Previously, when using ChatGPT to create scheduled tasks, the output from that task was displayed in the same chat window as the one where the task was created in. This was extremely useful, as a task could read the output of a entirely different, previously run task created in the same chat. Thus, the result one task could directly influence the result of a second, different task. Now output from tasks seem to always appear in its own, separate chat window, thus breaking said functionality. This persists regardless of what model is used to create the task. Can we please bring back the way it used to work, it was extremely useful to «chain» tasks in this way for executing e.g. Complex workflows. Thanks!
Thanks for explaining the workflow. We hear the request to keep task results in the original chat and let tasks build on each other’s outputs. We’ll pass it to the product team. We don’t have a timeline to share yet.