Over the short term, AI may create more jobs than it destroys. But bosses will use it to de-skill work and suppress wages unless workers make it a terrain of class struggle and fight to put it under public control.
The Economist recently ran a cover entitled “Prepare for an AI jobs apocalypse,” citing widespread fear among the American public of artificial-intelligence-induced job losses. Such fearmongering is far from new. Information technology breakthroughs have led to many such announcements over the last few decades.
In the 1990s, we had the computer scare, prompted by best-selling authors such as Jeremy Rifkin and Viviane Forrester. More recently, in 2013, we had the robot menace, when an Oxford study estimated that 47 percent of jobs in the United States were at risk of computerization. At both times, such prophecies coincided with mass unemployment, not caused by technology but by economic recessions driven by financial instability and austerity policies. Yet such stories helped to naturalize and depoliticize unemployment.
Old Stories About New Machines
As The Economist points out, the scenario today is different. The predicted mass destruction of jobs coincides, in fact, with historically high employment. Given the ongoing investment boom not only in data center infrastructure but also in upstream utilities such as water and electricity supply, the AI frenzy is now arguably creating more jobs than it is destroying. Another peculiarity is that the current narrative of the job-apocalypse story is driven by AI giants as a marketing strategy to sell their large language models (LLMs) to financial investors and, therefore, feed the ongoing financial euphoria. Nonetheless, the Left should not dismiss the potential systemic effects of AI in the labor process, notably its role in de-skilling and devaluing labor. AI needs to be historically understood and situated as a breakthrough born from concentration and centralization — and of capital’s struggle with labor.