Book Review: When AI Shops by Geoff Gibbins

Geoff Gibbins’s When AI Shops is a brisk, timely, and practically oriented exploration of what he calls “Human-AI Commerce” — the idea that buying and selling is being rebuilt as a continuous, agentic system in which AIs learn, anticipate, evaluate, influence and execute transactions on humans’ behalf.

Gibbins grounds the central argument in a simple and effective conceit (the Martinez family) and then builds a clear operating model — the flywheel of Learn → Anticipate → Evaluate → Influence → Exchange — that recurs throughout the book, giving readers both a vocabulary and a set of concrete problems to solve. The book’s purpose is unapologetically pragmatic: it’s written for business leaders who need to make decisions now, and Gibbins delivers playbooks, pilot designs, audit questions and metrics that teams can act on immediately.

Structurally the book is smartly organized: Part I sets up the diagnosis and the flywheel; Part II dives into discovery, the “two-sided web,” and the surprisingly odd psychology of agentic decision-making; Part III examines industry impacts, employment effects and systemic risks; and Part IV gives roadmaps for individuals and organizations plus five compelling scenarios of how commerce might evolve by 2035.

The Martinez vignettes work as recurring anchors that show how the same capabilities manifest in everyday life (from replacing cleat shopping with a 12-second decision to planning an entire vacation), and the scenarios — from “Harmony Garden” to the unsettling “Synaptic Economy” — are useful thought experiments for boardrooms and product teams alike.

The book’s strengths are severalfold. First, its synthesis: Gibbins pulls together examples from Amazon, Starbucks, Mastercard and many enterprise systems and stitches them into a coherent narrative that is actionable rather than merely descriptive. Second, its practical orientation: chapters on machine-readability, agentology, audit checklists (e.g., structured data, Schema.org, visibility) and phased pilots make the book more of a tactical manual than a high-level manifesto. Third, the ethical and organizational frame is sober — he repeatedly stresses “urgent patience” and governance before scale, and he offers readable exercises (three-week trust experiments, delegation levels) that make the abstract stakes very tangible. These practical takeaways and roadmaps are among the book’s most valuable contributions for practitioners.

At the same time, the book has limitations worth noting. Because its examples are intentionally recent (late-2024/2025) the statistical claims and platform citations risk dating quickly; readers will need to treat specific percentages and vendor features as snapshots rather than enduring truths. The emphasis on enterprise strategy and platform mechanics sometimes sidelines deeper engagement with public policy, antitrust dynamics and the lived experience of lower-resourced communities — topics that are touched on (for example, distributional effects and the “human premium”) but could use fuller treatment given their importance. There are also normative tensions that Gibbins wrestles with but cannot fully resolve — for instance, the balance between effective influence and covert manipulation — which means leaders will still need to translate ethics into enforceable rules inside their organizations.

Two other points we found noteworthy: Gibbins is candid about using AI in the book’s research and drafting process and frames that as an explicit human-in-the-loop collaboration, which mirrors the book’s larger argument about hybrid human-AI teams; that transparency is useful for readers evaluating the book’s provenance and for organizations thinking about their own disclosure practices. Second, his scenarios (including the provocative Synaptic Economy) are valuable both for stress-testing strategy and for prompting boards to ask which future they want to help create — not simply which one will happen.

Verdict: When AI Shops is an excellent operational primer for executives, product leaders, and strategy teams who must reckon now with the rise of agentic commerce. It succeeds best as a playbook — clear frameworks, concrete checklists, and course-of-action chapters that you can hand to a cross-functional team and use to run pilots. If you want deep academic theory or exhaustive public-policy prescriptions, you’ll need supplemental reading; but if your priority is understanding how commerce is changing and what to do about it this quarter and next, this book delivers.

Rating: ★★★★☆ (4.5/5)

Highly recommended for business and product audiences, with a small caveat about the need to pair it with policy and equity analyses for long-term societal planning.