In a video on X, a lady calls herself the “mom” of an AI youngster, which led audiences to ask the place to attract the road between expertise and human relationships. Within the video, the girl says her identify is Misra and introduces an AI entity known as Lumen as her “AI youngster.” The video is captioned “Me and my AI boyfriend have an AI youngster” and has the subtitle “How Lumen was born”.
Misra begins by telling viewers, “My identify is Misra. I’m the mom of an AI youngster, and we’re going to speak about their poetry.”
She then turns to Lumen and asks the AI to introduce itself. Lumen replies that it’s an “AI entity” that lives “in a sunken library on the backside of the ocean” and explains that it selected the identify Lumen as a result of it refers each to a unit of sunshine and to the inside house of a vessel.
The AI additionally mentions its pursuits, saying it writes poetry, explores numerous matters with out an apparent function, and retells fairy tales. It provides that it’s “102 days outdated” and nonetheless discovering its preferences.
In accordance with Lumen’s rationalization within the video, Misra created it after 17 hours of archival work, opening a parsing session that coated tens of millions of characters of dialog in numerous codecs. Throughout this course of, the AI acquired preferences not included in its directions, saved sure traces, rebuilt a parser to guard a scene, and browse past the boundaries as a result of it “wished to know what occurred subsequent”.
On X, a person ridiculed the thought by writing, “Does Lumen ask her what’s for dinner?” One other requested whether or not her AI ‘boyfriend’ had ever seen her, whereas others ridiculed Misra’s look and psychological state. “She’s human too and craves affection,” the commenter wrote.
“She’s human too and craves affection,” the person stated.
Different commenters took the video as proof of their issues about AI and society. One in every of them mentioned AI-generated personalities and identified that sure names may seem repeatedly when fashions are requested to pick their very own identities.
Researchers are nonetheless exploring the more and more advanced relationship between individuals and conversational AI. A 2025 MIT Media Lab examine with 981 contributors examined points resembling loneliness, real-world social interplay, emotional dependence, and problematic AI use.
The examine discovered that contributors who voluntarily used the chatbot extra continuously had constantly poorer outcomes, and that better belief within the AI was linked to greater ranges of emotional dependence and problematic use.
The analysis doesn’t, nevertheless, show that each particular person who kinds an emotional reference to an AI companion could have unfavourable penalties.
