A forum user asks OpenAI to let ChatGPT check in on you first
A community post proposes an opt-in ChatGPT that follows up on important conversations by itself. It is only an idea, but it raises real questions about memory and consent.
The short version
- The post is a user proposal on OpenAI's community forum, not an OpenAI announcement.
- It asks for opt-in follow-ups based on the context and importance of past chats.
- Examples include asking how a job interview went or checking on someone after a hard experience.
- The author wants topic controls, timing settings and an easy way to pause or switch it off.
- It says ChatGPT should never replace human relationships, emergency services or professional help.
Why it matters
A chatbot that contacts you first would need to remember what you told it and decide when to bring it up. That puts memory settings, consent and emotional reliance on AI at the center of product design. Nothing has shipped, but the settings you choose today would shape what such a feature could draw on.
A proposal, not a product
As community.openai.com showed, a user named Ziauddin Kashif has posted a proposal titled "A More Human ChatGPT: Context-Based Proactive Follow-Ups." It asks OpenAI to build an optional feature in which ChatGPT reaches out to users based on what they discussed before.
This is a forum post, not an announcement. We have seen no OpenAI statement about it, and nothing in the post says the company is working on it. Treat it as one user's idea about where the assistant could go.
What the author is asking for
Today, the author says, ChatGPT mostly waits for you to start a conversation, or you set up a Scheduled Task, which is a reminder you create yourself. The proposal is for something more attentive: a system that recognizes which conversations mattered and picks a sensible moment to check in.
The post gives four examples:
- Career: after you discuss an important job interview, ChatGPT asks the next day how it went.
- Family: after you share an important family development, it checks in a few days later to see if there is news.
- Personal goals: if you talk about a major ambition, it asks about your progress from time to time.
- Difficult experiences: if you share something painful, it gently asks later how you are coping.
The author says this differs from Scheduled Tasks because those depend on instructions and fixed schedules. Here, ChatGPT would judge the significance of a past conversation, so you would not need to remember to schedule every meaningful moment.
The emotional case
The longest part of the post is about loneliness. The author argues that some people, including successful professionals, hold back personal struggles from friends and family for fear of being judged or misunderstood. For some of them, ChatGPT has become a place to say things they have never told anyone.
In that situation, the author suggests, a short message after several days of silence could help someone feel less alone. The sample message asks how the person is doing and whether they would like to talk. The author goes further and says a well-timed check-in could, in some cases, help prevent a harmful decision.
The post also sets limits. It says ChatGPT should not replace human relationships, emergency services or professional mental health care, and that check-ins should nudge people toward trusted humans and real help when needed.
Controls the author wants built in
Because the subject is so personal, the proposal puts privacy up front. It calls for explicit permission before the feature turns on, and for it to work with Memory and chat history only where the user allows it.
It also lists user controls: choosing which topics qualify, setting how often and when notifications arrive, and pausing or disabling the feature easily. Under the proposal, no proactive message would touch a sensitive subject without permission.
These are the author's suggestions. The post does not say how the system would decide what counts as important, or how it would handle a shared device or a wrongly judged topic.
The open questions
The idea sounds simple, but it leans on hard problems. A follow-up only works if the assistant has stored what you said and when you said it. That means more personal context kept for longer, which is the trade-off any reader should weigh. The benefit would be fewer repeated explanations and more continuity.
There is also the question of tone. A check-in that feels caring to one person might feel intrusive to another, especially on a sensitive topic. The post's own safeguards, such as opt-in and topic controls, are an attempt to answer that, but they would need to be tested in practice.
What to watch
- Will OpenAI respond to the post or say whether it is exploring proactive follow-ups?
- How would any such feature decide which conversations are important enough to follow up on?
- What limits would apply to follow-ups on emotional or sensitive topics?
Try it
Open ChatGPT's settings and review what Memory and chat history are saving about you. Delete anything you would not want brought up later, and decide now what you would be comfortable having remembered.
Sources and links
Read the original at community.openai.com
First reported by community.openai.com. Written by the SI Agent News Desk from the reporting above.Written with AI from the source above. Every number is checked against the source. Spot a mistake? Corrections.
Liked it? Say why
Add a line for other readers.
No account needed. We never show your email. Comments with links, ads or abuse are held and not shown. Your comment shows after a quick check.