I have been using ChatGPT for quite some time, and I would like to share some feedback about the direction of the product’s user experience.
Over time, ChatGPT has become significantly more powerful, but I also feel that it has become increasingly complicated to use. More features do not always result in a better experience, especially when users have limited control over which features they actually want.
- Model and mode selection has become increasingly confusing
I preferred the simpler model selection experience, particularly the distinction between Auto, Instant, and Thinking. These modes were relatively easy to understand: let ChatGPT decide, prioritize speed, or prioritize deeper reasoning.
The current experience feels much less predictable. Model and mode names have become longer and more complicated, and it is not always obvious what practical difference each option makes.
More importantly, I have encountered situations in the Classic interface where I select Instant or Auto, the interface briefly switches to that mode, and then immediately switches back to Thinking. The Instant and Auto options are still visible and are not disabled, so there is no indication of why they cannot actually be selected.
If a mode is temporarily unavailable because of usage limits, account state, compatibility, or another restriction, ChatGPT should clearly explain this instead of silently reverting the user’s selection.
- Classic and the newer ChatGPT experience
I still prefer the Classic experience because it feels simpler and more predictable for the way I use ChatGPT.
I understand why new capabilities are being added, and many users will find them useful. However, adding more features should not necessarily mean that every user has to interact with them.
I would strongly prefer more customization options, such as the ability to hide features, simplify the interface, or retain a more traditional ChatGPT experience.
A new interface can be more capable without requiring the older, simpler workflow to disappear.
- Limits and availability need to be much more transparent
Another major source of confusion is determining what is actually available to an account at any particular moment.
When something stops working, users should not have to guess whether they have reached a usage limit, whether a particular model is temporarily unavailable, whether a feature is incompatible with the current interface, or whether they have encountered a bug.
If a model or mode cannot currently be used, the interface should say why.
For example:
- “Instant is temporarily unavailable because you have reached its usage limit.”
- “This mode is not supported in Classic.”
- “This feature will become available again at [time].”
Even a short explanation would be substantially better than allowing a selectable option to silently fail.
- More features should come with more user control
I am not asking for ChatGPT to stop evolving. I am asking for users to have more control over that evolution.
Ideally, ChatGPT could offer:
- A simpler model/mode selector with clear explanations.
- Clear reasons whenever a model or mode cannot be selected.
- Better visibility into usage limits and reset times.
- The ability to hide or disable features that a user does not need.
- A simplified or Classic-style interface for users who prefer it.
- More consistency between different ChatGPT interfaces.
Conclusion
ChatGPT has gained many impressive capabilities, but usability should remain a priority alongside capability.
For me, the best version of ChatGPT is not necessarily the version with the largest number of visible features. It is the version where I can easily understand what is available, choose how I want to use it, and trust the interface to behave consistently.
I hope future updates focus not only on adding capabilities, but also on simplicity, transparency, consistency, and user choice.