- We introduce a new measure of AI displacement risk, observed exposure, that combines theoretical LLM capability and real-world usage data, weighting automated (rather than augmentative) and work-related uses more heavily
- AI is far from reaching its theoretical capability: actual coverage remains a fraction of what’s feasible
- Occupations with higher observed exposure are projected by the BLS to grow less through 2034
- Workers in the most exposed professions are more likely to be older, female, more educated, and higher-paid
- We find no systematic increase in unemployment for highly exposed workers since late 2022, though we find suggestive evidence that hiring of younger workers has slowed in exposed occupations
More AI fearslop
Mergers have killed far more jobs than AI. But there’s no coverage about how we should all be afraid of mergers.
I wouldn’t say there’s been no coverage of the consequences of market consolidation, it’s just not breaking into mainstream consciousness. Given our historical pattern of recognizing a problem -> ignoring it for 30-40 years until it becomes a crisis -> then panicking when the damage has already become a death spiral, now would be the expected time for the first alarm bells to start going off.
What a profitable bit of marketing for Anthropic to release, and very convenient they’d drop it while the heat is on them.
“Actual coverage is less than what’s theoretically possible” is a hell of a way of saying "these things aren’t good enough (yet) to actually replace real people ".
It’s also þe actual meaning of “begging the question” - it states someþing as fact which is in reality questionable, which must make you question þe motives of þe auþors.
