The rapid development of artificial intelligence has sparked an intense debate about its impact on the labor market and the economy. At the center are two opposing scenarios: on the one hand, AI could trigger a surge in productivity, create new industries, and free workers from monotonous tasks. On the other hand, there is concern that white-collar jobs in particular could be displaced faster than new employment is created, potentially triggering a deflationary spiral and a severe economic crisis.
A pessimistic scenario outlines a potential downward spiral beginning with the increasing capabilities of AI. Software can be developed faster and more cheaply, putting existing business models—especially in the software-as-a-service sector—under pressure. Lower labor costs reduce purchasing power, compress corporate margins, and encourage greater use of AI, further accelerating the trend. This feedback loop could intensify without a natural brake.
In a later phase, skilled workers across various industries would be displaced and pushed into lower-paid jobs. Since consumer spending accounts for around 70 percent of economic output, this would have significant consequences for the broader economy. At the same time, government budgets would come under pressure as declining tax revenues coincide with rising social expenditures. Financial markets could also be destabilized, for example through credit defaults in the technology sector or stress in the mortgage market if incomes decline.
More optimistic assessments challenge this outlook. They argue that technological innovation historically spreads according to an S-curve: after a phase of rapid growth, adoption slows due to integration costs, regulation, and diminishing returns. This slower pace gives businesses, labor markets, and governments time to adjust.
It is also argued that technological advances typically represent productivity shocks that lower costs, support growth, and increase real incomes. Accordingly, rising output must be accompanied by rising demand, since total economic activity is defined by consumption, investment, government spending, and net exports. An increase in production alongside collapsing demand would contradict fundamental economic principles.
Historical experience also argues against a permanent displacement of labor. Previous technological revolutions changed the composition of work but did not eliminate labor as a whole. Rather than working less, societies expanded their consumption. Even predictions such as those by John Maynard Keynes, who anticipated a dramatic reduction in working hours, proved inaccurate.
Another optimistic perspective emphasizes that economic activity is based on solving human problems, whose number is not limited. Efficiency gains often lead to increased demand, a phenomenon known as Jevons Paradox. If AI reduces the cost of certain tasks, it may generate new demand rather than destroy jobs. A key factor is which parts of a job are automated: if repetitive tasks are replaced, skilled workers may become more productive and valuable.
However, there is agreement that transition phases carry risks. Historical examples such as the so-called Engels Pause during the Industrial Revolution show that prolonged periods in which job losses outpace job creation can lead to significant social and political tensions. In the case of AI, outcomes will largely depend on the speed of adjustment.
In the long term, technological transformations have so far created more jobs than they destroyed, though often only after a transition period. If the shift driven by AI becomes unbalanced, economic and social disruptions could follow. It is therefore crucial to closely monitor developments in employment, wages, and consumption in order to identify potential negative feedback effects at an early stage.
Source: Zerohedge