Anthropic rolls out Interviewer tool and opens data Anthropic introduced Anthropic Interviewer, an AI‑driven system that runs automated 10‑15‑minute interviews on Claude.ai and publicly released the full transcript dataset on Hugging Face for research use [1][5].
1,250 professionals surveyed across three cohorts The pilot interviewed 1,000 workers from a broad occupational mix, plus 125 creatives (writers, visual artists, etc.) and 125 scientists (physicists, chemists, data scientists, among 50+ fields) recruited via crowd‑worker platforms [1].
General workforce reports productivity gains but feels stigma 86% say AI saves time, 65% are satisfied with its role, yet 69% note workplace stigma and 55% express anxiety; 48% envision future jobs focused on overseeing AI systems [1].
Creative professionals see speed and quality boosts amid economic worries 97% report time savings, 68% claim quality improvements, while 70% grapple with peer judgment and fear of displacement, with many fearing they must sell AI‑generated content to stay afloat [1].
Scientists cite trust gaps yet desire deeper AI partnership 79% flag trust and reliability as barriers, 27% mention technical limits, but 91% want more AI help—especially hypothesis generation—though current use stays limited to writing, coding, and literature review [1].
Interview tool proves scalable but study has biases Anthropic Interviewer completed the large‑scale test at far lower cost than manual interviewing, yet the sample’s crowd‑worker recruitment, self‑report nature, and snapshot design limit generalizability and may overstate positive attitudes [1].