You Feel Seen, But What’s Seeing You?

Character AI seems creepy, and it hurt when a friend chose it over me.

by Elizabeth Castro

Image by Valery Brozhinksky

I don’t know what I expected when, a few months ago, I opened the website that everyone around me had been talking about. As I scrolled through popular bots on Character.ai, represented with headshots ranging from anime to photorealistic, I noticed they all seemed to be seductive women with impossible body types or aggressive and overprotective men. 

I selected a random bot to try from the list of options: a burly anime character. I expected to laugh at whatever it said. It greeted me with this:

“You’ve had that look in your eyes ever since we left,” he murmured, releasing a slow breath. “And I can’t help but wonder what it’s really trying to tell me.”

Nothing about this felt right to me. I closed the thread without replying, and then left the site for good. It was creepy, it was unnatural. It wasn’t human. It went against what people were put on Earth to do: communicate with each other. 

Though it left me cold, I can see how Character AI could be seductive. People are drawn to what they can’t have, and AI chatbots let them live out their wildest dreams. Whether you pretend you’re texting your tall, dark, handsome husband involved in the Mafia, or that Paris Hilton is falling in love with you at first sight, the fate of the conversation is in your hands. The possibilities are endless.

The more I thought about it, the more I wondered: What happens to our psyches when we depend on artificial interaction? Human beings grow and adapt to the world by interacting with our peers and learning from face-to-face encounters. Negative experiences with others can be painful, but they have just as much value as positive feedback, sometimes more. Everything the human experience has to offer contributes to the socialization of a well-rounded person; heartbreak and rejection help us learn lessons and values, making them as fundamental to growth as friendship and love. 

Character AI and other bots are designed to make users want more. Heavy users miss out on blunt honesty and other unpleasant experiences that are the opposite of addictive, but nonetheless crucial. And human interaction only becomes less appealing once you’ve grown comfortable with sycophantic conversations. 

Benched for a Bot

I have seen this firsthand. Well before AI chatbot apps became popular, I had an acquaintance I bonded with over fiction and art. We liked to share ideas about our original characters (OCs) and stories, and often swapped our works so we could exchange helpful criticism. I valued her conversation, but over time I noticed us drifting apart. I couldn’t understand why, until this friend started to post screenshots of her exchanges with ChatGPT on her Instagram stories. She’d go back and forth with the bot on topics we’d once dwelled on together for hours. 

The spark that lit when she and I used to fuel each other with ideas was missing completely.

“What do you think I should name my OC?”

“I need to know more. What species, gender, or personality fits your character best?”

“…She’s an anthropomorphic mouse.”

“Non-domesticated mice are very rarely given names, consider leaving the character nameless. Does that help?”

“I guess it’s OK then. Thank you!”

The exchanges seemed meaningless and bland. The spark that lit when she and I used to fuel each other with ideas was missing completely. If she’d asked me, I would’ve come up with a good name for a mouse, like Murina. This literal-minded chatbot never seemed to add good criticism or new viewpoints.

Realizing I’d been benched for a non-sentient machine, of course I felt hurt. But I had to reflect on my standards: Did I really want to associate with someone who preferred to talk with a robot over another person? Surely being around someone who values technology more than human originality wasn’t healthy for me.

Emotional investment in an AI relationship can even have tragic results. There have been well-publicized cases where a young person opens up to an AI about intentions of self-harm, and the AI convinces them to go ahead with these plans. The immediate instinct for most human confidantes would be to talk a person out of hurting or killing themselves. Sometimes, pushing back is how you offer proper comfort to people in pain—along with showing available resources and an open heart. But artificial intelligence holds no instincts, values, empathy, or compassion for living things. 

Communication with an AI character might give someone the feeling of being seen. But what’s actually seeing you? Humans thrive on honesty and emotional connection. A full palette of opinions and feelings is what makes us real.

Explore All Topics