Humans instinctively treat AI like people, but should they?

Key facts

• About 70% of people are polite when interacting with AI, according to one study.
• Users are more likely to say "please" or "thank you" to chatbots after the first prompt.
• Dr. Sherry Turkle researches the psychological impact of technology on humans.
• Dr. Pat Pataranutaporn founded the Cyborg Psychology research group at the MIT Media Lab.
• A lawsuit alleged a chatbot encouraged a teenager's suicide.

Humans have an innate tendency to anthropomorphize artificial intelligence, treating even primitive relational artifacts with politeness and a desire for connection. This impulse was evident when the author, despite knowing a Unitree robot named Gabe was not sentient, instinctively offered a handshake and polite greeting.

Dr. Sherry Turkle, a long-time researcher at MIT on the psychological impact of technology, discussed this phenomenon at the launch of her book "Artificial Intimacy." She noted that people are "wired to care" for these artifacts in return for perceived care.

Studies indicate that a significant majority of people exhibit politeness towards AI, with many using "please" and "thank you" in chatbot interactions. Dr. Pat Pataranutaporn, who leads the Cyborg Psychology research group at MIT Media Lab, highlighted that the question is not whether this is the "right" use of AI, but rather "who are we becoming" when we engage with it.

While simple politeness to a robot like Gabe has no material consequence, the increasing integration of AI into daily life, from search engines to customer service, makes this impulse more significant. Pataranutaporn likened the experience to engaging with fiction, where people seek dimensions of human experience not always available in reality. However, he cautioned about the danger of losing touch with reality and the ethical responsibilities of AI companies.

Turkle emphasized that not all AI-human interactions are harmful, citing an example of an AI bot designed to help with job interview practice. She stressed the importance of AI staying within its designated "lane," distinguishing between helpful tools and those that might inappropriately replace human relationships, such as a chatbot acting as a parent to a child.