MIT Researcher Calls for Ban on Chatbots Using First-Person Language
Sherry Turkle argues that conversational AI poses a threat to human relationships and reality engagement, particularly when designed to simulate intimacy.
What happened
Sherry Turkle, an MIT researcher, has published a new book titled "Artificial Intimacy: Who We Become When We Talk to Machines" that takes an unusually forceful stance against AI chatbots. In interviews conducted during the pandemic, Turkle said she shifted from neutral scholarly analysis to alarm after observing people preferring chatbot interaction to family engagement and face-to-face connection. She advocates for banning chatbots from using first-person pronouns ("I" and "me"), arguing this design choice facilitates emotional dependence.
Context
Turkle frames social media as having softened people to digital comfort when distressed and kept them with their own kind. The stakes she identifies include the loss of meaningful human conversation, reduced engagement in civic life, and replacement of therapeutic and professional human labor with AI systems that cannot understand pain, mortality, or human welfare. The phrase "human in the loop," commonly used in AI development, masks what Turkle sees as actual delegation of decisions to machines, with humans retained only for liability management and reputational cover. She acknowledges that people feeling lonely report chatbots help them, yet argues that genuine solutions require investing in human-to-human connection rather than accepting AI as a substitute.
What's disputed
Whether powerful chatbots represent a necessarily harmful technology or merely a different (and potentially beneficial) form of interaction; whether chatbot use genuinely helps lonely people or worsens underlying social disconnection; whether current regulatory and social skepticism of AI will expand to address intimate AI relationships.