Economist builds AI interviewer workflow to flag authors who can't explain their own papers
soumitrashukla9 · x · 2026-09-18
Economist John Horton built a workflow on Expected Parrot: respondents upload their paper PDF, a model generates an interview guide, and an AI interviewer quizzes the author with transcripts captured for analysis — to flag "authors" passing off AI work they don't understand as their own. Inspired in part by TMLR having editors interview submitters, he sees applications at research conferences and beyond, as academia buckles under AI-generated slop.
Related event: Economist Uses AI Voice Interviews to Test Paper Authors(2 posts)→
More from Apps
- OnCo launches an open cited map of cancer with 11,023 linked objects — dotey · 2026-09-18
- Meta Chief AI Officer Alexandr Wang Hypes New AI Product Muse as Smoother Than Grok — alexandr_wang · 2026-09-18
- Why everyone is getting excited about personal AI agents — The AI Daily Brief · 2026-09-18
- AI is a mediocre writer but perhaps the best reader a writer can have — TheZvi · 2026-09-18
- There's no 'accuracy' in AI search tracking — only precision, argues Mike King — VeryWellVersed · 2026-09-18
- All Day TA launches cheat-proof AI quizzes: pass-based grading with AI tutoring on misses — Afinetheorem · 2026-09-18