I’d like an option to start a temporary Voice conversation from an existing ChatGPT thread, using that thread’s context, and then bring only an approved summary back into the main conversation.
For longer projects, I sometimes want to talk through an idea, ask questions, or explore alternatives without adding every spoken exchange to the ongoing text thread. Starting an unrelated conversation can mean having to explain the background again. Keeping everything in the original thread can make it harder to follow the work and distinguish settled decisions from exploratory discussion.
The workflow I’m looking for is:
- Start a Voice side conversation from the current chat, with access to its relevant context
- Keep the spoken back-and-forth separate from the main thread
- At the end, generate a concise summary of conclusions, decisions, corrections and unresolved questions
- Let me review and edit that summary, then approve its addition to the original chat or discard it
- Continue in the original thread using the approved update
Dot already supports calls using selected conversation context, and Codex has side chats with separate transcripts. The specific addition I’m requesting is a Voice side conversation followed, once the Voice session ends, by a written summary that I can review and edit before choosing whether to add it to the original thread. Only the approved summary should update that thread’s ongoing context; the exploratory exchange should remain separate.
The review step is important because spoken discussions can include misunderstandings, transcription errors and ideas that are later rejected. Those shouldn’t automatically become settled context.
This would let me use Voice naturally while keeping a long-running project conversation organized and under my control.