@dalias @petergleick The only ethical thing to do is not to create large language models. Listen to your inner WOPR.