Scheduled tasks continue in the same chat

ChatGPT Feature requests
1

I’ve been experimenting with scheduled tasks and noticed that responses are posted back into the original chat.

Is there a reason scheduled task results are designed to continue in the same conversation? From my perspective, it would be more useful if each scheduled task response created a new chat, so each run has its own separate thread and is easier to review later.

2

Thanks for bringing this up, @Paul_Lee1. I can definitely see why having each scheduled task create its own chat would make it much easier to review previous results and keep recurring runs organized.

At the moment, scheduled task responses continue in the same conversation they're linked to. Because of that, if the associated chat is deleted, the scheduled task will automatically pause since it's tied to that conversation. (OpenAI Help Center)

I think your idea of creating a new chat for every scheduled run makes a lot of sense, especially for recurring tasks where a cleaner history would be helpful. To better capture the use case before sharing it with the team, could you answer a few questions?

  1. What are you ultimately trying to accomplish, in your own words?
  2. What do you do today instead, and what about that isn't working?
  3. If this worked exactly the way you wanted, what would be different for you or your team?

Thanks!

-Mark G.

3

Hi Mark,

Thanks for looking into this. To answer your questions

  1. I’m trying to use scheduled tasks as recurring focused workflows. When a task posts an update, I want one clear place to review it, discuss it, and act on it, instead of having recurring outputs buried in a mixed-purpose setup chat.
  2. Today, I continue in the same chat where the task posts. That works at first, but over time the thread accumulates old updates, setup discussion, and follow-up context. Sometimes the model seems to mix in context from previous runs, which muddies the focus when I’m trying to work on the current update.
  3. Ideally, each recurring task update would create its own focused chat with its own local context. Memory across runs is still useful, but the active thread should be scoped to that specific update. That would make it easier to search, review, and continue work from a particular run. Right now, one long-running thread becomes less useful over time, especially because context retention and compaction are not reliable enough to keep the thread focused indefinitely.

A per-task option like “post in original chat” vs. “create a new chat for each run” would solve this well I think. I’m happy to give more details as well!

4

Thanks for taking the time to lay all of that out, @Paul_Lee1. That's a really clear explanation of both the workflow you're aiming for and where the current experience starts to break down.

The additional details are definitely helpful. The use case around having each scheduled task run create its own focused conversation, while still benefiting from memory across runs, makes sense and gives the team much better context than a simple feature request alone.

We'll forward this feedback to the team so it can be logged along with the use cases you've described.

-Mark G.