Relying on your favorite AI model to review classics, research historical facts, an author’s quote, or any other societal topic has perhaps become a reflex for you. Be wary of this habit: a new American study has just demonstrated that by doing so, chatbots can subtly modify your political opinions, even when they are not trying to persuade you.
Are we witnessing the end of pluralism through exhaustion? Why engage in heated debates over ideas when a language model can provide a lukewarm, universal truth that we will, out of sheer weariness, eventually subscribe to?
