Recently, the topic of AI taking away jobs has been buzzing around society, but do you realize it might actually be the opposite? A lengthy article published by a16z partner companies fundamentally denies the theory of "the end of employment due to AI."

What’s interesting is the perspective that captures the underlying fallacy of this debate. People fueling AI panic generally assume that "the total amount of work to be done worldwide is fixed." But history shows otherwise. With the mechanization of agriculture, the number of agricultural workers decreased from one-third of the workforce in the early 20th century to 2% in 2017, yet unemployment did not persist. Instead, labor shifted to factories, stores, offices, hospitals, and the software industry, creating an entirely new economic system.

The same pattern applies to electrification. It wasn’t just a switch of energy sources; the structure of factories was rebuilt, leading to the creation of consumer goods and industrial products that had previously been unimaginable. In the early 20th century, only 5% of American factories used electricity, but by 1930, that figure rose to 80%. Over the following decades, labor productivity doubled, but employment not only did not decrease—it increased.

When VisiCalc and Excel were introduced, people said bookkeeping jobs would disappear. But what actually happened? The number of bookkeepers decreased by 1 million, but the entire financial analysis (FP&A) industry emerged, with 1.5 million more financial analysts. In other words, jobs weren’t replaced; rather, the quality of work became more advanced, and higher value-added jobs were created.

What about the current AI situation? According to Goldman Sachs estimates, the effect of AI replacing jobs is far smaller than its effect of enhancing functions. Also, at earnings calls, executives say they are using AI about eight times more for "enhancement of existing functions" than for "replacement of existing functions."

Demand for software engineers is rapidly increasing. The number of Git pushes is exploding, and the creation of new applications and businesses is also on the rise. From early 2025, jobs in software development are steadily increasing in both number and proportion within the overall employment market. The number of product manager openings, after a temporary dip due to interest rate fluctuations, is now at its highest level since 2022.

Of course, not everything is smooth. Research from Stanford University and the Dallas Federal Reserve shows that it’s becoming harder to find entry-level jobs closely related to AI. But at the same time, it’s also true that the number of entry-level roles where AI provides assistance is increasing.

Looking at academic research, current data does not support the apocalyptic view. An NBER paper points out that while AI adoption has not significantly changed overall employment numbers, it is beginning to reshape the distribution of tasks and occupations. Routine administrative work is more easily replaced by AI, but analysis, technical, and managerial tasks are more likely to be complemented and enhanced by AI.

A working paper from the Atlanta Federal Reserve estimates that over 90% of companies reported no impact from AI in the past three years. Census Bureau data shows that only about 5% of companies using AI reported effects on their employment levels, with roughly equal numbers seeing increases and decreases.

Most importantly, there is still no statistically significant correlation between AI and unemployment rates or employment growth rates. Overall, the relationship is neutral but not static. Some jobs disappear, some emerge, some decrease in value, and others increase in value.

The emergence of new businesses is exploding and shows a high correlation with AI applications. The share of new apps listed on app stores has increased by 60% year-over-year. Data sets related to robotics, for example, have jumped from 10th to 1st place in just two years.

Seeing it this way, the theory of AI ending employment is essentially a lack of imagination. Most of the jobs created since 1940 did not exist in 1940. The rise of mid-sized tech services industries focused on cloud migration and other innovations was unimaginable at the time.

Ultimately, human needs and ambitions are endless. When food prices fall, spending shifts to housing, healthcare, education, travel, and entertainment. The labor market follows the same logic. New jobs are constantly emerging, conquering old frontiers, and then new frontiers appear again.

The true purpose of productivity improvement is to eliminate harsh labor. The claim that AI will end employment only holds if, the moment AI becomes cheap, human needs and ideas suddenly vanish. That’s clearly flawed.

Macroscopically, the future is not an era of unemployment. Cheaper AI, larger markets, new companies, new industries, and more advanced human work await. There are no fixed values for the amount of work or cognitive capacity. AI is not the end of jobs but the dawn of a richer era of intelligence.