Every time a new automation technology emerges, the same question resurfaces: "Will it steal our jobs?". It happened with the mechanical loom in the 18th century (the Luddites destroyed them), with the assembly line in the 20th century, with computers in the 1980s, with the internet in the 2000s. Now it's artificial intelligence's turn. But history teaches us that the answer is never a simple yes or no.
When mechanical looms replaced manual weavers in 18th-century England, weavers lost their jobs — this is undeniable. But within a few decades, the textile industry employed far more people than before, in different roles: factory workers, engineers, merchants, transporters. The problem was not that work disappeared, but that benefits and costs were distributed in a brutally unequal way. Factory owners grew rich; manual weavers died in poverty. The transition created aggregate wealth but individual devastation.
The same pattern has repeated with every wave of automation. Economist Joseph Schumpeter called it "creative destruction": capitalism progresses by destroying old industries and creating new ones. But Schumpeter was more honest than many of his followers: he acknowledged that the destruction was real and painful, not merely a painless transition phase.
What makes AI-based automation different from previous ones? Three elements. First: speed. Previous industrial revolutions unfolded over decades; AI is transforming entire sectors in years. Second: breadth. Previous automation mainly affected manual, repetitive work; AI can automate cognitive, creative and relational tasks too. Third: concentration. Benefits of industrial automation were distributed among many companies; AI benefits tend to concentrate in a few global corporations that control models, data and infrastructure.
A 2023 Goldman Sachs study estimates that generative AI could "expose" around 300 million jobs worldwide to automation. But "expose" does not mean "eliminate": it means that a significant portion of tasks that make up those jobs could be automated. The difference between "task automation" and "job automation" is crucial. Few jobs are fully automatable; many jobs will see some tasks automated and others transformed.
The right question is not "will AI steal our jobs?" but "who will benefit from automation, who will pay the price, and what can we do to make the transition fairer than past ones?".