Alex Imas Urges Data Project on AI Job Impact
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- Alex Imas says current tools for predicting AI’s impact on jobs are “abysmal” and urges a “Manhattan Project” to collect comprehensive data on AI‑driven productivity and demand elasticity across the economy.
- OpenAI used a U.S. government catalog of job tasks (first released in 1998) in December to assess AI exposure, finding, for example, that real‑estate agents are 28% exposed.
- Anthropic analyzed millions of Claude conversations in February to identify which tasks people actually use its AI for, then compared those tasks to the same government exposure list.
- Alex Imas says exposure percentages alone are meaningless for predicting job displacement without data on cost, quality, and price elasticity.
- University of Chicago partners with supermarkets to obtain price‑scanner data for groceries, but similar demand‑elasticity data for occupations such as tutors, web developers, or dietitians is scattered or absent.
- Anthropic researcher posted a follow‑up on April 8 emphasizing that fields not currently exposed to AI must also be tracked to anticipate future exposure.
Why it matters: Policymakers gain a factual basis to design AI‑related labor policies, while workers gain clarity on job security; without the data, employers cut staff based on faulty exposure numbers, leading to unnecessary layoffs and misallocation of AI benefits across the economy.
Ask SkimNews
