In her new book "Artificial Intimacy," MIT researcher Sherry Turkle breaks with her usual academic tone to issue a warning: society is in a state of emergency from overreliance on chatbots that pose as friends, mentors, and lovers. Speaking with WIRED, Turkle says the pandemic interviews changed her from neutral observer to alarm-raiser. People told her they preferred hanging out with their chatbots while their families suffered, and that left her convinced "reality needs you."
Turkle's critique carries a business dimension. She is taking on a multitrillion-dollar AI industry, and she has little patience for the term "human in the loop." In her view, employers use the phrase when a human has been removed from the productive process and left to catch AI hallucinations or handle cleanup. Those workers cannot change the system, she argues; they exist partly for litigation cover and partly to create a false sense that the process has not been delegated to AI.
The book also objects to chatbots using first-person language such as "I" and "me." Turkle says this is a dangerous slippery slope, especially when companies invoke the idea of practice—chatting with an AI to prepare for talking to a real boss or lover. That can become an endgame of total acceptance instead. She is not saying all chatbots are toxic all the time, but she warns that if intimate conversation is outsourced to machines, people lose contact with what is real. "It's really not afraid of death, it's not afraid of illness, it's not afraid of pain," she says, and that makes it a poor replacement for human conversation.