The economist who was terrified of AI just found a rare reason for hope

The article argues that AI may not simply destroy demand and jobs, but instead shift spending and employment toward a "relational sector" where human interaction remains scarce and valuable. It cites Starbucks as an example of consumers preferring human touch over full automation, and says jobs in nursing, teaching, therapy, childcare, and hospitality may become more valuable as routine tasks are automated. The near-term market relevance is mostly interpretive rather than event-driven, though Morgan Stanley has flagged rising AI labor disruption risk as LLM capabilities improve.

Analysis

The market implication is not that AI destroys demand outright, but that it reallocates spend toward scarce human interface and trust layers. That is a structural tailwind for consumer services with high emotional content and low substitutability, while pure workflow automation remains vulnerable to a slower-than-advertised adoption curve. The key second-order effect is margin bifurcation: firms that can blend AI efficiency with a premium human wrapper should see pricing power expand, while “just efficient” businesses risk commoditization and lower lifetime value.

For SBUX, the signal is less about coffee and more about the value of atmosphere, ritual, and proximate human interaction in a world where convenience alone is easy to copy. That supports a modest multiple re-rating if management can prove that labor intensity is not a cost but a product feature that drives frequency and mix. The real risk is that investors extrapolate this into a broad retail-services renaissance before labor inflation and traffic trends validate it; the stock can still underperform if transactions weaken faster than experiential pricing can offset.

MS is a more nuanced beneficiary: the article reinforces demand for high-touch advisory, distribution, and relationship management, but it also highlights that internal AI productivity can compress fees in low-differentiation workflows. Net, AI should widen the gap between elite franchise banks and commoditized competitors, because the scarce asset becomes client trust and deal access, not code generation. The contrarian miss is that consensus may be too focused on workforce displacement and not enough on which jobs become more valuable when machines handle the easy parts; however, the transition is likely uneven and highly sensitive to adoption speed, so the next 6-12 months are about experimentation, not full economic re-rating.

Key Decisions for Investors

• Long SBUX vs. short a broad restaurant/retail basket for 3-6 months: if the market increasingly prices human-experience premiums, SBUX should hold up better than names whose differentiation is mainly convenience; target 8-12% relative outperformance, stop if same-store traffic deteriorates materially.
• Buy MS on 6-12 month dips, but hedge with a basket short of lower-quality regional brokers: AI should compress lower-value processing work first, widening franchise dispersion; upside is multiple resilience, downside is fee pressure if capital markets activity slows.
• Initiate a small long in labor-sensitive service winners and fund it by shorting pure automation beneficiaries that lack a human moat; use a 3-6 month horizon and size modestly because the thesis depends on slower adoption than headlines imply.
• Use SBUX call spreads rather than outright equity if positioning for a ‘scarcity of human connection’ re-rating; 6-9 month structure offers defined risk if the market rotates back to growth-at-any-cost and labor costs squeeze margins.