pallest·il y a 3 ans·discussOne solution is placing the ChatGPT response in another model that determines the accuracy of the information and revises where necessary