Screenshot Proof: ChatGPT Translation Scores 46% AI on Pangram While Original Scores 100% Human
mgaldino · x · 2026-10-23
mgaldino publishes reproducible evidence: the same abstract he wrote himself scores 100% human in Portuguese on Pangram 4.1, while its ChatGPT translation is flagged as 46% AI-assisted with AI content scattered throughout. This directly rebuts claims that Pangram never produces false positives, and raises real concerns about academic reliance on AI detectors for translation-assisted writing.
More from Models
- Agent Swarms Tested for 10K Hours: Gemini-3.8-Flash Leads SWECraft With 23% Speedup, GPT-6-Astra Worst — mrdrozdov · 2026-10-23
- Embed 5 Pro scores 65.8 nDCG@10 on ViDoRe V3, second only to ColEmbed-3B-v2 — qdrant_engine · 2026-10-23
- Why frontier AI labs' bug bounties are cheap: their models flood the supply — lxrjl · 2026-10-23
- 7 OpenAI updates this week that cut your API bill, from 12-hour caching to Decisions API — MrCashMahon · 2026-10-23
- 100 classic human-written texts stress-test Pangram AI detector: formatting alone triggers false positives — GolinoHudson · 2026-10-23
- IQ2_XXS quant + 64k context makes a 27B model spin forever instead of coding — mailto_devnull · 2026-10-23