AI tools and systems making their way into the business sector will have a serious impact on high-wage, white-collar and tech jobs according to a report by US research group Brookings Institution.
The study assigns an AI exposure score that predicts the likelihood of AI affecting cities, regions, occupations or demographic groups. The algorithm behind the score uses natural language processing (NLP) to analyze job descriptions and patent database texts.
The score results show that better-educated people and groups like White and Asian men are most likely to encounter AI. Industries with more women are less likely to experience AI exposure. This goes against traditional theories that suggest low-paid, manual jobs are most vulnerable to AI disruption.
Like previous reports, this new study found high exposure among textiles, manufacturing and motor industries. However, it suggested that health, retail and education would be almost immune to AI despite previously being highly vulnerable.
The difference in results is due to the Brookings exposure scoring system being less subjective than previous studies. It is also unique in the way it creates a distinction between automation and AI which impact roles differently.
Earlier this year, Brookings reported that up to 45% of jobs could be impacted by AI in the US.
An unrelated paper, published earlier in 2019, argues that economists lack the tools to accurately predict how labor markets will respond to AI.