Chronic loneliness has no medical imaging to display on the screen. There is no blood work drawn to test for it. There is no clinical protocol for routine screening for loneliness as there is for elevated blood pressure or pre-diabetes. Though we readily discuss diabetes, heart disease, and obesity in clinical and lay settings, loneliness is viewed as a third-rail topic. It shouldn’t be.
Chronic loneliness has been associated with inflammation, stress, decreased cognitive function, and even elevated risk of death. These are no longer considered to be soft science. So why do we treat loneliness as being less than other health issues, if it impacts our bodies in so many ways? Can AI help mitigate this problem, without making preposterous claims about replacing human companionship? If so, what are its limits?
Why Modern Life Is Making Isolation Worse
Remote work has meant fewer watercooler conversations. Social media gives us the headlines of our friends’ lives, not the details. We live in increasingly dense cities, but community is harder to find. Even the practice of medicine has become episodic as we pop in and out of the doctor’s office. With a few keystrokes and mouse clicks, many of life’s formerly mundane tasks are made more efficient.
This is a good thing. But in making many of life’s little tasks more efficient, we have removed many of the friction points that once brought us together as humans. Standing in line at the drugstore was a pain, but it was also a chance to chat with others. What we have gained in efficiency, we have lost in a sense of belonging. We have lost many of the everyday moments that remind us we are not alone.
This is a slow-moving process, so we don’t notice it in the same way we might notice a stock market crash or a natural disaster. But it’s a phenomenon that’s very real nonetheless.
Enter AI: From Chatbots to Digital Companions
The initial applications of AI, when the technology was in its relative youth, were straightforward, such as chatbots that answered frequently asked questions or virtual assistants that reminded us when to take our medication. These were useful, if somewhat clumsy tools. But over time, we started to see AI used in different ways. It could engage in longer conversations.
It could remember things about our habits, and remind us about tasks. It started to seem less tool-like and more companion-like. To be clear, I am not saying this is the same as human companionship. I am not suggesting we trick ourselves into believing that having AI companions will be the same as having other human beings in our lives. But if a chatbot could have a daily conversation with an elderly woman who can’t leave her home, and it can somehow reduce her feelings of loneliness even just a tiny bit, that’s good.
If it reminds her that she’s not alone, that’s good. If AI can remind someone to take a walk, or connect someone with a friend, that can help too. I don’t think it’s a technological fix for a very human problem, but I think it can be part of the solution to a problem that will only get worse as our lives become increasingly digitized.
Emotional Attachment to AI: Risk or Relief?
People get nervous about this question, but what if someone gets emotionally attached to an AI? Is that something we should worry about or is that fine? I think if someone is having a bad night and an AI companion makes them feel better and reduces some of their anxiety and makes them feel like someone listened to them, I wouldn’t knock it. A relief is a relief.
The problem is if that’s their only relief. If it’s the only thing they do, and if it supplants other social interaction, then I do think there is a problem there. I don’t think the attachment itself is a problem. I think the context of the attachment is a problem. I think this can be a tool that people can use. I just think there are times when it is used in the wrong way, and that’s when it becomes a risk.
Human + AI: Designing Augmented Social Support
I think the potential of this stuff is not that we are going to have this choice between human and AI, but what can AI do to facilitate human interaction? I think that’s the real power. For example, the AI can be used to do some of those daily check-ins, some of those daily reminders, some of those daily “Hey, how are you doing?” types of things, while the humans can then be freed up to do some of the more complex social work. I think that is where we will end up.
Technology can scale empathy, but it can’t scale lived experience. An AI can recognize that someone is pulling away from social interaction and flag a human to check in with that person. An AI can tell someone to call a friend. An AI can lower the barrier to call a friend. I don’t think that’s necessarily a bad thing. I think that’s actually a good thing. I think we can design these systems in a way that are really helpful.
The Future of Relational AI in Public Health
I think that the way in which relational AI is going to be impactful from a public health perspective is if we can get it to the point where it’s not a novelty and becomes more of an infrastructure-type thing. I think that’s really where the jump is. I think that we will have these systems that will be able to monitor community-wide trends of people and be able to see when someone is pulling away and say, “Maybe this person should get in touch with this other person.” Or, “Maybe we should alert a caregiver.”
I think that is where we are headed, but I don’t think we are there yet. I think that sounds a little bit like science fiction, but when we think about public health, we think about patterns. The difficulty is doing that in a way that is not clinical. I think that is where we are going to have to go with it.
I don’t think it is going to replace the friendship or the family member or anything like that. I think it is just going to be this thing that sits in the background that can help people. That is where I think we are going to end up with it. I think if we do it right, it will be one of those preventative measures that I think will be really, really valuable to public health. It is not going to cure disease. It is going to get at some of those social determinants that end up causing disease.
Conclusion: Technology Cannot Replace Humanity – But It Can Support It
You can’t replicate the human connection. That is not something you can program into an algorithm. I think that is what makes it so valuable. You are not going to be able to replicate a friend that you have known for 20 years without showing them a screen. But I do think that there are things that technology can do to help.
If technology can help reduce the isolation at the margins, if it can encourage someone to reach out to a friend, if it can be a presence in their life when they feel like there is no presence in their life, I don’t think that is dystopian. I think that is just compassionate.
I don’t think we are trying to supplant human connection. I think what we are trying to do is use technology to augment the places where human connection is not as strong as it could be.
