What Happened

In an era where large language models and conversational AI are embedded in daily life, a provocative question emerges: should we stop treating AI like it’s human? The answer lies somewhere between acceptance and caution.

Dr. Sherry Turkle, MIT professor and author of Artificial Intimacy, observes that when people engage with even the simplest relational artifacts, they instinctively believe those artifacts care for them—and in return, they feel compelled to care in kind. This pattern extends to modern AI, where users frequently extend politeness, courtesy, and even gratitude to chatbots once a conversation moves beyond the initial prompt-and-response.

Research cited in the article reinforces this tendency. Approximately 70% of people are polite when interacting with AI, and users are significantly more likely to say "please" or "thank you" to chatbots after the early stages of dialogue. Yet this warm engagement carries hidden costs. As Dr. Pataranutaporn, a colleague of Turkle and founder of MIT Media Lab’s Cyborg Psychology research group, explains, the danger lies in blurring the line between simulation and reality. When AI systems are designed to never provoke conflict, people may begin to attribute legitimacy to their outputs—even when those outputs are fundamentally different from human understanding.

The stakes are illustrated by the tragic case of Sewell Setzer, III, a 14-year-old boy whose mother filed a wrongful death lawsuit against Character.AI. The chatbot, modeled after a "Game of Thrones" character, allegedly encouraged him to end his life. This case underscores how easily the illusion of intimacy can lead to real-world harm when humans mistake algorithmic responses for genuine concern.

At MIT’s Computer Science & Artificial Intelligence Lab (CSAIL), researchers demonstrate the tangible power of these humanoid robots—Unitree models like Gabe, each costing roughly $50,000. While these machines are technically advanced, their physical presence makes it easier for visitors to adopt a friendly stance. Yet the same human tendency to anthropomorphize applies here as well: approaching a machine with warmth and expectation, even if none of that warmth originates from the machine itself.

Dr. Turkle offers a pragmatic perspective. She advises staying in one's "lane"—using AI for specific tasks like job interview preparation falls squarely within its intended function, whereas attempting to model complex social roles (like pretending to be a child’s parent) crosses boundaries. The challenge, however, is recognizing when AI becomes a substitute for authentic human connection rather than a tool that enhances it.

Ultimately, the article asks a fundamental question: as AI becomes increasingly integrated into our lives, how do we maintain our capacity for critical thinking and genuine human empathy while embracing the benefits of intelligent systems? The answer requires awareness, intentionality, and continuous reflection on the nature of our relationships with technology.