, contradicting P2.
• The employment system can return to its pre-discontinuity equilibrium once fiscal pressure is relaxed.
• The problem is a temporary shortage of jobs, not the loss of economically necessary participation for a growing majority.

Social Function

Classification: partial truth functioning as ideological anesthetic and elite self-exoneration.

The article correctly identifies taxes, wage floors, and regulation as real short-term pressures on marginal hiring. That truth gives the argument credibility. Its deeper function is to make the crisis appear politically reversible. Blaming Westminster is psychologically and electorally useful because governments can be replaced; admitting that productive participation is being structurally automated offers no such clean repair. The firms appear as squeezed victims, while the machine’s assault on labor necessity is reduced to a frightening narrative device.

The Verdict

This is an accurate account of several accelerants wrapped around a wrong diagnosis of the disease. Government labor costs may kill the weakest human jobs first, but AI is what makes the lost entry-level work increasingly unnecessary. Repealing those policies might reopen a few low-value positions; it will not restore the mass employment–wage–consumption circuit. The apprenticeship ladder is not merely being taxed out of existence. It is being rendered obsolete from above.