Improving AI Accuracy and Feedback Handling

Hi OpenAI Team,

I’ve noticed an issue with the AI’s handling of numerical calculations and feedback. Recently, the AI provided incorrect information on the equivalency between “crore lakh” and “lakh crore,” which was later corrected but highlighted a need for improved accuracy in calculations. Additionally, feedback provided during a session does not seem to persist beyond that chat.

I suggest prioritizing accurate calculations and improving how user feedback is retained and acted upon. These changes could greatly enhance the AI’s reliability and user experience.