AI Job Apocalypse: A Myth

The narrative of an impending "AI job apocalypse" is not only unconvincing but also a tired rerun of an old economic fallacy. This panic, often rebranded as the "permanent underclass" fear, is fundamentally the "lump-of-labor" fallacy in modern dress.

This fallacy posits a fixed amount of work in the economy, framing job competition as a zero-sum game. If AI takes on more tasks, humans supposedly have less to do. However, this premise ignores fundamental aspects of human behavior, markets, and economics. Our wants and needs are not static; history shows we adapt and create new productive endeavors.

While AI will undoubtedly reshape roles and eliminate certain tasks, the idea of economy-wide, permanent unemployment is poor economics and bad history. Productivity gains historically increase demand for labor by making it more valuable.

The "Checkmate, Humans?" Fallacy

The argument that AI performing cognitive tasks will render humans obsolete overlooks a crucial economic principle: when the cost of a powerful input falls, the economy doesn't stagnate. Instead, costs decrease, quality and speed improve, and new products and demands emerge.

This echoes the Jevons Paradox, where increased efficiency leads to greater overall usage and new applications. Just as cheap energy from fossil fuels spurred innovation beyond replacing manual labor, AI handling cognitive loads frees humans for more ambitious frontiers.

Technological transformations historically enlarge the economic pie. Successor industries have consistently been larger than their predecessors, making the overall economy bigger, more diverse, and complex. The doomer narrative freezes this progress at the point of task substitution, ignoring the subsequent expansion of work.

Lessons from History: Luddite Fears Put to Rest

History offers compelling counterexamples to the job apocalypse fear. In the early 20th century, agricultural mechanization moved roughly a third of the U.S. workforce out of farming. Instead of mass unemployment, farm output tripled, supporting population growth and fueling the rise of entirely new industries.

Electrification similarly revolutionized factory work and consumer goods, leading to decades of doubled labor productivity growth. This surge in productivity fueled demand for more manufacturing, sales, lending, and commercial activity, alongside second-order effects that pulled people into higher-value work.

Even specific software like VisiCalc and Excel, which automated bookkeeping, did not doom bookkeepers. Instead, these tools led to an explosion in financial analysis roles, transforming millions of jobs rather than eliminating them. This pattern, seen across technological shifts, suggests AI will similarly expand the frontier of useful work.

Augmentation Over Substitution

For every job at risk of AI substitution, numerous others stand to benefit from AI augmentation. Management focus increasingly leans towards augmentation, with AI-as-augmentation out-mentioning AI-as-substitution on earnings calls by a significant margin.

Software engineers, for instance, are a prime example of an AI-augmented role. AI acts as a force multiplier for coding, leading to skyrocketing development activity and, potentially, increased demand for engineers and product managers. This growth in both roles contradicts the simple substitution model.

The majority of new jobs created since 1940 did not exist in 1940. The doomer failure is largely one of imagination, fixated on tasks automated away while ignoring the emergent demand for entirely new roles and industries we cannot yet conceive.

While it's early days, current data does not support the widespread job destruction narrative. Research from institutions like the NBER, the Federal Reserve Bank of Atlanta, and Yale's Budget Lab suggests AI adoption has thus far led to modest employment impacts, with a reshaping of tasks rather than mass unemployment.

The weight of historical precedent and emerging data points not to an AI job apocalypse, but to an expansion of economic activity and the creation of new, unforeseen opportunities. The fear of technological unemployment has been a recurring theme throughout the history of technological unemployment, yet humanity has consistently adapted and innovated.