GPT-6 Conversational Quality Regression Across Thinking Effort Levels — Repetition, Lost Nuance, and Failed Corrections

🌐 链接: https://community.openai.com/t/gpt-6-conversational-quality-regression-across-thinking-effort-levels-repetition-lost-nuance-and-failed-corrections/1409642
🏷️ 分组: OpenAI Developer Community
🕒 时间: 2026-10-26 03:35:12
OpenAI Developer Community
GPT-6 Conversational Quality Regression Across Thinking Effort Levels — Repetition, Lost Nuance, and Failed Corrections
GPT-6 Conversational Quality Regression Across Thinking Effort Levels — Repetition, Lost Nuance, and Failed Corrections
I want to emphasize that this is not isolated to a single thinking-effort setting. I have encountered the same underlying problems across the available effort levels: predictable and formulaic writing, failure to recognize conversational subtext, excessive…