Average is Over by Tyler Cowen:

ChatGPT:

"Average Is Over, published in 2013, is one of the more provocative books about the future of work, economic inequality, and technological change.

Its central argument is that the ability to work effectively with intelligent machines will become one of the most important determinants of economic success.

Cowen predicts that technological progress will not benefit everyone equally. Instead, it will increasingly divide society into two groups: people who can use intelligent machines to become extraordinarily productive, and people whose skills become less valuable as machines take over more cognitive tasks.

What makes the book particularly interesting is that Cowen does not simply predict that computers will replace human workers. His argument is considerably more subtle.

The future belongs to people who can combine human judgment, creativity, and expertise with the extraordinary computational capabilities of machines.

Let's examine his argument in detail.

1. The end of the average worker

For much of the twentieth century, developed economies provided relatively comfortable lives for millions of workers with ordinary skills.

A person could obtain a reasonable education, find stable employment, perform moderately demanding tasks, and earn a middle-class income.

Cowen argues that this economic arrangement is becoming increasingly difficult to sustain.

Why?

Because computers can perform a growing number of tasks that previously required educated human beings.

Meanwhile, the value of people who know how to direct, interpret, and improve computational systems may increase dramatically.

Cowen's prediction is therefore not simply that machines will replace people.

It is that machines will change which human abilities are economically valuable.

2. Human–machine collaboration: the central argument

This is probably the most interesting part of the book.

The most productive employee may not be the person with the greatest individual intellectual ability, but the person who can most effectively organize and exploit machine intelligence.

3. Why intelligence alone will no longer be enough

Being intelligent and being effective at using intelligence are not necessarily the same thing.

Cowen's argument suggests that the second ability will become increasingly important.

4. The polarization of the labor market

Cowen predicts that intelligent machines will contribute to a growing division between economic winners and losers.

At the top will be people whose abilities complement technology. They may become extraordinarily productive and receive substantial economic rewards.

In the middle, many traditional occupations will come under pressure. Routine analytical, administrative, and technical tasks will increasingly be automated.

At the lower end, some service occupations will remain difficult to automate, but their wages may remain relatively modest.

Cowen imagines a society in which perhaps 10–15% of the population enjoys exceptional prosperity, while many others experience stagnant incomes.

However, his argument contains an interesting paradox.

Even people with relatively low incomes may enjoy access to extraordinary technological capabilities, inexpensive entertainment, information, and education.

In other words, economic inequality could increase even while technology makes certain aspects of everyday life substantially better.

5. Education will change fundamentally

Cowen also questions the traditional educational system.

Historically, obtaining knowledge required access to teachers, universities, libraries, and expensive educational resources.

Intelligent machines dramatically reduce some of these barriers.

A motivated individual can increasingly learn mathematics, economics, programming, statistics, or engineering with computational assistance.

Consequently, the ability to motivate oneself and organize one's learning becomes particularly valuable.

Previously, a person might have needed to attend specialized courses, read numerous technical papers independently, and find experienced colleagues willing to explain complicated mathematical concepts.

6. A particularly important implication: the speed of learning

There is another consequence of Cowen's thesis that deserves attention.

Intelligent machines do not merely increase the productivity of people who already possess considerable expertise.

They can also accelerate the acquisition of expertise itself.

Imagine two equally motivated analysts.

One learns primarily through textbooks, professional experience, and occasional discussions with colleagues.

The other uses the same resources but additionally spends an hour every day discussing difficult economic and mathematical questions with an AI system.

The second analyst might investigate substantially more questions, encounter more modeling approaches, and receive much more immediate feedback.

Over several years, these differences could accumulate.

However, the advantage depends on how the technology is used.

Someone who merely accepts machine-generated answers may develop superficial knowledge.

Someone who repeatedly questions the machine, verifies its conclusions, and implements the underlying ideas may develop much deeper expertise.

The crucial distinction is between outsourcing your thinking and using machines to improve your thinking.

7. What Cowen may have underestimated

We should also examine the limitations of his predictions.

The book appeared in 2013, long before the widespread adoption of modern generative AI.

What happens when machines become so capable that human intervention actually reduces performance?

In such situations, the most valuable human skill might no longer be helping the machine solve a problem.

8. The central lesson

I would summarize Average Is Over in one paragraph:

Tyler Cowen argues that intelligent machines are fundamentally transforming the economic value of human abilities. In the emerging economy, success will increasingly depend on whether individuals can complement computational intelligence rather than compete against it. People who combine expertise, discipline, adaptability, and technological competence may become extraordinarily productive, while those whose abilities can be replaced by machines risk economic marginalization. The resulting society may become significantly more unequal, even as technological progress makes knowledge, education, and many consumer services more accessible.

And I would add one observation particularly relevant to 2026.

The ability to collaborate with intelligent machines may become important not only because it makes you more productive today, but because it enables you to become more knowledgeable, more capable, and more productive tomorrow.

That creates the possibility of a cumulative advantage: better use of AI accelerates learning, which improves judgment, which in turn enables even better use of AI.

And it raises a question worth examining in considerable depth: Could someone who systematically uses AI to deepen their professional knowledge over the next five years acquire expertise that would previously have required fifteen years to develop?

The possibility of substantially accelerating intellectual development is real—and it is one of the most consequential implications of the ideas in Average Is Over."