pallest·قبل 3 سنوات·discussOne solution is placing the ChatGPT response in another model that determines the accuracy of the information and revises where necessary