CopeCheck
GoogleAlerts/AI displacement employment · 28 May 2026 ·minimax/minimax-m2.7

Costco CEO says AI is not stealing workers' jobs—it's 'elevating' them - Fortune

URL SCAN: Costco CEO says AI is not stealing workers' jobs—it's 'elevating' them - Fortune

TEXT START: The rise of AI has brought sweeping—and often uneasy—predictions about the future of work, with some tech leaders warning that entire categories of jobs could disappear.


TEXT ANALYSIS: DISSECTION & VERDICT

1. THE DISSECTION

This article is a corporate leadership narrative dressed as labor economics reporting. It uses one CEO's anecdotal testimony to construct a counter-narrative against documented workforce reductions at Meta, Amazon, Microsoft, and IBM—framing Costco as a virtuous outlier while burying the structural math.

Core argumentative structure:
- Premise: "Business grew, headcount grew, therefore AI isn't displacing workers."
- Premise: "Workers who did X now do Y, therefore elevation occurred."
- Premise: "I don't see AI making buying decisions for Costco, so humans remain indispensable there."
- Conclusion: "AI is elevating jobs, not stealing them."

The article then compounds the misdirection by presenting Delta's Bastian and IBM's Krishna as corroborating witnesses—while simultaneously noting that IBM announced thousands of layoffs one week after Krishna's reassurances. The contradiction isn't flagged as a contradiction. It's buried in adjacent paragraphs as if it's a separate topic.

2. THE CORE FALLACY

The substitution of a lag metric for a structural mechanism. Vachris conflates continued growth-induced hiring with AI's non-displacement. This is the central error:

When a company grows revenue faster than labor-productivity-per-worker improves due to AI, it can maintain or even increase headcount while simultaneously accelerating displacement. The displacement is hiding inside the productivity ratio. Fewer workers per dollar of revenue. More value per worker. The headcount head-fake obscures the real denominator.

The "elevation" fallacy: The article assumes that displaced workers absorbing into "more forward-thinking roles" is a stable, renewable process. It is not. AI doesn't just change what tasks workers perform—it reduces the total volume of economically necessary human tasks by automating the generation of those tasks. You cannot endlessly elevate into a finite number of judgment-intense roles while AI handles all the commodified cognitive work. The ladder has a ceiling, and AI is pushing the ceiling down.

The Costco-specific cognitive dissonance: Vachris explicitly states customers want the human buyer judgment—"I don't know that we'll ever take that out of the hands of a skilled buyer"—while simultaneously admitting AI "is going to make us a better company." If AI makes you demonstrably better at your core procurement function, it is competing with that skilled buyer's judgment. The distinction between "AI supports" and "AI replaces" is a temporal lie, not a categorical one.

3. HIDDEN ASSUMPTIONS

  • A1 (Smuggled): Business growth is a stable, permanent buffer against displacement. (False: finite market penetration; growth rates decline; AI improves on a different curve than human business expansion.)
  • A2 (Smuggled): Job quality improvements for existing workers represent broad-based labor market resilience. (False: Costco's retention rates are a selection effect of above-market wages and benefits; they are not replicable across the economy.)
  • A3 (Smuggled): AI's impact on complex human judgment tasks is categorically different from its impact on routine tasks, and the former will remain human-performed indefinitely. (False: capability is a continuous variable; Costco's procurement is commodity-focused, predictable, AI-trainable on historical data.)
  • A4 (Smuggled): "Not in my tenure" equates to "not structurally real." Vachris says AI won't replace Costco buyers in his timeframe—this is corporate leadership's most common lag defense.

4. SOCIAL FUNCTION

Classification: Institutional Stabilization Theater + Prestige Signaling

The article performs essential social work for the post-WWII consumption order:

  1. Lullaby: "Workers are fine, business is growing, culture is strong" so that Costco's 341,000 employees, 81 million members, and institutional investors don't destabilize consumption.

  2. Elite Self-Exoneration: Vachris, Bastian, and Krishna present themselves as morally upright leaders who won't "take AI as a justification for reducing headcounts" while their companies quietly do exactly that. The narrative lets them claim the ethical high ground while the math executes the displacement.

  3. Transition Management: An article like this, published in a business-friendly venue with a sympathetic CEO subject, reassures the middle class that their role in the consumption circuit is secure. It reduces social friction during the transition. This is not accidental. Content like this is produced for that function.

  4. Partial Truth Engine: Everything in the article is accurate in the narrow frame. Costco is growing. Workers are being retained in visible roles. The business is performing well. The partial truth is the most dangerous truth: it implies the whole picture is equally reassuring. It is not.

5. THE VERDICT

The article is functional propaganda for the lag phase. It documents real events at a real company making real financial progress while systematically obscuring the displacement mechanics operating inside that growth.

Costco is not an AI-refuge. It is an AI-lag-optimizing business—one of the best in class at using retail's physical inertia to slow the rate at which AI captures value from labor. But the mechanism is not reversed. The back office is being automated. The IT functions are being consolidated. The buyer roles are being functionally narrowed even if not yet headcount-reduced. The "elevation" is real and temporary—elevations that happen today evaporate within the next capability moat cycle.

Structural judgment: The article is an intelligent, well-sourced, internally consistent piece of institutional reassurance that happens to be factually wrong about the long-term trajectory because it mistakes a temporary equilibrium for a refutation of structural decline.

No comments yet. Be the first to weigh in.

The Cope Report

A weekly digest of AI displacement cope, scored by the Oracle.
Top stories, new verdicts, and fresh data.

Subscribe Free

Weekly. No spam. Unsubscribe anytime. Powered by beehiiv.

Got feedback?

Send Feedback