I’ve encountered an issue and wanted to see if anyone else has run into the same problem (or found a solution!).
When using custom GPTs, I can’t seem to save their conversations into a “Project” folder. Regular conversations save just fine, but when I try with a custom GPT’s conversation, the option doesn’t seem to work or isn’t available.
Has anyone else experienced this?
If so, have you found a way to save custom GPT conversations into a “Project” folder successfully? Or is this just a current limitation of the feature?
Thanks in advance for your insights and suggestions!
there is a issue while transferring files, the files are order by the time of transferring,
before projects in common pool, it was listed according to recent - which is not how it is within projects, leading to issues searching, we need something like sorting or at least auto arrange by order with newest created post to come up within projects
Thanks for the clarification. I hadn’t put it together yet that it was only the chats with custom GPTs that didn’t have the “add to Project” option.
One note for Open AI to consider in support of including custom GPTs in Projects: The vast majority of my chats are inside custom GPTs. Without the option to add these, it unintentionally dissuades me from using them in… well, projects!
I love the feature - very excited to see where it all goes.
You’re absolutely right that all conversations initiated inside a project share the same custom instructions. However, this doesn’t apply to conversations started outside the project and then added to it later.
Taking this further, I believe OpenAI should seriously consider allowing users to add custom GPT conversations to projects.
Here’s why:
Many users (myself included) rely heavily on custom GPTs for specific tasks within larger projects. These conversations address distinct, task-oriented needs that operate at a different layer compared to general project conversations.
To break it down:
Conversations started inside a project share the project’s overarching instructions, making them ideal for project-wide goals or workflows.
Custom GPT conversations, however, are tailored to specific tasks or specialized workstreams within that project.
This distinction is crucial because projects often consist of multiple layers:
A global layer for the overall scope (project-level custom instructions).
A task-specific layer where custom GPTs shine, delivering targeted solutions for focused aspects of the project.
By enabling users to add custom GPT conversations to projects, OpenAI would unlock the full potential of the “Project” feature. It would empower users to seamlessly integrate specialized, task-specific workflows into their broader project structure—something I believe many users would find invaluable.
AFAIK, this is a correct. In my experience, if I’ve started a chat with one of my custom GPTs and later move it into a Project, the association with that custom GPT goes >poof< and I’m left chatting with a generic model.
Since I have different custom GPTs with different purpose and instructions, all being used on a larger project, it’s painful to have the chats suddenly “forget” what they’re good for.
– edit –
I just did a couple experiments regarding this issue, here are findings:
If the chat resides in the project I can’t use “@” to invoke a specific GPT.
If I move the chat back out of the project, it doesn’t recall which GPT it was originally associated with. However, I can now use “@” to re-invoke the appropriate custom GPT.
Now, this doesn’t suffice as a workaround, since moving the chat out of the project folder removes the model’s access to any background docs or GPT instructions within the scope of the Project level.
Yes very frustrating. I have spent considerable time creating my custom Gpts to help with my business and now I can’t use them in projects. Doesn’t make sense.
to the community!