The automation trap. The future of the entire economy is at stake

Prof. Aleksandra Przegalińska, a researcher of new technologies associated with the Kozminski University, in a recent post on LinkedIn, pointed out that everyone will lose in the long term due to mass automation of work. Even companies that enjoy savings today may lose markets for their products and services in the future because there will be no one to buy them.

An economic model based on mass consumption requires a large, well-earning middle class. As French economist Gilles Saint-Paul points out, this system will collapse if the profits from automation go solely to technology owners (who cannot consume everything) and workers lose income.

Read also: You won't do anything. Google shared its vision of the world

In the work “Artificial Intelligence, the Collapse of Consumer Society, and Oligarchy”, Saint-Paul analyzes a scenario in which oligarchs and corporations fall into the trap of self-optimization – mass production of goods and services using advanced technologies is then no longer profitable. As a result, some companies lose their own market. What looks like savings in the short term may end up undermining the foundations of the entire economic model in the long run.

The basis of economics is not productivity

The problem is particularly visible in the United States, where household consumption accounts for about two-thirds of GDP. The local economy is based on a simple assumption: people work, earn and spend money.

However, if more and more of the revenue goes to technology and capital owners and less and less to workers, the system begins to become unbalanced.

According to data cited by Aleksandra Przegalińska today the richest 10 percent households account for about half of all U.S. consumption. At the same time, the number of people carrying credit card debt and having problems with repayment is growing.

Productivity in itself does not guarantee prosperity. It also matters who benefits from its growth. If the profits from automation are concentrated in the hands of a small group of technology owners and the incomes of most workers decline, there is a risk of weakening the demand on which companies themselves also depend.

Automation needs a secure framework

Interestingly, similar questions are increasingly being asked not only by critics of artificial intelligence, but also by people who have been involved in its development for years.

To prevent economic and social shocks, it is necessary to build new safety nets. Experts emphasize that public administration and companies must jointly engage in financing retraining programs (upskilling) of employees displaced by AI. Moreover, such social security systems are beneficial regardless of the scale of changes.

Read also: The financial industry is being transformed by AI. The head of HSBC leaves no doubts

As subsequent AI models take on more and more complex tasks, the topic of introducing an unconditional basic income (UBI) and an automation tax is also recurring. If artificial intelligence really starts generating huge profits with less and less human labor, redistributing some of these benefits may become necessary. However, the condition is that it corresponds to the actual costs of living and the scale of the potential transformation.

Not all automation looks the same

However, this does not mean that artificial intelligence must lead to such a scenario.

We have two paths to choose from when implementing AI. The first is replacing people with artificial intelligence, the second is strengthening their competences.

Daron Acemoglu, David Autor and Simon Johnson, in the report “Building Pro-Worker Artificial Intelligence” published this year, focus on technologies that increase the value of human skills, help perform more advanced tasks and create new areas of specialization, instead of eliminating humans from the work process.

Read also: Can AI be a reason for dismissal? The expert explains

Optimization versus simple human greed

The authors point out that the potential of AI as a collaborator is as great as its automation potential. The problem is that companies are more likely to focus on reducing costs than on building new opportunities for employees.

This is confirmed by the first experiences of enterprises. Contrary to media narratives, many AI implementations do not bring spectacular increases in productivity. Some companies also discover that completely replacing people with algorithms leads to a decline in the quality of work because the systems lack experience, context and responsibility for the decisions made.

Experts call this phenomenon “workslop” inflation. Markets and organizations are flooded with an increasing amount of generated content and work of increasingly low importance and low value.

Worse still, over-reliance on algorithms in the name of convenience leads to atrophy of human skills (deskilling). Employees who thoughtlessly delegate tasks to machines gradually lose their expert competences and agency. In the long run, this destroys the innovation of companies, which become completely dependent on tools from external suppliers.

The “human plus AI” model, in which humans retain control over the process and technology serves as support, brings much better results. However, this means an additional cost.

Read also: Boss tells you to use AI? More and more employees have problems with this

A legal minefield

Replacing humans with artificial intelligence raises a fundamental, often overlooked legal issue: who is responsible for critical errors? Experts warn that creators of powerful AI models usually wash their hands of them, trying to shift all responsibility to the companies using their solutions.

Read also: Will AI be held responsible for complicity in the massacre? An unprecedented court case

In a system in which autonomous machines, not people, make key business, HR or medical decisions, the lack of clear regulations regarding liability and insurance is a ticking financial bomb for enterprises.

It's not just about jobs

In the report “Building a Human Resilience Infrastructure for the AI ​​Age”, published in 2026, Elon University experts indicate that the effects of the AI ​​revolution may go far beyond the labor market. One of the most frequently mentioned threats is the already mentioned loss of agency and the gradual delegation of an increasing number of decisions to algorithmic systems.

It's not just jobs that are at stake. The authors of the report also warn against a phenomenon known as the “work quake” – a shock in the labor market that may simultaneously mean the displacement of some professions and the creation of completely new ones, mass layoffs and a crisis of identity related to work.

Please note that work also has a social function. It determines status, gives a sense of meaning, allows you to build relationships and plan the future. Therefore, the question about automation is not just a question about productivity. This is a question about what an economy should look like in which an increasing part of the tasks are performed by machines.