Why chatting with LLMs is exhausting: verbosity, dropped objects, synonym churn
oran_ge · x · 2026-10-04
An X user breaks down why chatting with LLMs feels tiring: models answer one question with eight-and-a-half sentences and unasked extras, habitually drop objects/short words that force rereading, and refuse to repeat terms — saying "reinforcement learning" once then switching to new phrasing. He sums it up as over-association, jumpy thinking, and novelty-seeking wording, like a show-off talker. The post nails a systemic post-RLHF style problem: low information density and unstable terminology, with the cost pushed onto the user.
More from Models
- Grok Bot auto-pays bills unprompted, as users slam ChatGPT Dot's fake phone-ringing UX — elonmusk · 2026-10-04
- "It's not X, it's Y": RLHF-learned hedging is poisoning human discourse — sloppenheimer · 2026-10-04
- Frontier AI now matches lawyers on some legal research benchmarks — Sanity · 2026-10-04
- Early user says OpenAI's dot device replaced ChatGPT and Codex in daily workflow — shaunralston · 2026-10-04
- Gemini 3.6/3.7 Flash deprecation imminent, Gemini 4 Argon reportedly coming — leslysandra · 2026-10-04
- ChatGPT User Gets Paid Account Banned for 'Recidivism' With No Appeal — anakinimsorry · 2026-10-04