AI to truly achieve productivity improvement will inevitably require large-scale unemployment to occur first.

Recently, Federal Reserve Board member Waller publicly revealed that many CEOs of large American companies have privately expressed that artificial intelligence will lead to mass layoffs.

Many people dismiss this statement, because it is clear that the current level of AI development and costs have not yet reached the point of replacing large numbers of human workers.

However, this complacent attitude is fundamentally wrong; the mainstream narrative about AI and employment today is as follows: AI technology is not yet mature, and its impact on productivity is minimal; only when AI becomes sufficiently powerful in the future will it start replacing human jobs, potentially triggering large-scale unemployment. In other words, most people place “large-scale unemployment” at the end of the story—it is seen as a consequence of AI maturing.

But if we carefully examine economic history and current macro data, we find an entirely opposite, yet more logical conclusion: large-scale unemployment is not the result of AI productivity gains, but rather its prerequisite.

Companies must first undergo a profound organizational restructuring—eliminating old roles, breaking down existing processes, and redesigning production methods around AI—before productivity can truly take off. Unemployment waves will not appear after AI “matures”; they will precisely occur during the transition period when AI shifts from a “useful tool” to a “core production engine.”