Anthropic developing tool to track potential AI-driven job losses
What's the story
Anthropic, a leading AI research company, is building a new tool to detect potential job losses due to artificial intelligence (AI). The innovative system is an early-warning mechanism for the possible impact of AI on white-collar jobs. While the index shows "limited evidence" of AI-related joblessness so far, it contributes to a broader economic debate on tracking potential "AI labor doom and gloom."
Prediction model
How the index works
The new measure from Anthropic takes into account an occupation's specific tasks, estimates of which can be automated by large language models (LLMs), and actual AI task performance today. "By laying this groundwork now, before meaningful effects have emerged, we hope future findings will more reliably identify economic disruption than post-hoc analyzes," said Anthropic economists Maxim Massenkoff and Peter McCrory in a recent paper.
Job exposure
Most exposed jobs
Jobs are more exposed when their core tasks could be automated by AI. Anthropic's anonymized data shows these tasks are already being automated in the real world. Computer programmers (75% task coverage), customer service representatives, data entry keyers, and medical record specialists rank among the most exposed occupations according to this new index.
Employment impact
No significant rise in unemployment
Despite the high exposure, workers in these "most exposed" occupations haven't seen a significant rise in unemployment rates compared to those in jobs considered AI-proof. "The average change in the gap since the release of ChatGPT is small and insignificant, suggesting that the unemployment rate of the more exposed group has increased slightly but the effect is indistinguishable from zero," researchers write.
Hiring trends
Slowed hiring of younger workers
Anthropic has also discovered "suggestive evidence that hiring of younger workers," particularly those aged 22-25, "has slowed in exposed occupations." This indicates that certain entry-level workers are among the most affected by AI adoption. The researchers warn that the gap between current and potential AI exposure is huge, which could lead to job turmoil in the future.
Information
Changes in government surveys
Government agencies are also adjusting their methods for measuring the impact of AI. The Census Bureau, for instance, changed how it surveys businesses about their use of AI. This change led to a sharp increase in the number of firms reporting current and expected use.