Big 4 Exploration Series: The Rise of a Useless Class

Dozens of movies, perhaps hundreds, have looked into the future through an artificial intelligence (AI) lens, such as the 2001 classic Space Odyssey, speculating at the enormity of a self-aware machine. Arguably none are more mind-bending than the Wachowskis directed The Matrix, which reshaped the way movies are filmed. 

Beyond the realm of Hollywood, Social historian Yuval Noah Harari speaks of another AI-derived danger, one which already has a ubiquitous presence in our society: unconscious AI. “What will happen…when computers push humans out of the job market and create a massive new ‘useless class,’” he inquires.

The accounting firm, EY, briefly converses on the idea of a “useless class” in one of its workforce series assessments: In a digital world, how can being more human be key to unlocking more growth?

Through analysis one arrives at two pressing questions:

  1. What exactly is a “useless class”?

  2. Can a “useless class” actually happen on a large scale in the developed world?

On the face, the phrase “useless class” has a visceral impact, but its meaning in the context of Harari’s usage seems to describe a severe, perpetual, and large-scale structural unemployment epidemic.

Structural unemployment refers to a scenario where the skills of unemployed workers don’t “line up” with the desired skills of the market, e.g. unemployed oil drillers lack the skills necessary to install solar panels. Structural unemployment is often the most-feared type of unemployment, as other types such as cyclical unemployment (which refers to unemployment resulting from a downturn in the economy, or frictional (natural economic job-hopping) generally are self-correcting or temporary.

In basic terms, Harari’s phrase “useless class” is a hypothetical environment where AI possesses the specific skills of most laborers, and the new skills required of those workers take a generation or multiple generations to learn. This would result in an indefinite period of unemployment for potentially a majority or at least a plurality of society, creating a lasting economic depression and severe income inequality. 

So is such an event plausible? Or even possible?

If we rely on the past half-century as historical evidence, the link between technological advancement and higher rates of long-term unemployment is often ambiguous and in some cases even positively correlated.

The Beveridge Curve, which plots the unemployment rate versus the job openings rate, paints a slightly different picture. It suggests that over the last seven years since the economic recession there has been a steady structural deficiency in America’s workforce. However, that deficiency has not created a post-recession higher rate of unemployment, but a higher rate of job openings. In simple terms, and in an apparent contradiction to Harari’s claim, the post-recession economy has created a field of jobs without enough American workers with the necessary skills or desire to fill them. BeveridgeCurve02


Definitively, the technological transition in both accounting and other service industries is indeed startling. After all, for decades it was manufacturing and other goods producing sectors that shed jobs and worried about the repercussions of an automated world. However, over the last seven years at least five of the top ten job-losing subsectors in the U.S. have been in services. In addition, according to a 2013 study conducted by Oxford academics, Carl Frey and Michael Osborne, accountants are among the highest paying jobs to be most susceptible to automation in the coming years.  

However, Frey and Osborne’s assertions work against the grain in two different manners. One, as discussed above, they don’t fall in line with historical precedent. At a time when technology has never been more advanced, America’s domestic unemployment rate is actually below the rate of five percent at which most economists agree is full employment.

Finally, Fray and Osborne’s claims dissent with government agencies, including those of the Bureau of Labor Statistics (BLS). The BLS projects that the accounting industry will add 142,400 jobs between 2014 and 2024, a rate faster than the cross-industry national average.

While several studies do indicate that a material percentage of current jobs will be obsolete in the coming years and decades, history reveals workers have always been resilient in their ability to learn new skills in a constantly changing technological environment. For now it appears that Harari’s claim won’t be realized in the near-future. Beyond that? Only time will tell.