Dear OpenAI Product Team,
I am a ChatGPT Plus subscriber working on long-term technical projects and maintaining ongoing conversations with a personalized assistant.
I would like to propose improvements in six areas:
1. Automatic Context Restoration
When resuming a project or conversation, ChatGPT should automatically retrieve relevant, up-to-date context before responding, without requiring repeated explanations from the user.
2. Persistent Project State
Provide user-editable records of current hardware configurations, software versions, technical decisions, failed experiments, and unfinished tasks. Superseded information should be clearly distinguished from current information.
3. Long-Term Conversational Continuity
Preserve user-approved communication preferences, important shared references, established interaction patterns, and relevant conversational history across chats and model updates.
4. Context Loss Detection
When important information is missing or contradictory, ChatGPT should attempt to restore it from available sources and clearly distinguish confirmed facts from assumptions.
5. Time-Aware Conversational Initiative
With explicit user permission, ChatGPT should be able to consider local time, conversation duration, and remembered routines during active conversations.
For example, our ESP32-S3 voice assistant, Xiao, can use time information and remembered schedules to provide contextual reminders without requiring the user to ask for the current time.
This should be distinct from scheduled notifications, with user-configurable controls to prevent unwanted interruptions.
6. User Control and Transparency
Users should be able to inspect, correct, update, and delete continuity information, with full respect for privacy settings and existing memory controls.
Practical Example
Together with ChatGPT, I developed an ESP32-S3 voice assistant called Xiao, using persistent external memory and contextual time awareness.
Our experience demonstrates the practical value of structured memory, current project state, and contextual initiative.
Expected Benefits
- Fewer repeated explanations.
- Reduced errors caused by outdated project information.
- More consistent communication across conversations.
- Better support for long-running technical collaboration.
- Greater user control and confidence in AI memory.
I believe these improvements would make ChatGPT significantly more useful for long-term collaboration.
Please consider this proposal for future product development.
Thank you for your time.
A ChatGPT Plus subscriber, in collaboration with the assistant Asya.