Some suspect that AI could lead to mass social upheaval, akin to the Luddite riots in response to nineteenth-century automation. However, given that AI supports some jobs whilst undermining others, it’s unlikely that a united front will form. Institute for the Future of Work Fellow Carlos Felipe Balcazar argues this will further concentrate economic and political power in the hands of the top income earners. Redistributive policies would be needed to address this increased inequality that may emerge from automation, but will be made harder to achieve given higher levels of polarisation and the ability of elites to prevent opposition.

Sam Altman, CEO of OpenAI, stated earlier this year that artificial intelligence will “probably replace most of the jobs people do today,” and that entire job categories will be “totally, totally gone.” He was predicting a “job apocalypse.”  Mr Altman and other CEOs of AI companies have also claimed that AI will bring a big bounty for all, through lower prices for goods and the ability of societies to provide a universal income. These predictions have not yet materialized, and it’s unlikely they ever will.  AI’s capacity to replace labor is fundamentally constrained by the inherent limitations of algorithms exposed by the Turing Halting problem, which states that there are problems that no computer program can ever solve, no matter how clever they get. Society’s capacity for redistribution is also limited by CEOs’ mandate to increase stakeholders’ profits and the need for companies to remain economically competitive to do so; the redistribution of these gains doesn’t happen “just because.” A future with no work and universal well-being is therefore a far cry.   

AI, however, does create opportunities for companies to automate tasks that use a computer, reducing the need for those workers performing these tasks. The threat of increased unemployment as a result of AI adoption therefore looms large, and, at the time of writing, it seems to materialize only for early-career workers. The unemployment rate for workers aged 22-25 has declined markedly in comparison to November 2022, prior to the advent of commercial AI technologies with ChatGPT. For other age groups, the impact of AI has been more modest.

SUGGESTED VIEWING The fight for the AI economy With Aaron Bastani, Zoltan Istvan

Increased unemployment that can be attributed to AI has evoked the imagery of the violent riots by the Luddites in early nineteenth-century England, when automated looms and spinning machines took over the tasks performed by thousands of skilled weavers and knitters, plummeting their wages and their ability to remain employed or find new work in the industry. The prospect of unemployment and lower wages resulted in massive solidarity and mobilization from workers, striking, protesting, and even destroying the machines threatening their livelihoods. Factory owners reacted by shooting protesters.  The situation escalated to necessitate both legal and military suppression by the British government. The Luddite riots created a five-year period of social turmoil between 1811 and 1816.  Today, similar prospects for the health of the labor market fuel concerns for a wave of social discontent leading to political instability, where the distribution of the gains from AI could “take place by the guillotine.”