Workers feared AI would kill jobs. Instead, 72% of employers say they'll pay more for AI skills

TEXT START: Workers are on edge that AI will swoop up their jobs and make it harder for them to find another.

The Dissection

This article converts a narrow scarcity premium—employers paying more for scarce AI-capable recruits—into a mass-labor reassurance story. Its actual signal is harsher: firms are rewarding the small group that can make AI productive while using the technology to reduce the need for everyone else.

The Core Fallacy

It confuses a temporary scarcity premium with durable, broad-based employment. AI skills command money because supply is thin and firms are racing to capture returns. Under P1, that scarcity invites automation, standardization, and software embedding. As AI capability spreads, the premium compresses while displacement expands.

The claim that 36 million jobs will disappear while 41 million appear is gross arithmetic, not continuity. It ignores timing, access, geography, credential barriers, pay, and whether displaced workers can reach the new roles. More jobs is not the same as the same people remaining economically necessary.

Hidden Assumptions

• Employer intentions become durable wages rather than recruiting signals or temporary bidding.
• New AI roles scale faster than AI destroys adjacent roles.
• Workers can become fluent before their underlying tasks are automated.
• New jobs are accessible to displaced workers rather than concentrated among owners, engineers, and insiders.
• Productivity gains flow to labor instead of capital.
• Institutions can preserve stable human-only economic domains at scale, contradicting P2.
• A $177,000 average AI posting represents a normal worker’s path rather than a concentrated market.
• Gross job counts measure productive participation. They do not.

Social Function

Primary classification: copium and transition management, with elite self-exoneration.

The article tells workers to adapt individually, allowing employers to present capital substitution as empowerment. The pay premium is real, but it is also bait: it rewards the narrow class currently needed to install the machinery that makes broader human labor less necessary.

The Verdict

This is a partial truth packaged as a false systemic conclusion. Some workers will profit during the AI buildout. That does not rescue the post-WWII employment circuit; it confirms its fracture. A thin layer of AI owners, operators, and indispensable implementers gains leverage while the majority become replaceable, transfer-dependent, or excluded. The article mistakes wages paid to the technicians installing the guillotine for proof that the guillotine is creating jobs.