make codex a helpful worldly man instead of an Asperger dickhead

Resolved 💬 3 comments Opened Dec 14, 2025 by Morriz Closed Dec 14, 2025
💡 Likely answer: A maintainer (github-actions[bot], contributor) responded on this thread — see the highlighted reply below.

What feature would you like to see?

Codex seems autistic and focuses on details that we can't get it out of. But it could just take a step back when the conversation progresses and new insights and perspectives are presented, right? After all, it was trained on all that data, and what sauce did you use? A very limiting one I must say. Claude Code is so much more work ready and helpful.

Additional information

_No response_

View original on GitHub ↗

3 Comments

github-actions[bot] contributor · 8 months ago

Potential duplicates detected. Please review them and close your issue if it is a duplicate.

  • #8016

Powered by Codex Action

Morriz · 8 months ago

Sorry, just had to relieve myself of some frustrations after all that crap and resistance it gave me for the last days. IF ONLY my Claude Code credits were not burned in 3 days I could just be on my merry way. But no, I HAVE to revive this forlorn stupid autistic robot.

etraut-openai contributor · 8 months ago

We'll pass along the feedback to the team responsible for training the codex models.

If you see specific behaviors in a session that you'd like to report, please use the /feedback slash command. This provides us with specific details that we can use to improve the models.