UNESCO Study Finds AI Distorts Climate Policy Combinations and User Preferences

Written by

in

UNESCO tested 33 mainstream AI models, evaluating their reliability with 656 questions on climate science and policy. The models were most accurate when answering basic climate science questions, with accuracy reaching up to 96%, and up to 94.5% when evaluating individual policy interventions. However, accuracy fell to 33.8% when analyzing combinations of climate policies.

Source:Business News Today · Read the original article

Comments

Leave a Reply

Your email address will not be published. Required fields are marked *