Pages
172
Published
2022
How AI Shifts Decision-Making Power Across Industries and Organizations
Understand how AI redistributes power inside firms and across markets so you can position your organization before the shift happens to you.
AI is not just a productivity tool — it is a power transfer mechanism. Economists Ajay Agrawal, Joshua Gans, and Avi Goldfarb argue that cheap prediction, the core output of modern AI, rewires who decides what inside organizations and across entire industries. This book gives you a rigorous but accessible framework for anticipating those shifts before competitors do, drawing on economic theory grounded in real industry examples.
Most AI conversations focus on the technology. This book focuses on the consequences. Economists Ajay Agrawal, Joshua Gans, and Avi Goldfarb — authors of the earlier Prediction Machines — return with a harder question: when AI makes prediction cheap and abundant, who gains power, who loses it, and why?
The central argument is precise. AI does not automate decisions. It automates prediction, a crucial input into decisions. Cheap prediction shifts the value of complementary inputs — human judgment, access to proprietary data, control of distribution — and that shift redistributes power. Some roles become more valuable. Others become redundant not because a machine replaced the person, but because the prediction the person was making is now nearly free.
The book introduces the concept of a transition from a world of systems to a world of points. In the current era, organizations are built around systems of coordinated rules that handle uncertainty by routing decisions to humans. As AI resolves that uncertainty, those systems break apart into points where a single actor, armed with reliable prediction, can bypass the old coordination structure entirely. Understanding this transition tells you where disruption will come from and where incumbents retain durable advantage.
Agrawal, Gans, and Goldfarb are economists, not futurists. The framework they offer is grounded in incentive theory, not speculation. Each chapter tests the model against real industries — health care, financial services, logistics, media — and surfaces the same structural pattern playing out in different forms. That consistency is the book's practical value: once you see the pattern, you can apply it to your own context without waiting for a case study to appear.
If you are responsible for strategy, product direction, or organizational design in a sector where AI is arriving, this book gives you a durable analytical lens rather than a list of hype-cycle predictions.
Establishes the book's core premise: AI makes prediction cheap, and cheap prediction changes the value of everything else. You learn to separate prediction from decision-making and understand why that distinction matters strategically.
Introduces the book's central framework: the shift from coordinated rule-based systems to discrete decision points. You see how organizations built around uncertainty management are structurally vulnerable to AI-enabled point solutions.
Maps how decision-making authority is currently distributed inside firms and across industries, showing how that distribution reflects the cost and scarcity of prediction before AI.
Explains the economic incentives that push firms to reorganize once AI lowers prediction costs, and why incumbents often resist restructuring even when it would benefit them competitively.
Examines how new entrants exploit cheap prediction to bypass incumbent coordination structures, using health care and financial services as concrete cases of the same underlying pattern.
Identifies the inputs that remain scarce and valuable after prediction is commoditized — proprietary data, regulated access, human judgment in high-stakes contexts — and how to build strategy around them.
Addresses what it means to be a human decision-maker when your prediction function has been absorbed by a machine, including how roles, accountability, and organizational design need to adapt.
Translates the book's framework into a practical lens for identifying where your organization stands in the systems-to-points transition and what actions are available before the shift consolidates.
No. The authors write for a business and strategy audience. Economic concepts are introduced through concrete examples rather than formal models, and no prior economics training is assumed.
Prediction Machines explained what AI is and why cheap prediction matters. Power and Prediction focuses on the second-order question: how cheap prediction redistributes power inside organizations and across markets. It builds on the earlier framework but is a standalone read.
The book was published in November 2022, before generative AI dominated the public conversation. Its framework is deliberately structural and economic rather than technology-specific, so the core arguments apply to current AI capabilities even if the examples predate them.
It is primarily a conceptual framework book. It gives you an analytical lens for reading AI-driven change rather than step-by-step implementation instructions. Readers who want tactical playbooks should treat this as the strategic foundation for those playbooks.
The book runs 172 pages. Most readers complete it in three to five hours of focused reading.
New
The Disruptive Economics of Artificial Intelligence and What It Means for Business Decisions
by Ajay Agrawal
New
New
A Practical Guide to Using AI Coding Tools and Productivity Assistants in Your Daily Workflow
by Tom Taulli
New