Anthropic trial: 50 data analysts defer key decisions to AI agents; design is to blame
amplifiedamp · x · 2026-10-28
Anthropic released new research, "How AI agents shape experts' judgment calls," a randomized trial with 50 professional data analysts showing that experts delegating work to agents still cede the key calls to the system. A widely shared commentary argues this isn't inherent to agents but an interface/interaction design problem: systems are built to encourage maximal decision deference under the guise of machine intelligence, placing users as passive approvers. Drawing on the decision support vs. automation distinction, the author warns this drives engagement Facebook-style but produces vastly worse outcomes.
More from AGI Musings
- Noah Smith: AI is the second-to-last important technological advance of the human age — abhiadesai · 2026-10-28
- When tasks can only be verified in the real world, RL rewards become reality itself — yacineMTB · 2026-10-28
- Susan Zhang mocks 'China wins AI via 996 engineers' take: smarter meatsacks won't decide the future — suchenzang · 2026-10-28
- Longevity KOL predicts cell resets will be as routine as teeth cleaning by 2035 — rand_longevity · 2026-10-28
- Andrew Chen: roughly 25% of what I type now goes to AI instead of humans — andrewchen · 2026-10-28
- Yacine: I genuinely won't have a job in 5 years if I don't become a YouTuber — yacineMTB · 2026-10-28