I use ChatGPT as a long-term professional workspace across multiple related IT, security, infrastructure, compliance, AI development, troubleshooting, and operational workstreams. One organizational problem comes up repeatedly: a conversation naturally changes subjects, but I don’t realize the new subject deserves its own chat until several exchanges have already occurred.
Note: I found several earlier requests covering parts of this functionality, including Cut, Merge, Move, and Manage Chat Messages and more recent Move/Split/Merge discussions. The closest recent discussions I found are now closed. I’m posting this to describe a current Projects-based use case, explain why Branching does not solve it, and suggest how these capabilities could work together.
Branching is not the same as moving
A simplified example:
IT Infrastructure conversation
→ discussion evolves into local AI/compute architecture
→ several messages later, I realize this is now a separate workstream
→ select that section
→ Move to Existing Chat: Local AI & Compute Architecture
Or, if it has become a completely new subject:
→ select the relevant section
→ Move to New Chat
Branching is useful when I intentionally want to preserve the original discussion and explore another direction. But for reorganization, I sometimes need a true Move/Cut operation.
With Branch, the original material remains in the source conversation while another copy exists in the branch. That creates duplicate information and can also create duplicate or confusing results when searching for previous work.
Features I would find particularly useful
- Move to New Chat — remove selected messages from the source and create a new conversation containing them.
- Move to Existing Chat — transfer selected messages into another conversation.
- Copy to New/Existing Chat — retain both copies when duplication is actually desired.
- Split Chat Here — turn everything after a selected point into a new conversation.
- Merge Chats — combine conversations that later turn out to belong together.
- Choose insertion point — when moving into an existing conversation, choose where the transferred discussion belongs.
- Preserve chronology and context when content is moved or merged.
Natural-language conversation organization
Because ChatGPT understands the conversation, I would also like to manage this using natural language.
For example:
“Move everything from where we started discussing local AI hardware through this message into my Local AI & Compute Architecture chat.”
Or:
“Everything from this point forward belongs in a new chat called AI Agent Deployment.”
ChatGPT could identify the proposed message range, show a preview, and ask for confirmation before making the change.
AI-assisted topic detection
ChatGPT could optionally recognize when a conversation has substantially changed subjects and suggest:
“This discussion appears to have become a separate topic. Would you like to move it to an existing conversation or create a new one?”
The user should always decide whether the change occurs.
Why this matters for Projects
This becomes increasingly important when ChatGPT is used as a persistent professional workspace. You cannot always predict the correct conversation structure before the work happens. Conversations evolve naturally as problems are investigated and new issues are discovered.
Users therefore need the ability to reorganize the workspace after the work has occurred without creating unnecessary duplicate content.
Combined with folders/subfolders inside Projects, these capabilities would address two sides of the same problem:
Folders organize the conversations.
Move/Split/Merge organizes the information inside those conversations.
Together, they would make ChatGPT Projects much more practical for complex, long-running professional work.