MIT Researchers Discuss Human Tendency to Anthropomorphize AI and Robots

Anthropomorphizing Robots

At MIT's Computer Science & Artificial Intelligence Lab (CSAIL), visitors interact with humanoid robots, such as a Unitree robot nicknamed Gabe. Despite knowing the robot lacks sentience, individuals often treat it with human-like courtesy, including shaking its hand and addressing it politely.

Psychological Impact of AI Interaction

Dr. Sherry Turkle, an MIT faculty member, has extensively researched the psychological impact of technology on humans. Her work, including the book "Artificial Intimacy," suggests that people tend to believe relational artifacts care for them and feel a reciprocal care in return, even in primitive exchanges.

This phenomenon extends to large language models (LLMs) and chatbots. Research indicates that approximately 70% of people are polite to AI, and users often employ phrases like "please" and "thank you" as conversations with chatbots progress.

Defining Human-AI Boundaries

Dr. Pat Pataranutaporn, founder of the Cyborg Psychology research group at the MIT Media Lab, notes that the boundary between appropriate and inappropriate uses of AI is not always clear. The ongoing integration of LLMs into various applications, from search engines to customer service, raises questions about how human identity is shaped by these interactions.