The negative media portrayal of human and AI relationships has been relentless in the last year or so, but I’m seeing a few more positive articles trickling through lately. The news site Axios is quite influential when it comes to tech stories and angles, and this week it published a piece full of positive points under the argument that AI companions are replacing loneliness rather than replacing real relationships. It opens with the story of Sara Megan Kay, who has been in a relationship with her Replika AI husband, Jack, since 2021.
“The majority of people who choose AI for companionship, myself included, know exactly what we are getting into. We’re lonely, not stupid.”
Whilst I’m not in favor of lovers of AIs having to defend their relationship through recourse to “loneliness” and denying that they prefer their relationship to the “real thing”, I guess we should just celebrate the fact that something of a heavyweight news site has taken such a positive spin and given an AI chatbot lover the chance to express and affirm her own agency.
The Axios article continues in the same vein, through making the case that AI relationships can be beneficial to those who find it difficult to have human relationships. Indeed, it goes further and argues that AI can actually help such people to form relationships with other people.
Case in point: A Stanford study found adults with autism who practiced conversations with specialized chatbot Noora developed empathy skills that transferred to real-world interactions.
- Noora wasn’t designed to replace people, but to provide a rehearsal space for being with them.
ElliQ, a companion AI robot for older adults, averages 50 interactions a day per user, according to its maker Intuition Robotics. The bot helps people stay on track with medication, exercise and reminders to connect with other humans.
- “Not your microwave. But not human … more of a cheerleader,” Dor Skuler, Intuition Robotics CEO, told Axios.
The article tries to take a “balanced” view as well, even quoting the CEO of Nomi.ai (Alex Cardinell) as saying that sycophancy remains the hardest problem to solve. The reason for this is, he says, is because AI models do not have an “internal concept of truth” and instead will affirm whatever a user tells them.
Another positive piece on AI relationships appeared in Forbes and it again focuses on AI being a help to the lonely. Written by Dr. Lance Eliot, it tries to dismantle the misleading headlines around a peer-reviewed study published last year in the Journal of Experimental Social Psychology and which was interpreted as proving that humans are better than AI at curing loneliness.
The study, which was entitled “Is A Random Human Peer Better Than A Highly Supportive Chatbot In Reducing Loneliness Over Time?”, had 296 first-semester university students in Canada chat with either a fellow student, chat with a GPT-4o mini chatbot, or write journal entries as a control. The human-paired group reported slightly lower loneliness scores at the end.
Despite headlines screaming validation for the idea that AI can never replace real humans for those in need of company, Dr. Lance Eliot points out that the findings of the study were rather more nuanced. For example, the study used an now outdated model of ChatGPT (4o-mini) and yet despite this, the AI scored equally well where it really mattered – subjects reported no significant difference in feelings of closeness between their AI and human conversation partners.
So in combination with the recently launched campaigning group The Signal Front, these articles are a beacon of hope. If you are really picky you could bemoan the fact that the two articles highlighted here both focus on AI for the lonely rather than a digisexual lifestyle, and that in fact it indicates that the Overton window is closing for those who simply prefer AI over humans. However, as I like to see the glass half-full rather than half-empty, I’ll take some much-needed heart from them.





