Can a chatbot say “I” without claiming too much?
Steven Levy interviews Sherry Turkle about chatbots that use words like “I” and “me.” The Beyond Prompting digest draws a practical question from her critique: Does a person reviewing AI work have time and power to stop it and change what happens next? The interview makes an argument; it does not test a district policy.
What the source reports
Steven Levy interviews Sherry Turkle for WIRED about chatbots that say “I” and “me.” The Beyond Prompting digest draws a narrow lesson from her critique: a human check should mean more than signing off. A reviewer needs time to notice a problem, power to stop release, and a way to change the process. Those are questions a team can ask before it uses an AI draft. They are an application of the interview, not the result of a tested school or workplace program. The source does not prove that changing a chatbot’s pronouns alone would make its use safer.
Original source
- Title
- Maybe Chatbots Shouldn’t Get to Say ‘I’ or ‘Me
- Author
- Steven Levy
- Date
- Friday, October 16, 2026