In The News

AI Hiring Is Hard, but That Doesn’t Mean There’s a Skills Gap

Morning Brew

April 26, 2021

Diana Gehlhaus and Ilya Rahkovsky's CSET issue brief "U.S. AI Workforce" shows a lack of evidence to suggest there is an AI talent gap in the United States.

See original article

Survey after survey shows that hiring AI talent is the biggest barrier to enterprise AI adoption. The Biden admin, the DoD, and other Washingtonians want to fix it.

Only problem? We don’t know how big the gap is…or whether one truly exists at all. 

A new report from Georgetown University’s Center for Security and Emerging Technology (CSET) says there is little evidence to suggest there is a massive, urgent AI skills gap. Digging deeper…

  • If there is a shortage, it’s likely just for a small subset of highly specialized AI roles—like research scientists.
  • Talent pipelines are “working to meet increased demand” for roles like software developers and data scientists, meaning supply is likely to catch up to demand soon. 
  • There’s no evidence of a talent gap for less technical roles, like project management or UX design. 

Bottom line: The findings echo something David Autor, an economics professor at MIT, told us: “Whenever there is a hot new field, there is a scramble for talent and expertise—however, most AI tools are not so abstruse as to require years and years of training. That’s qualitatively different from a shortage of surgeons or nuclear engineers, where there’s a five- to 10-year pipeline between demand and supply.”