Not feeling like winners: The higher educated and the AI revolution

Technological redundancy was so far seen as mainly a concern for lower to medium educated workers, but recent advances in machine learning and algorithmic decision-making technology – i.e., “artificial intelligence” (AI) – have raised concerns that the higher educated could now be affected as well. Adding to an emerging literature on the effects of AI exposure, I show that higher educated workers indeed feel that they are at risk of technological redundancy, and that these perceptions are related to objective exposure to AI technology. Specifically, I combine comparative survey data on workers’ subjective technological vulnerability with indicators of objective industry-level AI exposure and find a U-shaped pattern among higher educated workers: There is high perceived vulnerability in some “traditional” automation-exposed sectors (e.g., manufacturing) but also in a group of strongly AI-exposed white-collar industries (finance, real estate, IT), with a remaining pocket of low vulnerability in strongly interpersonal sectors (e.g., education, healthcare). These findings have broader societal and political implications, namely that further advancement in AI technology may contribute to a de-polarization of social risk and thus create the basis for new cross-class coalitions and social solidarity – but it may also reinforce current authoritarian tendencies.