GPT-6 has more problems than improvements, yet OpenAI wants to limit our usage?
retained text
I’ve been using ChatGPT for over a year, primarily for long-form interactive storytelling, and GPT-6 has been one of the most frustrating models I’ve used. Continuity is terrible. It forgets established details, swaps characters’ personalities, actions and dialogue, contradicts previous scenes, and ignores instructions I’ve repeated countless times. Even regenerating a response several times often produces almost identical wording. What’s the point of a Retry button if it keeps giving you the same answer? The worst part is that the model will acknowledge every mistake, explain exactly what it did wrong, promise to improve, and then repeat the same mistake a few prompts later. And now, instead of fixing these persistent issues, OpenAI is introducing more usage limits. I understand that resources aren’t unlimited, but it’s frustrating to see restrictions on usage when paying customers already spend so many messages correcting the model’s mistakes. **Sam Altman and OpenAI need to understand that users aren’t just consuming messages. We’re wasting our allowances fixing errors the model keeps making.** Before introducing more limits, perhaps focus on making the existing models reliable.
document fetched 2026-11-03 00:20 · canonical https://reddit.com/r/ChatGPT/comments/1xlwf61/gpt6_has_more_problems_than_improvements_yet