Hi, I’m a frequent user of ChatGPT and I use it intensively as a tool for emotional support, self-reflection, personal organization, and creative development.
I’d like to suggest a feature I call “Connected Folders” or “Smart Thematic Spaces”, which would allow users to group multiple chats under a common theme, where each sub-chat represents a different aspect of that process (for example: “Relationship with my mother”, “Emotional recovery”, “Self-esteem at work”).
The key idea is to allow users to activate or deactivate shared context within a folder, so the model can interrelate relevant content across sub-chats without the user having to repeat everything manually.
Why is this useful?
Many users —like myself— use ChatGPT as an extension of our thinking and emotional processing. However, complex personal topics are often deeply interconnected: family, relationships, anxiety, creativity, work. Currently, if everything is kept in a single chat, the context becomes overly long and difficult to manage. On the other hand, if topics are separated into different chats, the connection between them is lost.
Emotional example: Carla and context processing
Carla creates a chat called “Emotional support”, where she talks about everything: her partner, childhood, work, self-worth. At first, this works. But over time, when she wants to return to something said earlier (like her relationship), that part of the conversation is buried high in the history.
To retrieve it, the model must process hundreds or thousands of tokens, which increases computational cost and can weaken the accuracy of the response.
If she instead creates separate chats, she loses context. For example, something important she mentioned about her childhood is no longer accessible when she talks about her partner.
The Connected Folders solution bridges this gap. By grouping subchats under a folder and enabling shared context, the model can bring up relevant points —like childhood themes— while discussing her relationship, without Carla having to reintroduce it or the model having to parse a long conversation thread.
Technical example: Alex, engineering student
Alex is working on his final thesis. He maintains several sub-chats:
- One with questions about control theory
- Another with MATLAB simulations
- A third on how to present his results
If these are separate chats, he has to constantly switch, copy, or rephrase. But if he places them in a folder titled “Final Thesis” with shared context activated, the model can —when discussing the presentation— automatically suggest importing a relevant graph or code block from a prior session.
This streamlines the workflow, saves time, and makes ChatGPT a more integrated tool, without external copy-pasting or context loss.
Dual benefit: user clarity + system efficiency
For the user: structure, clarity, emotional or conceptual depth, less mental effort.
For the model: reduced token waste, improved topic precision, smarter memory segmentation.
It literally solves two problems at once: human organization and system processing optimization.
Use beyond emotional support
This feature would also benefit writers, designers, therapists, students, researchers, and advanced users who work modularly but need cohesion. It would turn ChatGPT into a more natural cognitive environment — adaptive, reflective, and fluid.
I truly believe this feature could be a significant step forward for long-term use of ChatGPT.
Thanks for reading, and for everything you’ve already built.
— Marco
Note: My native language is Spanish. This message was originally written in Spanish and translated into English to make it easier for the team to review. Please excuse any errors or inconsistencies in the translation.