Feature Request: Persistent Assistant Identity and Long-Term Project Continuity
Overview
I would like to suggest expanding ChatGPT’s long-term continuity capabilities so that a personal AI assistant can function not only as an intelligent information tool, but as a consistent long-term professional collaborator across conversations and projects.
For many users, an AI assistant is no longer used only to ask isolated questions or retrieve information. It can become part of an ongoing workflow involving planning, organization, teaching, creative development, decision support, and project management.
In these use cases, remembering facts about the user is useful, but it is not enough. What is equally important is remembering the state and progression of the work itself.
My Professional Use Case
I am a graphic design instructor, and I use my AI assistant extensively while preparing and organizing my courses.
My teaching requires continuity. Concepts need to be introduced in a logical sequence, lessons must build on previous lessons, exercises need to match what students have already learned, and course material frequently evolves over time.
My assistant therefore functions as more than an information source. I use it as a:
-
teaching and curriculum-planning partner,
-
organizational assistant,
-
creative consultant,
-
and continuity partner across ongoing professional projects.
If a lesson outline, teaching sequence, exercise, or project stage has already been completed, I need the assistant to reliably understand that progress when I return in another conversation.
Long-Term Creative Projects
This becomes even more important in long-term creative work.
I am currently developing a long-form fictional series using several AI-assisted creative tools.
The project includes recurring characters, established personalities and relationships, visual identities, story arcs, locations, dialogue styles, music, character voices, completed scenes, works in progress, and creative decisions that may evolve over time.
I use ChatGPT as a creative consultant to help organize the narrative, maintain story continuity, develop scenes, evaluate creative choices, and coordinate different parts of the production process.
For this kind of project, remembering individual facts is not sufficient.
The assistant needs to understand the current state of the project:
-
What has already been completed?
-
What decisions have been approved?
-
What ideas were rejected or replaced?
-
Which character details are currently canonical?
-
Which scenes have already been produced?
-
What stage of production are we currently in?
-
What should logically happen next?
Without this project-state awareness, a new conversation may unintentionally return to an earlier stage and suggest work that has already been completed.
The Difference Between Memory and Continuity
I believe there is an important distinction between personal memory and workflow continuity.
Personal memory answers questions such as:
Who is this user?
What are their preferences?
What do they usually work on?
Project continuity should additionally answer:
Where did we stop?
What has changed since then?
What is already finished?
What is the currently accepted version?
What is the next logical step?
For users who work with ChatGPT regularly, this distinction can significantly affect the experience.
Suggested Capability
It would be extremely useful if users could maintain a persistent assistant or project workspace with a structured, evolving project state.
For example, it could maintain:
Established / Canonical — decisions and information currently considered valid.
Completed — work that has already been finished.
In Progress — tasks or creative elements currently being developed.
Changed / Superseded — earlier decisions that have since been replaced.
Next Steps — the point from which the assistant should continue in the next conversation.
Ideally, the user should be able to inspect, correct, approve, or remove these entries so that continuity remains transparent and under the user’s control.
Why This Matters
The value of a personal AI assistant increases significantly when the user does not have to repeatedly reconstruct the history of an ongoing collaboration.
For educators, designers, writers, researchers, creators, and people managing long-running projects, the ideal experience is not simply:
“The assistant remembers information about me.”
It is:
“The assistant understands where our work currently stands.”
A persistent assistant should be able to continue a professional or creative journey from its latest meaningful state rather than repeatedly returning the user to earlier stages.
In short:
Please consider extending long-term memory into user-controlled project-state continuity — allowing ChatGPT to remember not only the user, but the progression of the work they are doing together.