Apollo Economist Says AI Affecting Wages More Than Employment - PYMNTS.com

URL SCAN: Apollo Economist Says AI Affecting Wages More Than Employment - PYMNTS.com
FIRST LINE: Artificial intelligence is affecting the labor market, though not in the way people might think.

The Dissection

This is a normalization piece disguised as empirical caution. It takes early labor-market data—AI-exposed wages growing 6.7% more slowly, especially among lower-income workers—and frames the absence of a broad employment collapse as the primary fact. It then uses business formation as a counterweight, implying that new firms and a “more dynamic economy” can absorb the shock. It documents the first visible symptom of displacement while treating the missing terminal symptom as reassurance.

The Core Fallacy

It confuses employment counts with economic viability.

A worker can remain employed while losing wage growth, bargaining power, hours, training, promotion paths, and future replaceability. Slower wage growth is not benign because payroll headcount remains positive; it is the price signal of labor becoming less scarce and less valuable. For lower-income workers, this is the front edge of exclusion.

The article also treats new business formation as evidence that AI is creating offsetting opportunity. That only proves AI is lowering the cost of starting businesses. It does not prove those businesses will require mass human labor, pay viable wages, or restore productive participation. Under P1, more firms can operate with fewer workers. Under P2, workers cannot coordinate a protected human-only economic domain. Under P3, payroll survival can coexist with the collapse of the wage-consumption circuit.

Hidden Assumptions

• Jobs retained today will remain economically secure tomorrow.
• Employment totals are sufficient to measure worker viability.
• Wage suppression is a temporary adjustment rather than evidence of declining labor scarcity.
• AI-created businesses will generate enough human demand to offset substitution.
• Aggregate payrolls reveal no meaningful deterioration in job quality, training, or progression.
• Lower-income workers can absorb disruption despite less training and thinner financial buffers.
• The lag between mechanical displacement and official employment data is evidence against displacement.

Social Function

Primary classification: ideological anesthetic with partial truth.

The partial truth is that the supplied evidence shows AI weakening wage growth more clearly than reducing total employment, while business formation is increasing. The anesthetic is presenting this as a manageable transition rather than as early-stage labor devaluation. It gives institutions permission to call the injury “adaptation” until the lag defenses fail.

The Verdict

The article records the beginning of the kill mechanism and mistakes the delay for a contradiction. The critical number is not the 0.2% job decline; it is the 6.7% slower wage growth among AI-exposed workers, concentrated among people with the least buffer. Employment is the lagging indicator. Wages, training, and bargaining power are the leading edge.

Business formation is not a rescue mechanism unless it restores broad access to productive income. On the supplied evidence, it is equally compatible with an economy where AI makes firms easier to launch and humans easier to underpay. The post-WWII circuit is not dead because every job vanished; it is dying because labor is losing the power to command the income needed to sustain that circuit. This is an early autopsy report mislabeled as reassurance.