Why the AI economy is like a bad dating app — drowning in decks, pilot purgatory - Fortune

TEXT START: Amy Webb was on her long Sunday bike ride, the one she takes when she’s not training for a race, when the thought arrived fully formed.

The Dissection

The text is an autopsy of enterprise AI adoption failure, not of the AI economy. It accurately documents the first-order symptoms: pilot purgatory, exploding verification costs, decision paralysis, generic output, weak integration, and executives unable to identify where saved capacity should go.

But it keeps the analysis trapped inside the existing corporate order. The question is framed as how companies can productize AI efficiently, not whether AI destroys the labor relationship that sustains mass purchasing power. “The economy makes weird decisions” is a boardroom description of systemic disorientation, not an explanation of the underlying break.

The Core Fallacy

The central error is treating implementation friction and poor short-term returns as the main economic story. Under the Discontinuity Thesis, expensive integration is a lag defense, not a refutation. Legal bottlenecks, IT failures, pilot churn, and managerial incompetence delay automation; they do not preserve human economic necessity.

The article also treats AI-cited layoffs as meaningful evidence. They are not. Firms can eliminate labor through restructuring, overstaffing corrections, attrition, or productivity targets without labeling the cuts “AI.” Layoff attribution measures corporate messaging, not the erosion of labor demand.

The decisive question is absent: when a small number of workers and AI systems can produce what formerly required entire departments, what happens to the wages that purchase the output? Missed targets are Wall Street’s reckoning. Productive participation collapse is the systemic reckoning.

Hidden Assumptions

• Corporate AI value is adequately measured by near-term ROI and budget performance.
• Productivity gains will be redeployed into new human jobs rather than used to reduce headcount.
• Better planning and more AI expertise can repair the transition.
• Human creativity, judgment, and ownership remain durable economic scarcities.
• Firms can manage adoption deliberately despite competitive pressure to automate.
• AI displacement must be explicitly identified in layoff announcements to be economically real.
• Employment, wages, and consumption remain structurally coupled without requiring analysis.

Social Function

Classification: partial truth, transition management, elite self-exoneration, and ideological anesthetic.

The article provides a useful warning about wasteful experimentation and unpriced organizational costs. Its larger function is more comfortable: it converts a structural labor shock into a consulting problem involving bad workflows, inexperienced boards, verbose models, and insufficient governance. That lets executives continue buying AI while postponing the ownership question and the fate of displaced workers.

It is not pure copium because it predicts a reckoning. It is more sophisticated than that: it narrows the reckoning to missed targets, wasted budgets, and managerial embarrassment, leaving the wage-consumption circuit outside the frame.

The Verdict

The text sees the smoke but misidentifies the fire. Pilot purgatory and drowning in decks are real symptoms of institutions trying to digest a technology that cheapens cognitive production while preserving obsolete organizational rituals. They are friction in the transition, not the terminal disease.

The article’s “AI bubble” is a spending and governance bubble. The Discontinuity Thesis concerns the deeper rupture: AI severs mass employment from productive necessity, then severs wages from consumption. This piece describes the hospital bills of the old system while refusing to name the patient’s terminal condition.