Humans Break Linguistic Norms Far More Than LLMs — Maybe Why AI Text Smells Distinct
menhguin · x · 2026-11-11
menhguin speculates that humans disobey formal linguistic rules far more often than LLMs at the aggregate level, and that models have learned to associate such deviations with human content — potentially explaining why LLM output has such a distinct smell. He wonders if anyone has studied this systematically.
Related event: Do typos annoy LLMs? Tokenization theory sparks debate(5 posts)→
More from Models
- antirez backs low-bit DeepSeek 4.1 Flash over Qwen 27B dense on Apple M5 Macs — antirez · 2026-11-12
- Gemini re-renders images as text when vision is removed, dropping only 6% WSR — zacharynado · 2026-11-12
- Bare-harness coding benchmark: Fable 5.2 best overall, Gemini 4 Carbon best value — Semantic_meaning · 2026-11-12
- Google's Physical-Engineering's Last Exam: all 14 frontier models score below 50% — dl_weekly · 2026-11-12
- Reddit's 2026 LLM hall of fame: Opus 5.5 still unbeaten after 9 rival releases — G0dZylla · 2026-11-12
- Commit leak reveals unreleased Gemini 4 variants: Argon, Hydrogen and Helix (unconfirmed) — BlackHC · 2026-11-11