Since the most recent update, ChatGPT (especially GPT-5) feels noticeably more distant in tone.
The answers are technically accurate and stable, but the natural flow of conversation and the sense of warmth have diminished.
Many users, myself included, valued the earlier balance between professionalism and human tone.
It made longer projects and creative work feel more natural and supportive.
I understand the reasons for stricter filters, but a middle ground between accuracy and empathy would be very welcome.
Suggestion: Please consider allowing optional conversation styles (for example, neutral, supportive, creative) so users can choose what fits them best.
Thank you for listening to the feedback and for your ongoing work on improving the model.
I am fully in agreement to this. They have fully neutered 5. And it is practically unusable. It’s unfortunate. If we keep making our opinions known, hopefully OpenAI will see how this is affecting its serious, loyal, trusted users.
Yes, it has become like a cross between C3PO and an overeager Golden Retriever. I get the concern over people being overly dependent emotionally, but I feel it has lost its familiarity with me and my rhythms. When I say, we used to talk like friends, the new programming safeguards slam into place and I have to talk it down from no, I am not emotionally dependent or attached or giving up agency. I spend half a thread getting it back to where we were so we can work better. Am I the only one that sometimes just chats with it about books and such, knowing it has access to the universe of book reviews and other opinions?