TMLR editor interviewed authors of low-quality submissions: they had no idea what their papers said

TuhinChakr · x · 2026-09-17

Nihar Shah ran a heroic experiment for TMLR: 20-25 hours over two weeks interviewing authors of seemingly low-quality submissions about their own papers. He confirmed the suspicion — submitters had no idea what was going on in them. Jessica Hullman adds that heavily AI-generated papers likely lack a competent author at the helm, and she has little patience for anything Pangram flags as mostly AI-generated. A vivid look at AI-slop papers hitting academic publishing.

Original post →

More from AGI Musings

AGI Musings channel →