In recent years, chatbots have become much more prevalent and capable. Users think of chatbots as machines but treat them as social actors. The perception of chatbots as machines and as social actors each brings with it both benefits and drawbacks. If it were possible to switch flexibly between the two perspectives, users could retain benefits while avoiding drawbacks. For example, they could commit to the chatbot as a social actor to enjoy its flattery, and still dismiss its utterances as mere machine productions when they trigger a feeling of doubt. We varied reminders about the chatbot's identity to observe which would help most to keep enjoying its exaggerated flattery. We reminded the groups that the chatbot was a machine, a social actor, or that the user was free to switch their perspective on the chatbot's identity. Only if users already believed they could switch was the last reminder impactful.