Pangram defends AI detector: false positive rate far below 1%
anecdotal · x · 2026-11-10
Responding to criticism over AI detection debates, Pangram argues its false positive rate is far below 1%, citing tests on internet text, pre-2022 arXiv/NBER papers (zero detections), and internal student-writing experiments with near-zero FPR.
The exchange also captures frustration with LinkedIn "AI snake oil" consultants publishing AI-written manifestos about how universities should adopt AI.
More from Fun
- Fable 5.2 Demo: Hoodie Models Step Out of Product Strips and Dance on Fashion Sites — EricBuess · 2026-11-11
- Guy trolls Burger King's AI hiring bot by wearing a shoe on his head in interview — craigmullins · 2026-11-11
- Chinese SMBs play cat-and-mouse just to keep using Claude Opus — xiaosun86 · 2026-11-11
- Scoreboard meme: Gemini 4 Carbon 1, human patience 0, as releases pile up — patience_cave · 2026-11-11
- Google says 93 parallel agents built a working OS in 12 hours for under $1K — zacharynado · 2026-11-11
- 'Long descriptive names are a good habit' dunked on: enterprise Java did it for decades, universally hated — seatedro · 2026-11-11