You can't bypass AI detectors — a random word list still scores 100% AI on Pangram
rubenhassid · x · 2026-11-17
rubenhassid argues AI detectors can't really be bypassed, with Pangram being the only one he trusts (no affiliation).
- His test: prompt an AI for the "most normal-but-random" 50-word list (chicken, rice, salt, pepper...). Pangram still flags it as 100% possibly AI — because model word choices are never truly random, and detectors pick up that systematic bias.
- He claims one specific bypass method exists, shared with pros and cons, mainly as a lesson in how AI generation works.
- Attached charts: a month of Claude Ultra ($200) equals $14,662 of API-priced Opus 5.5 alone ($10,353 if Fable 5.2 is used first), and METR data showing task-length capability doubling every 1.7 months since Opus 4.5, with Anthropic's internal "Model 4" estimated at 38 hours.
More from Models
- Reuters exclusive: China taps US AI models via distillation to train homegrown rivals — ns123abc · 2026-11-17
- IFM releases K2-Horizon: six fully open LLMs up to 375B MoE, trained on ~10T synthetic tokens — IFM · 2026-11-17
- Optimization researcher: Astra proved with one prompt what we couldn't — bremen79 · 2026-11-17
- Gemini's AI Overview caught fabricating brake rotor dimensions for a Fisker Ocean query — AWTom · 2026-11-17
- Dev Slams Claude for Building a Content Repository With No Way to Add Content — dejanseo · 2026-11-17
- Tencent Hunyuan highlights user demo: Hy4 Flash builds a stunning castle with only 75k tokens — TencentHunyuan · 2026-11-17