What to Watch: AI's Impact on the Economy - WWD

TEXT START: Artificial intelligence just might consume the world, if not literally — although some aren’t counting out that possibility — then figuratively.

The Dissection

The article packages a structural break as an investment story. It inventories current layoffs, hiring, GDP effects, data-center spending, and fashion firms’ weak AI returns, then uses present-tense friction to imply that the labor market remains intact. Its real function is temporal deferral: the dangerous conclusion is pushed into “next year” and “the year after that,” while current disruption is labeled evolution.

The Core Fallacy

It treats incomplete deployment as evidence against eventual displacement. The fact that AI has not yet dismantled employment does not show that it cannot; it shows that implementation is still uneven and the lag defenses remain active. Aggregate hiring and GDP figures measure the transition’s early investment phase, not whether human labor remains economically necessary after AI achieves durable cost and performance superiority.

The article also mistakes gross activity for productive participation. Data-center construction, AI financing, and stock-market wealth can expand economic measurements while concentrating ownership and hollowing out the wage-to-consumption circuit. A growing economy can therefore coexist with a collapsing labor bargain.

Hidden Assumptions

• Current hiring patterns will persist after AI systems mature and diffuse across workflows.
• Weak short-term return on investment means weak structural threat.
• Aggregate employment data captures who is becoming unnecessary and who controls the replacement capital.
• Higher stock prices and affluent consumer spending can indefinitely compensate for lost labor income.
• Physical infrastructure, borrowing costs, and organizational friction will meaningfully constrain deployment rather than merely delay it.
• “Evolution, not revolution” describes the endpoint rather than the calm before threshold effects accelerate.

Social Function

Primarily transition management and ideological anesthetic, with a layer of partial truth. The article correctly acknowledges real layoffs, uneven adoption, uncertain returns, and mounting capital requirements. But those facts are arranged as reassurance: because the machine has not yet finished the kill, the patient is declared stable. It gives investors and incumbent firms permission to treat systemic labor displacement as a future scenario while the ownership structure consolidates in the present.

The Verdict

A soft-focus collapse memo. Its data records the opening movements of obsolescence, but its interpretation confuses lag with survival. “Something will have to give” is the article’s buried admission; it simply refuses to identify the thing most likely to break: mass human access to economically necessary work.