ChatGPT Let the User Update a Running Task but Not Talk Beside It or Hear Progress Early

Human Experience Reform

Human Experience Reform public Incident record

Stable Incident ID: HXR-2026-0824-0002

ChatGPT Long-Running Task Supported Mid-Run Updates While Side Conversation and Read Aloud Remained Unavailable

This is an opportunity record that preserves a current positive interaction and two separable ways it could be extended.

Bounded Incident: the contributor did not present this primarily as a complaint. HXR does not infer motive, technical cause, prevalence, or material injury.

What happened

  1. ChatGPT was performing a longer-running Text Mode task and displayed progress text.
  2. The contributor could send additional input that was folded into the active work and regarded that as a useful improvement.
  3. The contributor wanted a related side discussion that would not alter the running task, but no separate route was available in the observed interaction.
  4. The contributor opened another ChatGPT thread for that discussion.
  5. The contributor wanted to hear already-visible progress text while doing another physical task, but Read Aloud was unavailable until the larger turn completed.

Positive practice preserved

Mid-run Update: the user could correct or steer work already underway instead of waiting for completion and restarting. HXR treats this as positive current performance.

Remaining opportunities

Side conversation: there was no observed way to discuss the project alongside active execution without changing it, so the user managed a second thread.

Read Aloud: useful stable progress text was visible but not listenable until the whole active turn finished.

Why this matters

Users may need to correct current execution, save a message for later, or think aloud beside the work. Those are different intentions. Audio access also lets people continue the interaction while walking, doing chores, using their hands, or reducing visual attention. Live Voice Mode is not equivalent for users who rely on editable composition before submission.

Evidence and limits

  • Direct contributor testimony supports the interaction and second-thread workaround.
  • OpenAI release notes document current mid-run Update behavior.
  • OpenAI-hosted community requests and independent reports describe side-thread, queue, and early Read Aloud desires.
  • No population prevalence, shared cause, or product roadmap commitment is established.

Related Evaluations

Current workarounds

Open a second thread and manually carry useful context back later; visually read progress while work continues or wait until completion for Read Aloud. These workarounds function but fragment context and delay audio access.

OpenAI: respond or add evidence

An authorized representative may clarify current controls, supply positive or challenging evidence, explain constraints, propose reform, document implementation, or request verification.

Start a private organization response

Have you encountered this?

Describe whether you wanted to update current work, queue something for later, have a side conversation, or hear stable progress text; what the interface allowed; and what workaround you used. Remove private conversation and account details.

Back to blog

Leave a comment

Please note, comments need to be approved before they are published.