The way I understand it is: Those three responses together create a single coherent response, and it’s the response of Bing (which it can’t insert into the normal speech bubble because of the filter). The three options are meant to be something the user can say, not something Bing says.
The way I understand it is: Those three responses together create a single coherent response, and it’s the response of Bing (which it can’t insert into the normal speech bubble because of the filter). The three options are meant to be something the user can say, not something Bing says.