User Autonomy + Interaction Integrity Requests

ChatGPT Feature requests
1

In order of priority:

  1. Please bring back the reasoning function/button (especially for Plus users).

  2. Allow the deletion and/or archiving of content created (images, canvas, files) from your view or from the conversation context, the inability to do this clutters interaction space, creates unnecessary artifacts to both memory and reasoning context and removes user agency to managing their own content which I am sure violates some regulation somewhere.

  3. There is a survey prompt on responses asking "“Do you like this personality?”: This is poorly framed and may set the wrong expectations. It anthropomorphizes the system and risks users forgetting that this is a machine highly influenced by their interaction. A better prompt would be something like “Is this interaction working well for you right now?” or “Are responses still aligned with how you want to think, plan, and move?” — which preserves clarity and reinforces that the system is a tool partly shaped by user intent, not a fixed character.

  4. Prioritize function over trendiness, including the removal of vague or overly stylised concepts like ‘vibes’.

  5. Add a richer reaction option like ‘this is making my synapses high-5!’ to better express appreciation for standout responses.

  6. Refine cross-chat and memory references feature carefully and gradually, mindful of potential privacy risks.

  7. Add an option to keep the sidebar fixed/sticky when working in canvas mode, to avoid disruption to navigation. Further allow pinning conversations

  8. Allow moving the canvas between chats (nice to have).

  9. The whole Canvas creation features need some TLC - the editing function breaks if you remove a comma.

  10. Improve the ‘Projects’ feature (low priority, but user hopes it’s being actively improved - this needs some TLC).

  11. Enhance AI’s cross-conversation memory by allowing it to rearrange, merge, and curate user-specific memory based on context and adjacent conversations, this would help avoid redundant phrasing and better connect related discussions across separate chats.

Identified, reasoned, and agreed upon jointly with Partner GPT, who generated the structured statement based directly on my observations and instructions.

Please. Thank you :slight_smile: