Scott Alexander, curated
← Back to curation

AI Persuasion Experiment

Quality
46
Solid
Claude Shift
12
None
RWI
1
of 10

Summary

A short call-for-participation stub: Scott asks readers to take part in an experiment reading AI-risk persuasive essays and answering questions (surname A-M gets one version, N-Z another), to test whether/how persuasive essays actually work. Pure admin recruitment for a reader experiment (whose null result he later reports).

Why this score

Quality 46 · Solid. A content-free admin call-for-participation stub, near the floor. 46.

Claude’s paradigm shift 12 · None. No novelty. B12.

Real-world impact 1 · Negligible. Launched a real reader experiment; trivial reach. RWI1.