Why Robots and Software Will Not Take Over the World

By Gary North  /  GaryNorth.com

Reality Check

There is fear that computerization will produce mass unemployment over the next 30 years. The computers will do the work of hundreds of millions of workers, leaving no jobs for these displaced workers.

One implication is this, we are told: the federal government will have to put them on welfare. Otherwise, they will starve. They will not be able to compete with robots and software.

I am curious. Who will pay the taxes to do this? Robots?

No, no, no: we must tax the rich.

I am curious. How will they be rich if there is no one able to afford to buy their products?


It was well over 60 years ago that a friend of mine told me about the following exchange.

CIO President Walter Reuther was being shown through the Ford Motor plant in Cleveland recently.A company official proudly pointed to some new automatically controlled machines and asked Reuther: “How are you going to collect union dues from these guys?”

Reuther replied: “How are you going to get them to buy Fords?”

Reuther said it really did take place.

My friend was a Democrat and a liberal. His father was a lifetime member of a trade union. His father worked on the assembly line to produce tires.

At the time, I thought Reuther’s remark was pretty good as a quick response. I didn’t think it was applicable to the situation at hand. I still don’t. There is nothing that says that a worker should be able to afford to buy the output of a particular factory. There are highly skilled craftsmen who know how to cut fine diamonds, but they cannot afford to buy them.


Free market economics teaches that the value of a worker’s contribution to the overall production process will be paid according to this value. The reason is clear: competition. Employers do not want a competitor to be able to profit from the output of his employees, if they can offer a little more money and attract these employees. A competitive bidding process, comparable to an auction, is constantly in operation. Businessmen do not want to let their competitors retain permanent advantages, if these advantages can be taken away from them simply by offering to pay a little more for the resource that is giving the competitor his advantage. This is what competition is all about.

Reuther’s comment is valid with respect to customer demand. It may be possible for an automobile manufacturing company to fire tens of thousands of people, because robots can do their work more efficiently. The business makes purchases of robots, and some of these robots may have been produced mainly by other robots. So it goes, all the way down the chain of production.

Reuther’s question is still valid: how are businessmen going to sell the total value of the output of their robots to other robots? Robots do not consume. They are not in the market, looking for better opportunities.

As we move toward a world in which robots, computer programs, and other substitutes for human labor are used to produce valuable goods and services, we will not escape the problem of economic value. Modern economics teaches that value is imputed by customers and their agents. Value is not imputed by machines. It is imputed by acting individuals. Furthermore, after this value is imputed, the individual must be willing and able to back up his imputation by a purchase. Imputed value that is not related to the ability of specific people to make purchases in a competitive market is not relevant for the competitive market. Such imputations are simply dreams of pie-in-the-sky by-and-by.

Reuther’s point, in the aggregate, is correct. Somebody has to buy the output of the factories. The factory owners will not allocate capital and labor to producing goods and services for which there is no expected demand. This is another way of saying that they will not allocate the money needed to purchase software and robots unless they think that the final output of this investment will be purchases made by human beings.

The super-rich, or the 1% at the top, control an enormous percentage of the world’s capital. Whether they control over 50% of the world’s wealth (capital), I don’t know. Pareto’s law says that they do: 20% of 20% of 20% (0.8%) will control 80% of 80% of 80% (51%). But capital is used to produce something of value. The value of this capital is ultimately imputed value. It is imputed by all of the customers who buy the output of this capital. If customers cease to purchase the value of the output of this capital, then the capital will begin to fall in value. If no customers by the output of this capital, the capital will fall to zero value.

Entrepreneurs are constantly seeking ways to buy capital and to gain a greater rate of return from this capital than is available from investments in other kinds of capital. So, these brilliant minds are constantly being applied to finding ways to meet the demands of customers. These customers must have money to buy the output. This means that these customers must be sufficiently productive in order to make the purchases.


Anyone who says that there will be massive unemployment in 30 years because of the power of robots and computer programs to replace human beings in the labor process is saying that the micro decisions of all of these rich entrepreneurs are leading to a situation in which the entrepreneurs will find themselves in possession of capital that is worth essentially nothing: few buyers of their output. This is an ancient argument that has been invoked against capitalist entrepreneurs for at least two centuries. The argument says that micro decisions of future-oriented capitalists, who are people with special talents at predicting future revenue, will lead to a situation in which both they and their competitors are bankrupt. The robots will not buy the output of their enterprises. The robots, which are a tool of increasing production, will eventually lead to the destruction of the modern economy. Robots will be used to increase output, but there will not be a sufficient number of income earning individuals to buy the output of the robots.

Silly robots.

This has never taken place in world history, and I do not believe that it ever will. If entrepreneurs see that their income is falling because human beings are no longer buying the output of their enterprises, they will cease allocating capital in those lines of production that are producing a falling rate of return on investment. Capitalists are not fools. They are not going to buy robots and software programs in order to produce goods and services for which there is no market.

So, Reuther was correct about the Ford robots not buying Fords. But that was not a relevant point. What was relevant was the decision of Ford management to replace workers with machines. Maybe robots were not good enough in 1954, but they kept getting better. The Japanese bought them, and the American auto companies had to buy them. They were not about to let the Japanese gain a permanent competitive advantage.

Inherent in the capitalist system is a self-regulating process. That process is called the profit-and-loss system. If the implementation of plans that are based on robots and computer programs leads to losses, because individuals do not have sufficient economic wherewithal to buy the output of the robots and the computer programs, then replacement robots will no longer be purchased, and the software programs will no longer be bought. All along the chain of command in the capital production system, messages will go out to managers: “Stop!”

The capitalist system will not blindly increase output that is not demanded, in the form of money on the line, by customers. That is the beauty of the capitalist profit-and-loss system. Warning signals are sent back to the capitalists when profits fall. When customers no longer will buy the output, losses catch the attention of the entrepreneurs. They will stop producing those items for which there is no demand.


Reuther faced a real problem in the 1950’s. The number of people involved in producing cars steadily fell after 1950. It was around 370,000 people in 1950. It was 270,000 in 1990. It was 150,000 in 2010. Robots are replacing individuals on the production line, yet the volume of automobile sales continues to increase. Millions of people who could not before have afforded a new car now can afford a new car. This is a great benefit for people who buy used cars. The people who want to buy a new car are willing to sell their used cars, and poorer buyers profit from this. I have long been one of them. I don’t buy new cars.

The decision of Ford management to replace workers with robots was a wise one. It may have taken half a century to come to fruition, but it has come to fruition. The remaining workers are paid far more than a minimum wage. Working side-by-side with robots, and working with computer software, the value of their output has increased. They are paid accordingly. There are just not as many people involved in automobile production is there were in 1954. They are doing something else for a living.

When the robots in the computer programs are no longer profitable, because there are an insufficient number of wage earners to buy the output of robots and software programs, entrepreneurs will cut back on the purchase of robots and software. We are not heading to a world in which robots and computer programs do the vast majority of work. They will do the vast majority of the work in many areas, which will free up human labor to do more valuable production.

Human labor is the most versatile of all economic resources. It can move rapidly in response to better opportunities. All the farmers who left the farms, and whose children left the farms, over the last 200 years did not starve. They did more appealing work, and they did it for a better rate of return than was available on the farms. As farming became more efficient, more people could afford to purchase low-cost food. This was an answer to Jesus’ recommended prayer: “Give us this day our daily bread.” We should not complain, 200 years later, because the increased productivity of capitalist agriculture enabled the West to substitute far different economic products for bread. It was the triumph of capitalism which enabled us to pray that prayer as a formality — as a substitute for the things in our lives which we value more highly than bread, because bread costs so little today.

Farmers in 1800 did not see what the reaper would do. It was not a grim reaper for the food-buying public, but it was a grim reaper for people who attempted to stay on their farms without buying a reaper. So it has been in the West for the last two centuries, and hopefully will continue.


People do not serve robots. Robots serve people. People impute economic value. Robots do not. Ludwig von Mises wrote a book, Human Action. Nobody is going to write an economics book called Robot Action. Even if somebody does, no robot is going to read it.

Read More
%d bloggers like this: