OpenAI's economics team: 'We don't have the nouns yet' for the jobs AI will create

paulnovosad · x · 2026-09-23

In a new Substack post, OpenAI's Economic Research team argues the AI-and-jobs debate over-focuses on existing occupations while systematically underestimating future ones, because we lack the vocabulary for jobs built on technologies and business models that don't exist yet. Examples: explaining 'optimizing search ads for a cloud-hosted video app' to someone in 1976 was impossible; research by Autor et al. finds most 2018 US employment was in job specialties absent from the 1940 Census; cybersecurity is a concrete example of an unforeseen job category. The post urges balancing displacement fears with hard-to-predict new opportunities.

Original post →

More from AGI Musings

AGI Musings channel →