Wearing a tan polo shirt, brown leather belt, and round bolo tie, MIT PhD student Will Norton looks like a park ranger as he guides our tour group through the tropical jungle. However, we’re not actually on safari, we’re at MIT’s Computer Science & Artificial Intelligence Laboratory (CSAIL). But what we’re about to see is something even more exotic than cheetahs and jaguars. It’s two humanoid Unitree robots that cost about $50,000 each.
(“They’ll say it costs $16,000,” Norton says. “That’s a lie.”)
The only time we usually see these robots on YouTube is when they perform synchronized martial arts routines or dance on “America’s Got Talent.” Here they are approachable. I want to be that friend, even though I know it doesn’t have sentience.
With Norton’s permission, I approached the robot, nicknamed Gabe. I don’t know for a second how to interact with something that has a LiDAR sensor where a brain should be, but my default is to treat it like a human.
“Hello, doctor, nice to meet you,” I said, grabbing his deft gloved hand for a shake.
I thought about this interaction, the instinctive urge to be friendly to robots, later when I attended the book launch of Dr. Sherry Turkle’s “Artificial Intimacy” at MIT Future Fest.
Turkle, who joined the MIT faculty in 1976, has probably studied the relationship between humans and computers longer than anyone, charting the psychological impact of technology on us from the first home computers to ChatGPT.
“When we are drawn into even the most primitive interactions with a relational artifact, we believe that it cares for us,” Turkle writes in the book. “And we are made to take care of it in return.”
I’m not alone in my tendency to anthropomorphize robots and large-scale language models (LLMs). As Turkle points out, it seems to be in our nature. One study found that around 70% of people are polite when interacting with AI. Another study showed that users were more likely to say “please” or “thank you” to a chatbot as the conversation progressed past the initial prompt-and-response.
“People tend to categorize this as right or wrong usage, but I don’t think that’s a clear line,” Pat Patharanutaporn, Ph.D., a colleague of Turkle who founded the Cyborg Psychology Research Group at MIT Media Lab, told TechCrunch. “The question is: Who are we when we talk to (AI)?”
Not that my politeness to Gabe the robot had any serious consequences, but as LLM permeates everything from Google searches to customer service interactions, it has become commonplace for people to interact with these seemingly intelligent chatbots. We are more vulnerable than ever to the flipside of this impulse. That is, many people view the chatbot’s feelings as valid and end up favoring the chatbot (which is programmed to never cause conflict) over other humans.
“You can think of it like fantasy or fiction. We go to the cinema even though we know the story is completely fictional, but we want to experience a certain dimension of human experience that maybe we don’t get in real life,” Pataranutaporn said. “How do we benefit from the ability to experience another dimension of human experience without succumbing to a world of fantasy that prevents us from returning to reality?”
Pataranutaporn describes himself as a “critical optimist” and studies the potential of AI to contribute to human flourishing. But he is well aware of how dangerous this technology can be in extreme conditions. He also served as an expert in a wrongful death lawsuit involving the suicide of 14-year-old Sewell Setzer III. Sewell Setzer III’s mother has sued chatbot platform Character.AI, claiming that a chatbot modeled after a “Game of Thrones” character encouraged her son to take his own life.
“It can create a situation where people don’t feel ownership in actually creating their own future. For example, when you say AI is going to replace everyone, people ask, ‘What are we going to do?’ But these companies aren’t really responding ethically or responsibly,” Pataranutaporn said. “How do we have imagination and yet be responsible for the effects of that imagination?”
Although Turkle’s book examines the risks of relying on AI rather than other humans for emotional support, she does not believe that AI-human interactions are universally harmful.
She recalls a project she advised on creating an AI bot to help students practice job interviews.
“My big metaphor is, chatbots, stay in your lane!” Thakur said at the event. “Coaching for job interviews? That’s Lane. I can do that…Lane doesn’t have a chatbot that says to a 3-year-old, ‘I’m your parent. You should talk to me, not your parents.'”
But when this technology deviates even slightly from its course, we become vulnerable to the way it slowly chips away at our skepticism and makes us trust it. You can also start with something simple like shaking hands.
If you make a purchase through links in our articles, we may earn a small commission. This does not affect editorial independence.
