pull down to refresh

I'm going through the ILO's workshop on AI/job friction and there was a pretty interesting presentation from Laura Nursky that compares a set of studies about the expected exposure of jobs to AI disruption (with overall tech as a comparison) over time, focusing on job classification, sector, age groups, education and so on. You can see the entirety of the first half day of the workshop here, with the part I am highlighting starting around 02:09:00.

By job classification

Earlier studies indicate that many occupations will be exposed to AI in the future. Initially, there was little differentiation among specific groups of occupations, but a consensus emerged around 2021 suggesting that high-skilled professions such as managers and clerical support workers would primarily be impacted by AI. More recent studies split their scores between augmentation and automation, revealing that high augmentation potential is found in high-skilled occupations while most automation potential lies in lower-skilled groups.
Focusing specifically on generative AI (2nd column), there is a strong agreement that high-skilled professions are significantly impacted, particularly clerical support workers due to tasks like writing, drafting emails, and summarizing documents. This focus on AI often overshadows other technologies affecting the labor market, such as robotization, software, and automated guided vehicles; wider technological changes show that almost all occupations are being impacted in some way.

By sector

Sectoral exposure analysis shows high exposure in IT sectors like coding assistance and banking, as well as professional and scientific activities such as consultancies. Wider technological changes primarily affect agriculture, manufacturing, and mining, with less exposure in public sectors.

By education

Education level correlates positively with exposure to AI and Gen.AI but shows an opposite trend for wider technological changes.

By age

After a flat trend, recent age analysis suggests younger workers are slightly more exposed to Gen.AI than older workers, though occupational classification doesn't always distinguish between junior and senior roles within occupations.