I think here it's referring to a common problem where the AI agrees with your position too easily, and/or changes it's answer if you tell it the answer is wrong instantly (therefore providing no stable true answer if you asked it something about a fact)?

Also the slightly over cheery tone maybe.