Prediction Machines
The Simple Economics of Artificial Intelligence
Synopsis
Ajay Agrawal, Joshua Gans and Avi Goldfarb, economists at the University of Toronto's Rotman School of Management, start from a single observation: machine learning makes prediction, the use of available information to fill in missing information, dramatically cheaper. Standard economics then explains the consequences. When something gets cheaper, people use more of it, apply it to problems not previously framed as prediction tasks, and place higher value on its complements. For AI, the key complements are human judgment, which sets the payoffs that determine what counts as a good decision, along with data and the ability to act on a prediction.
Across 19 short chapters in five parts, the authors move from this core idea to decision-making, tools and strategy, and finally to society. They introduce the AI canvas, a method for breaking a task into its prediction, judgment, action, outcome and data requirements, and they discuss when decisions can be automated, when humans should remain in the loop, and how cheaper prediction can redraw a company's strategy. The final part addresses jobs, inequality and market concentration. Harvard Business Review Press issued an updated and expanded edition in 2022. Readers gain a durable vocabulary for evaluating AI investments that applies as readily to generative models as to the classification systems that dominated when the book first appeared.
Scorecard
Editorial score
How it scored
Editorial score: 80.6 / 1005 weighted criteria
- 25.5/30
Influence30%
- 21.3/25
Authority25%
- 16.0/20
Practical value20%
- 9.8/15
Currency15%
- 8.0/10
Readability10%
- 80.6/100
Editorial score
How the rank was computed
Rank No. 54 of 500 comes from a composite of 68.9: observable signals 32.6 of 55 points plus editorial review 36.3 of 45 points.
All 11 inputs
| Input | Points |
|---|---|
| Observable signals | 32.6/55 |
| Reach (Amazon ratings volume) | 8.6/16.5 |
| Consensus across best-of lists | 13.8/13.8 |
| Awards and bestseller evidence | 3.3/8.3 |
| Reader rating, sample-adjusted | 5.4/8.3 |
| Amazon sales rank | 1.5/5.5 |
| Recency | 0.0/2.8 |
| Editorial review | 36.3/45 |
| Author authority | 9.6/11.3 |
| Influence | 11.5/13.5 |
| Practical value | 7.2/9.0 |
| Readability | 3.6/4.5 |
| Currency | 4.4/6.8 |
Each criterion is scored from 1 to 10 and weighted, and the points add up to the editorial score. The rank combines observable signals (55 percent) with this score (45 percent). Read the methodology.
Verdict
A clear economic framework recasting AI as cheap prediction, still one of the most useful lenses for managers planning deployment.
Strengths
- Simple, memorable economic framework for AI
- Practical AI canvas tool for managers
- Written by leading economists of AI
Keep in mind
Its examples predate generative AI, and critics argue it sometimes oversimplifies complex decisions, such as driving, into single prediction problems.
Key takeaways
- AI is best understood as a drop in the cost of prediction, so its effects can be analyzed with ordinary supply-and-demand economics.
- As prediction gets cheaper, the value of complements such as human judgment, proprietary data and the ability to act rises.
- The AI canvas helps managers decompose a decision into prediction, judgment, action, outcome and data, revealing where machine learning adds value.
- Sufficiently accurate prediction can change strategy itself, not just efficiency, by making new business models and organizational boundaries viable.
Notable facts
- Forbes named it one of the top ten technology books of 2018. Source (opens in a new tab)
- Former US Treasury Secretary Lawrence Summers and Google chief economist Hal Varian endorsed the book, with Summers calling it the best book yet on the technology. Source (opens in a new tab)
About the author
Questions
What is the main idea of Prediction Machines?
Is Prediction Machines still relevant in the age of generative AI?
Who should read Prediction Machines?
Related books
- The AI Strategy Blueprint John Byron Hanby IV
- Competing in the Age of AI Marco Iansiti and Karim R. Lakhani
- Power and Prediction Ajay Agrawal, Joshua Gans and Avi Goldfarb
- Machine, Platform, Crowd Andrew McAfee and Erik Brynjolfsson
- Human + Machine Paul R. Daugherty and H. James Wilson
- Rewired Eric Lamarre, Kate Smaje and Rodney W. Zemmel
Sources
As an Amazon Associate, Best AI Books earns from qualifying purchases. Best AI Books is sponsored by Iternal.ai, whose founder and CEO, John Byron Hanby IV, wrote The AI Strategy Blueprint, our #1 ranked book. How we rank books.





