AI rebound effect: Why cheaper services could mean more demand, not fewer jobs

TEXT START: A few years ago, a business website was a big investment.

The Dissection

The article argues that AI will reduce service costs, expand demand, and shift human value toward judgment, explanation, trust, and accountability. Its examples identify genuine niches, but quietly convert those niches into a theory of broad employment survival.

The Core Fallacy

It confuses more services with more human labor. AI can generate vastly more websites, legal analyses, images, and reports while requiring fewer human hours per unit. A larger market does not restore the mass employment → wage → consumption circuit. “More output” is not “more economically necessary people.”

The article also assumes that judgment and responsibility will remain labor-intensive. Under the Discontinuity Thesis, those functions can become narrow supervisory layers, standardized processes, automated verification, or privileges concentrated among AI-capital owners. The rebound effect expands production; it does not guarantee human bargaining power.

Hidden Assumptions

• Demand will expand indefinitely rather than saturate or collapse in price.
• Authenticity, trust, and human judgment will remain difficult to automate.
• Businesses will preserve human checkpoints despite competitive pressure to remove them.
• New high-value roles will scale fast enough to absorb displaced workers.
• Displaced workers can access those roles rather than being excluded by skill, capital, or ownership barriers.
• More client volume will require proportionally more people instead of merely more AI-supervised output.
• Increased consumption will preserve wage-based participation rather than become dependent on transfers or concentrated ownership.

The web-development example proves only that demand can rise when costs fall. It does not prove that the sector still needs the same number of humans, or that this pattern generalizes across the economy.

Social Function

Classification: partial truth, transition management, and ideological anesthetic.

The article usefully warns companies not to automate away customer-facing value. But it treats temporary human moats as a systemic solution while avoiding the decisive question: who owns the AI capital, and how many people remain economically necessary? It offers adaptation advice to firms where a mass-participation crisis requires an ownership analysis.

The Verdict

The rebound effect is real but strategically overrated. AI may create more demand for services while destroying the labor required to produce each one. Continuous website updates, higher radiographer throughput, and cheaper legal advice can all coexist with a shrinking human labor share. This article describes survivable niches for Sovereigns and indispensable Servitors—not a rescue of post-WWII capitalism.