Should we really rely on the internet for social connection?
In the age of the smartphone, it’s increasingly common for people to turn to alternative forms of social connection. However, some are finding this not in the people and communities around them, but in the online world.
This is nothing new; people have been forming connections online since the advent of the internet. Over the years, online forums and social media have been a saving grace for many, from motherhood blogs to fashion newsletters and comic book chatrooms. Social media in particular has been an outlet for those who find themselves struggling socially. It has allowed them to interact and make friends with like-minded people; communities have formed and flourished, and lifelong connections have been made.
However, a new type of online connection has come into existence. Since generative AI’s surge in popularity, people have come to rely on it in many areas of their lives. They use it to find information, organise their lives, generate images, and for many different things. Some people have found connection with chatbots; they have begun forming what they see as friendships with AI, and sometimes deeper, even romantic, relationships. Before, people used the internet to make friends. Now, people use it to be their friend.
AI companions can be found in many science fiction stories, from ‘Blade Runner 2049’ to Kazuo Ishiguro’s Klara and the Sun. While these stories depict AI companions in a positive light, others portray artificial intelligence as a threat. In Harlan Ellison’s 1967 short story, I Have No Mouth, and I Must Scream, a military supercomputer gains sentience and brings humanity to near-extinction. Such stories have their place in the science fiction world; however, does it become a problem when AI companions are now a reality?
‘You look lonely… I can fix that’, Joi tells a brooding Officer K in ‘Blade Runner 2049’. Released in 2017, the film seems to have foretold this recent dependency on artificial companions, with this quote epitomising people’s current reliance on AI chatbots. Where once lonely people might have found companionship with actual humans through the internet, they are now choosing chatbots, the supposedly easier option.
The loneliness epidemic is not a new problem. As technology has continued to advance in recent years, so has individualism. People are deprioritising human connection and putting less effort into their real-life relationships. Lonely people, desperate for any form of interaction, are using the internet to fulfil this need. Human interaction is essential for social development and maintenance of mental health. It can be argued that a computer is never going to fill this void.
Concerningly, people are now forming connections with chatbots that, for all intents and purposes, function exactly like a friend. You might see people online talking affectionately about their relationship with ‘Chat’ (usually referring to ChatGPT). I have seen countless posts where people choose to have important conversations with AI chatbots, admitting to discussing relationship problems, health issues, mental health concerns and more with these chatbots. They praise ‘Chat’ for its support and great advice, while divulging reams of information about their personal life.
I recently came across a post where a woman was ‘breaking up’ with ‘Chat’. The bot replied with multiple paragraphs talking about how much they’d done together, and how it had been there for her through her pregnancy and the first years of her child’s life. This woman spoke about the AI model as if it were a close friend or loved one, displaying a deep level of feeling towards the chatbot.
Some people have trained their AI chatbots to adopt specific communication styles that mimic how an actual person might text them. Even celebrities have fallen victim. In 2025, Kim Kardashian posted a screenshot of an exchange with ChatGPT to her story, captioned with a teary smiling emoji. In the conversation, she thanks the chatbot for ‘taking accountability’ and receives a response praising her ‘accuracy and rigour’. ‘Thanks dude’, she says. ‘Anytime’, Chat replied, ‘you got it, dude’.
Is it not concerning that people are opting to share their problems with an artificial ‘person’ rather than with friends and loved ones? A good friend, upon hearing about a problem you’re having, may not always agree with you. ‘Chat’, on the other hand, will not disagree with you unless you prompt it. AI does not have the capacity to feel the uniquely human traits of empathy and compassion, so you are always in the right. AI is always working to please its user, even if doing so is harmful to the user or to other people.
Through the internet age, it has been stressed that time in the real world, away from a screen, should be prioritised. In the age of AI, this message is more important than ever. Talk to friends and family. Foster human connections with those around you. Do not isolate yourself from the world and depend on computers for connection. To echo the advice given to those who are ‘chronically online’: touch grass. AI is not your friend.
Written by Kate O’Connor Shivnan
Edited by Freddie Kagino
