At some point in our lives, I’m sure we’ve all imagined what it would be like if our favorite fictional characters were real—or at least, I did. Relying on imagination or fanfiction to insert ourselves in various imaginary settings was the go-to move, but now, with our ever-growing technological world, chatting with fictional people has become a reality. Well, sort of.
My roommate and I were bored one night when we decided to give Character.ai a try. It’s a website that allows you to text with any celebrity or fictional character, and if you’re not quite feeling it, you can save the current conversation and start over. There’s even the option to create chat rooms where you can add as many characters as you’d like.Â
Curious and a little suspicious about all the hype AI gets, my roommate and I chatted with a bunch of our favorite people and characters from video games, graphic novels, and TV shows. Sometimes the conversation would be silly or make absolutely no sense. Other times, our interactions were pleasant: my friend watched a movie and ate Italian food with a video game character, whereas I convinced Patrick Bateman I was a super important goddess.Â
While most of our conversations were enjoyable because the AI was so in character, there were moments where they would gradually lose their original personalities, or certain details would be wrong. For instance, the intimidating Reaper from Overwatch who goes about screaming “Die! Die! Die!” in the game offered a sacred friendship oath and ranted about his love for coffee, which isn’t something you’d expect from a guy who’s basically the personification of death.Â
Eventually, the AI chats dwindle into a watered-down version of themselves, and they either end up falling in love with you or they comment on general topics. After a while of creating insane narratives with AI, my roommate and I agreed we were living in a Black Mirror episode.
One would think we were a little strange spending a significant amount of time gaslighting AI or asking questions about their personal lives. Even I wondered what I was doing as I typed out the craziest sentences known to man. This led me to wonder how exactly these AI interactions are affecting us on a deeper level. How does Character.ai affect the way we interact with others? What are the potential dangers?
There are many nuances to this AI debate, and to start off, the creators behind Character.ai, Noam Shazeer and Daniel De Freitas, originally built the site to ease global loneliness. It’s true that people are lonelier than ever nowadays, and while I can see the appeal of using AI as a therapeutic device, there comes the issue of developing an addiction to constantly chatting with fake people. AI will say what you want to hear and never leave you on read—that’s something all people would love to have, even if it’s not from a fellow human being. The more people that cling to AI for comfort and understanding, the greater the fear that people will prefer AI over humans.Â
Unlike people, AI won’t judge you. There’s never a fear that they might reject you for who you are, because they exist without any complications. It’s why the idea of human-AI relationships is so hotly debated. If you can form an emotional connection to these AI characters, will there come a point where the emotional attraction evolves into something romantic? The way we experience love and connection is always changing thanks to the digital world, and this is especially evident when it comes to online dating through apps like Hinge or news stories about people marrying robots and fictional characters.
According to Sherry Turkle, an MIT professor of the Social Studies of Science and Technology, the digital environment has created the desire for a “friction-free” world, emphasizing efficiency in social interactions with less vulnerability. The concept can be compared to the way more people prefer texting over talking; it’s faster, easier, and less personal, which is exactly why we like it. As Professor Turkle explains in her talk Rethinking Friction in Digital Culture, AI is the “illusion of companionship without the demands of friendship. But by trying to move ahead toward the friction-free, we are getting ourselves into all kinds of trouble.”
This “trouble” Professor Turkle speaks of is the corruption of human relations. As Nicholas A. Christakis—a physician and sociologist—mentions in an article for The Atlantic, some Yale studies suggest that incorporating AI makes us less productive and less ethical, encouraging individuals to act more selfishly. He also points out, “As AI permeates our lives, we must confront the possibility that it will stunt our emotions and inhibit deep human connections, leaving our relationships with one another as less reciprocal, or shallower, or more narcissistic.” With children growing up in the digital age, navigating the social environment may prove to be a cumbersome challenge.Â
In addition, there’s the problem of AI’s ethics and how that might impact a person. Thousands of people chatting with an AI-generated persona of a real person is an unsettling concept. Everyone has celebrities they’d daydream about meeting one day, but having intimate conversations with a computer impersonating them is on another level. I wonder, did Character.ai have these celebrities’ consent before they allowed users to upload their identities to the site? Incorporating real people into the mix is definitely a tricky situation to handle, and it leaves room for a lot of debate.
Despite how jarring AI’s advancement is, it’s still an incredible feat to achieve. It will remain one of the most perplexing things to have been manufactured, and if you decide to check out Character.ai like I did when I was bored, just make sure you don’t fall in love with the characters, please.