Hi OpenAI team and community,
I’d like to suggest two connected improvements for dot. These are requests for future product behavior, not claims about capabilities available today.
1. A compact, persistent task queue inspired by Age of Empires II
When I have parallel tasks running, such as a physics research project and a video storyboard project, progress is scattered across chat messages. I keep having to ask whether each task is still running. An idle local GPU doesn’t tell me whether work is happening in the cloud, waiting for me, or blocked. This makes it harder to manage usage, budget, and when to game or work on my computer.
The attached Age of Empires II reference image has its four production queue icons circled in red (annotation added with AI). Notice how little screen space the queue occupies. In dot, I would place a similarly compact, persistent horizontal row in the unused upper-left area beside the conversation. The default view should show only small project icons, count badges, and status indicators; task details should appear on hover or click, keeping the conversation as the main workspace. The red circle marks the game reference, not an existing dot feature. The queue could provide:
- Task/project icons and count badges. Icons could be generated automatically or designed/uploaded by the user, making projects easier to recognize and creating more personalization, ownership, and emotional connection.
- Clear states: queued, running, waiting for user input, blocked, and completed.
- The real current stage, such as checking data, generating images, reviewing page quality, or saving outputs.
- Cloud versus local execution, plus CPU/GPU resource indicators where they can be measured reliably. If the system cannot know, say so. Creating a task must not be presented as proof that computation is active, and percentages should not be invented.
- A click-through view with the latest meaningful progress, any blocker or required action, and the latest output.
This would reduce the effort of managing multiple tasks, improve transparency, surface blockers earlier, and give users better control over resources and scheduling.
2. One-time opt-in for automatic continuity with ChatGPT projects
After explicit opt-in, let important dot updates sync automatically into the latest relevant conversation within each corresponding ChatGPT project. Users shouldn’t have to keep pasting links, copying summaries, or re-uploading files whenever work moves between dot and a project.
The sync should carry approved corrections, decisions, key findings, latest files, and open tasks, with source, time, and version information and clear labels for confirmed versus tentative information. Routing should be topic-aware, not simply based on whichever conversation has the newest timestamp. Please include deduplication and loop prevention, a visible sync state, and review/undo controls. Ask the user only when the destination is genuinely ambiguous.
For someone moving between research and creative projects, this would preserve context, prevent lost updates and conflicting versions, and reduce repeated work and wasted usage credits.
The two ideas work together: the queue shows what is happening now, while project sync records the results where the work will continue.
And, on a slightly self-interested note: if you find these suggestions useful, I would happily accept a few extra usage reset cards—so I can keep stress-testing the product and bringing you the next round of ideas. 
