Hello OpenAI Community,
I’ve been using ChatGPT for a while now and recently encountered an issue that I believe needs addressing. During my session with the model, I provided a significant amount of information to it. However, I noticed that it struggled to retain the information within that single session, leading to inaccuracies and inefficiencies in the subsequent interactions.
- Context: I was sharing detailed information about an entity called “SOAP Health” and its relationship with an individual named “Keika.”
- Expected Behavior: ChatGPT should be able to reference and recall information shared earlier in the same session to ensure consistent and accurate responses.
- Observed Behavior: The model failed to accurately recall the role of “Keika” despite the information being provided earlier in the same session.
The ability to retain and recall information within a session is crucial, especially when users are sharing complex and detailed information. The observed behavior can lead to inefficiencies and potential miscommunication, making the user experience less seamless.
Suggestion for Improvement:
It might be beneficial for the model to have a more robust mechanism to reference immediate prior information within a session, ensuring that it can consistently provide accurate and contextually relevant responses.
I understand that no model is perfect, and there are limitations. However, highlighting this issue might help in refining future iterations of ChatGPT.
Thank you for your attention to this matter, and I look forward to any feedback or insights from the community and the OpenAI team.