I’d like to share some UX feedback after actively using ChatGPT Projects as a long-term thinking and knowledge management environment.
The core issue is not individual features, but “continuity of thought.”
Here are the main friction points I experienced:
- Inconsistent draft/input persistence between Web and App
-
Web keeps unfinished input reliably
-
App sometimes loses draft input
This creates a strong feeling of “thought interruption.”
For users doing long-form thinking or structured project work, draft persistence is critical.
- UX inconsistency between Direct Chats and Projects
Examples:
-
Different reaction systems (e.g. “Nice”)
-
Different interaction feeling
-
Different lightweight behaviors
From a user perspective, both are still “ChatGPT,” so behavioral differences create cognitive load.
- Projects are evolving into knowledge workspaces
Projects are not just folders.
For advanced users, they become:
-
thinking environments
-
structured knowledge systems
-
iterative workflows
In that context, UX consistency becomes more important than adding isolated features.
My suggestion:
“Unify the UX philosophy, optimize the UI per use case.”
Examples:
-
Consistent draft persistence everywhere
-
Consistent interaction basics
-
Preserve the deeper, focused nature of Projects without turning them into social/chat spaces
One more future-looking request:
Partial conversation extraction / re-edit based migration between projects would greatly improve knowledge refinement workflows.
In short:
“UX is not just usability — it is continuity of thought.”
Thank you for building this platform.