#Bug Title Memory Full warning is unclear about what still works and what stops working

1 messages · Page 1 of 1 (latest)

lofty waveBOT
#

Reported by @snow bison

Bug Report: Bug Title Memory Full warning is unclear about what still works and what stops working
`Steps to Reproduce`

Steps to Reproduce

  1. Enable:

    • saved memories,
    • chat history reference,
    • project usage,
    • and other personalization/memory features.

  2. Fill memory storage until ChatGPT shows:
    “Memory full. New memories can't be saved.”

  3. Continue using ChatGPT Projects, memories, and chat history features.

  4. Observe that several memory-related systems still appear enabled and active.

Additional observations:

  • The UI does not clearly explain which memory systems are still functioning.
`Expected Result`

Expected Result

The Memory Full warning should clearly distinguish:

  • saved memories;
  • chat history reference;
  • browser memories;
  • project context;
  • and whether only new memory creation is blocked.

Example clearer wording:

“Memory storage is full. Existing saved memories can still be referenced if enabled, but new memories cannot be saved until some are deleted. Project files, project instructions, browser memories, and chat history are separate systems.”

`Actual Result`

Actual Result

The warning says:

“Memory full. New memories can't be saved, so responses may feel less personalized.”

But the UI still shows multiple memory/context systems enabled, making it unclear what functionality is actually affected.

`Environment`

Environment ChatGPT Pro Projects enabled Saved memories enabled Chat history enabled Atlas browser testing macOS