AI
Anthropic staff expect far faster AI progress and greater job disruption than the public
An Anthropic survey finds a wide gap between its staff and US adults in expectations for AI by 2030. Ninety-one percent of staff, versus 39% of adults, expect AI to make a Nobel-level discovery, while staff estimate that AI could perform 75% of their own work compared with 28% among workers overall. Respondents who foresee more automation also expect higher unemployment. The results measure beliefs, not demonstrated capability, and Anthropic staff are a highly selected sample.
Editorial summary · Publisher article
What matters
- Anthropic staff expect AI adoption on 65% of technically feasible tasks by 2030, compared with 42% among US adults.
- Fifty-nine percent of staff say AI cuts suitable tasks to a quarter of the time or less, versus 26% of employed AI users.
- Workers who expect AI to perform at least 80% of their work forecast average 2030 unemployment of 12.2%.
- The report’s model maps survey responses to a projected 17.1% 2030 GDP growth rate for the median staff scenario, versus 5.4% for the median US adult scenario.
Expectations reflect different exposure
The report says both groups largely agree about current AI capabilities, but Anthropic staff expect faster improvement, wider adoption and greater autonomy. Workers who already use AI regularly are more optimistic than nonusers, suggesting experience may shift public expectations as adoption spreads.
Survey results are not labor-market outcomes
Anthropic staff expect the company to keep hiring even while believing AI can perform most of their work. The authors therefore distinguish task automation from job elimination, while noting that the survey cannot establish whether those expectations will become reality.
Context & caveats
- The staff sample is internal to Anthropic, and the growth figures are outputs from a model applied to survey responses rather than observed economic results.
Why this made the edition
A detailed survey-based analysis compares public and Anthropic staff expectations about AI capability, adoption, productivity, employment and growth, while identifying that the figures are expectations rather than forecasts or measured outcomes.
Source material