Pages
347
Published
2022
The Simple Economics of Artificial Intelligence
Understand AI as an economics problem — and make sharper decisions about where, when, and how to deploy it in your business.
Prediction Machines reframes artificial intelligence as a drop in the cost of prediction. That single insight, developed by three economists who study AI professionally, unlocks a clear framework for deciding when AI creates genuine value, when it doesn't, and what changes across your entire workflow when prediction gets cheap. This updated and expanded edition extends the original analysis to cover the latest developments in machine learning and their organizational consequences.
Artificial intelligence is not magic, and it is not a strategy. It is a technology that dramatically lowers the cost of one specific thing: prediction. Once you see it that way, an enormous amount of confusion about AI in business dissolves.
Prediction Machines was written by three economists who study AI and its effects on firms, markets, and decision-making. Their central argument is precise: when prediction becomes cheap, complementary inputs — judgment, data, action — become more valuable. That shift reshapes job roles, competitive advantage, and investment priorities in ways that a purely technical lens misses entirely.
This updated and expanded edition builds on the original framework with new material covering the rapid advances in machine learning since the book first appeared. The core economic logic holds: the right question is never "should we use AI?" but rather "what decisions does cheap prediction change, and for whom?"
The book works through that question systematically. It distinguishes prediction from intelligence, separates the value of data from the value of prediction, and shows how risk tolerance, incentive structures, and organizational design all need to adapt when a key input gets cheaper. It gives you a vocabulary and a set of tools that transfer across industries, functions, and specific AI applications.
If you are an executive, analyst, product manager, or technologist trying to build a durable view of where AI fits in your organization, this is the analytical foundation that makes the rest of the literature make sense.
Establishes the core thesis: AI is best understood as a technology that lowers the cost of prediction. This framing sets up the entire book's analytical approach.
Defines prediction precisely — filling in missing information about an unknown state of the world — and separates it from the broader, looser claims made about AI's capabilities.
Examines what happens across an economy when a key input gets dramatically cheaper, drawing on historical analogies and economic theory to set expectations for AI's impact.
Distinguishes between data as a raw input and data as a source of competitive advantage, and explains the conditions under which proprietary data actually matters.
Introduces judgment as the complement to prediction — the human input that specifies objectives and evaluates outcomes — and shows why cheap prediction makes judgment more, not less, important.
Maps how AI shifts the boundary between human and machine tasks, focusing not on job titles but on the specific decisions and micro-tasks that change hands.
Presents a framework for auditing any business decision to identify whether cheaper prediction changes the right answer, the right process, or the right person responsible.
Applies the framework to competitive strategy: when AI creates durable advantage, when it commoditizes, and how to position a firm for either outcome.
Addresses the downside scenarios — model failure, misaligned objectives, and data vulnerabilities — and shows how organizational design can reduce exposure.
Extends the analysis to labor markets, policy, and societal adaptation, covering the updated material on recent AI advances and their macroeconomic consequences.
No. The book is written by economists, not engineers, and requires no programming or statistics knowledge. The argument is conceptual and strategic throughout.
No. It deliberately avoids covering specific products or vendors. The framework is designed to apply regardless of which tools or models are current, which is why it has held up since the first edition.
The updated edition adds new material addressing advances in deep learning, large language models, and their organizational implications. The core economic framework is unchanged but extended.
It is both. The framework is grounded in economic theory, but each chapter applies it to concrete business decisions, organizational structures, and strategic tradeoffs.
Readers looking for hands-on implementation guidance, model-building tutorials, or tooling recommendations will need a different book. This one focuses on strategic and economic reasoning, not technical execution.
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