‘The amazement is kind of wearing off a little bit’: Wharton survey reveals managers are learning what Gen Z already figured out

A Wharton/GBK survey found daily AI use among managers and above rose to 62% from 46% last year, while excitement fell 8 percentage points to 51%; 80% reported positive financial returns, but only 28% called them strongly positive. Seventy-two percent said their focus had shifted to managing AI-generated outputs, and 48% said AI does the work with a human approving it. One-third reported AI-related hiring slowdowns or freezes, while departments measuring growth and redesigning workflows were more likely to report strong returns; the findings indicate implementation challenges and varied outcomes, not a uniform effect across companies.

Analysis

The investable signal is not declining AI use; it is a shift in where value accrues. If human review remains necessary for most consequential workflows, near-term AI deployment can increase demand for verification, governance, security and workflow redesign while delaying labor-cost savings. That weakens the simple “more agents = fewer workers = immediate margin expansion” thesis and raises the risk that model and infrastructure spending outruns measurable customer ROI.

The strongest returns reported alongside workflow redesign and growth metrics are suggestive, not causal: the survey uses separate samples, self-reports, and no matched company comparisons. Treat it as evidence that implementation capability may differentiate adopters—not proof of realized, durable earnings uplift. Near term, watch enterprise software budgets and AI vendor pricing for signs that buyers are reallocating spend from experimentation to integration and controls. Over 1–3 months, earnings disclosures on AI-related revenue, labor productivity and implementation costs matter more than usage anecdotes. Over 6–18 months, a downturn could accelerate redesign and hiring substitution, but also pressure discretionary technology budgets.

Contrarian angle: cooling enthusiasm may be healthy normalization rather than demand destruction. More routine use alongside lower excitement can broaden adoption; however, value may accrue to firms redesigning processes, not automatically to model providers. Ford is cited only as a historical analogy; the article provides no Ford-specific AI evidence or basis for a company trade.

Key Decisions for Investors

• No direct trade in Ford (F) on this article: its mention is an analogy, not company-specific news.
• Prefer a watchlist of enterprise workflow, cybersecurity, governance and verification providers over a blanket long on AI infrastructure. Require evidence of customer renewals, AI-linked revenue or lower delivery costs before adding exposure.
• Avoid underwriting near-term margin expansion from headcount reduction alone. Track disclosed implementation and oversight costs alongside productivity, hiring and revenue-conversion metrics; persistent review bottlenecks would falsify the rapid labor-savings thesis.
• Reassess over the next 1–3 months as company earnings clarify whether AI spending is shifting from pilots to redesigned, revenue-generating workflows. Strong usage without improved unit economics is a caution signal; credible growth or productivity gains with controlled costs would support selective longs.