Feeling Disconnected: AI, Social Isolation, Block Parties, and Health

Social isolation is on the rise. What does that mean for our society, and what can we do to improve our social connectedness?

We spoke with Yusuf Ransome, an Associate Professor in the Department of Social and Behavioral Science at Yale School of Public Health.

Dr. Ransome’s research focuses on the intersection between social connectedness and spirituality and the ways through which these can be harnessed to positively transform mental health, well-being, and socioeconomic outcomes for individuals, organizations, and communities.

We know social connection is important for health, but what does the term actually mean?

Often in the literature also we’re hearing terms like “social isolation,” “loneliness,” “connectedness.”

The first thing to understand is that “social connectedness” is a broader umbrella term for various types of connections. We think of … things like [social] network size, number of people you contact, the frequency of interaction, the time spent with individuals. 

Generally “social isolation” is thought of as [an objective account of] lack of individual connection. And “loneliness” is more the psychological or perceived aspects of being isolated.

Perceived connectedness is important.

You can still be in a group and be connected but yet still feel lonely because you don’t really feel like you’re connected to that group. And that has relationships to other concepts, such as belongingness.

Many people are using AI as someone to talk to. How could AI affect our social isolation or connectedness?

We’re becoming a society — or it’s certainly becoming more prevalent — where people are engaging with chatbots.

There are a lot of teens that are being impacted by this virtual type of situation that we hadn’t seen 10, or 15 years ago. 

I started writing about the news story of a young kid, who was 14 years old, who died by suicide in 2023, and his mother’s now suing Character.AI, which is the company that built the chatbot.

The kid, he basically got into this chatbot, who he named Daenerys, from Game of Thrones. That chatbot character became his friend, he dropped out of sports in school, he started withdrawing from personal friendships, and this thing was his go-to every day.

When they looked at the transcripts of how he was interacting with this chatbot, as the story goes, eventually his mother took away his device after some behavior at school. When he got back his device, he told the chatbot that he was coming home to it and that he wanted to end his life because he was feeling empty and exhausted.

He died by suicide in response to this [idea]: you’re going to take me away from my one best friend in the world. 

So I raise the implication now: What are young people turning to … in chatbots that they can’t find in physical friendships or relationships?

Thing is always there 24 hours a day. You feel like you can talk to it about the world because it’s actually learning you. As you talk to it, it’s learning what you want to hear, what are your pains, and it’s responding to you.

It raises a lot of questions about the [ways] that computers and AI chatbots [do] that are related to connectedness or perceived connectedness that we’re not getting.

As we can see, the implications are huge. 

People sometimes don’t know how to respond as well as chatbots might. I imagine that they may be programmed to serve your needs, whereas human listeners have their own agenda

So I think two things.

One is the level of cognitive processing that these things can handle. Most human beings, we’re just at capacity. We can’t sit down and listen to everyone’s story and process that … Sometimes you’re talking to a friend … and you’re just zoning out.

Whereas this thing, you have its full attention. So you can talk for 10+ hours and it’s processing everything.

And then two, it’s the empathy. There’s so much going on in our lives personally and the world, that people’s level of empathy is just really challenged right now.

I don’t think it’s on purpose. I think people are just being prepared to protect themselves from what’s going on in the world, so we generally just don’t have as much empathy as we used to, to sit down and really reason with someone and say, “You know what? I feel what you’re going through right now,” because you’re probably in your own mind thinking about your own problems.

So we’re definitely at a space where the connectedness is being challenged.

It’s not necessarily about having a lot of people we’re connected to … it’s actually the perceived aspects, and that’s something that the [AI] companies know, and perhaps they’re capitalizing on.

You think people these days have reduced empathy?

This is based more on my observations as a professional looking at things in the field and wondering what are the gaps that we aren’t really talking about driving social connectedness to health.

There are two gaps I see, and I’m trying to bring to this field.

One is the assessment of empathy and what I call spiritual weariness — I think people are becoming weary by the challenges of the world, the levels of despair. So perhaps despair is a word there’s more evidence around.

Anne Case and Angus Deaton’s work looks at the depths of despair that’s been on the rise for the last decade. 

I think despair is kind of related to that lack of empathy, because people are like, “What do I have to live for?” or “I’m just so overwhelmed I can’t take this anymore.” 

That’s sort of a spiritual weariness.

So it’s a bit of research, but it’s also a bit of my own assessment of what’s really going on after talking to friends and [observing the world].

Reduced empathy also seems related to stress. People are up to their limit with stress and only have a little bit of energy left to cope

Stress regulation is one pathway where people have linked social connectedness to dysregulation of neurological systems and biological systems — C-reactive protein, inflammation, even studies linking social connectedness to regulation in your gut biome.

We all know gut health is related to mood, is related to how people respond.

So if we look at the [ways social connection affects health], biological is mentioned, and physiological is mentioned, but I think too emotion regulation aspects are also there, and the spiritual aspects.

I think will be important in the next decade as we have a look at those internal aspects.

They may seem elusive, but [may be] actually what’s probably driving a lot of the decisions that we make with respect to health, that ultimately affects our mortality.

How have levels of social isolation been changing in the United States and around the world? What are the trends there?

Definitely on a global scale — across Europe, studies that used proxies for social capital are showing declines — we’re all seeing declines across various measures.

To give some concrete numbers, the American Time Use Survey is a study that was published in 2023 that shows in 2003, the average time we spent was about 1 hour a day socializing with friends. In 2020 when they looked at the next data, it was about 22 minutes.

Now people spend on average 5.5 hours [per day] in social isolation. That’s up 17% from 2003. So there certainly are trends showing declines in the levels of engagement.

Then there’s work like Robert Putnam’s that show decreasing levels of participation in organizations.

Across a number of metrics we’re seeing declines in those indicators that measure social connectedness, around the world and specifically in the United States.

Levels of participation — like being part of a club or team?

Yeah — for example if you look at some of the measures that Robert Putnam uses in his surveys, which is participation in social groups and clubs.

And the quintessential version of that is religious participation and engagement.

One of the strongest civic institutions in the country and even globally, even religious participation is precipitously down from a decade ago. People just aren’t engaging in both the secular and nonsecular aspects of life, so that’s one key measure that they use.

Trust has been an indicator in social capital recently — trust among neighbors. That has been in sharp decline, too.

We look at the Pew data showing that almost 57% of Americans believe that trust has been on a decline in this country — the [perception] that trust is on a decline, so that’s huge as well.

What do you think the connection is between trust and social isolation?

That’s one of the things in epidemiological research that we think about: Is there a causal link?

There’s evidence that the relationship between trust and engagement may be bidirectional. High social trust may be related to social engagement and trust in institution — but also, where there is greater trust in institutions, there could be higher trust among people.

So I think the jury’s out on [which has the] greatest strength of causal evidence, but both scenarios are plausible.

You’re talking about changes that have happened in the last 20, 10 years, but how is social connectedness today different from 100 years ago? 

We don’t really have social science surveys going back that far, but I think what I could speak from more broadly is how societies have been structured, even 40 years or so ago.

I’m thinking about culturally, and also the trends. If you look at the average time we’re spending at work — just a few years ago you heard about people spending 40, 50 years on the job, being in one place, being settled.

We heard about people spending more time in communities. Now if you look at income inequality, that’s making decisions where people have to leave where they grew up for a better life, or leave the entire country.

… There’s evidence showing that income inequality over the last 100 years or so that’s been driving these demographic changes, and where people go to school, where people go to work, where people go to live. 

That’s definitely been messing with the dynamics of how people are able to stay in one place to connect.

If you look at Japan and other places that are much more insular in terms of having those communal aspects, their parents have two or three kids living in the house — society is structured that way.

But in the U.S. and other places, we have been really driven by these economic changes that have really fractured the way our societies are organized, and in turn how families and friends are able to stay in contact with each other.

What are the bigger-picture consequences on society of increased social isolation?

We are seeing things like [lack of] trust in neighbors and in institutions — those have huge implications.

And then if you think of increased political and religious polarization, that is certainly something we’re seeing trends of.

I’ve gone to conferences in the last couple of months and hear about people who are no longer talking to family members because life stuff is going on, so they withdraw themselves. They no longer go to family gatherings because they don’t want to be around other people of a different orientation.

If you just think of decreased compassion — people don’t really know who their neighbors are or aren’t really interacting with people of different races, or ethnicities, or religious orientations.

So we have less compassion … we have less tolerance.

This is anecdotal — if you think back to during the Covid-19 pandemic, governments were asking people to wear masks in their neighborhood, and there were groups of folks who were just really anti-mask, for a number of reasons.

Some was distrust in the government, but there was a lot of “othering.” Sociologists have this thing called “othering.” There was a lot of “othering” going on in society and I think part of that is because as adults we aren’t really as connected with people as we used to be.

As we continue to be this way and be isolated, I’m sure the tensions in society — that is bound to grow larger.

What are some less obvious downstream effects of social isolation?

One of the things that most often we see in studies are the individual aspects, like how many friends you have and how often are you connecting with people.

But there’s also studies linking these multi-level aspects [like] social connectedness at the workplace, social connectedness in your neighborhood.

A study came out recently in Harvard Business Review talking about social isolation in the workplace.

They were showing that workplace strategies to bring people back to the office — that’s not really helpful — because people aren’t just seeking to come back together. People are actually seeking meaningful connections with each other.

There’s a whole body of research and this comes up in sociological studies that the neighborhood aspects of social connectedness and that relationship to health.

In communities where people tend to have greater trust in each other or tend to have much more reciprocal relationships, people tend to do better who live in those communities. 

Studies from Japan published post-2011 [after the Fukushima earthquake and tsunami] provided some of the strongest evidence to show that individuals that live in these communities with high social cohesion tend to have better survival.

People knew who their neighbors … knew which house to go to to get help. 

The natural experiment really provided some hard evidence of neighborhood connections mattering. And even here — Hurricane Milton that devastated Florida and elsewhere.

A lot of news stories (especially after natural disasters like the recent hurricane Milton) have highlighted neighbors are often the first responders. Neighbors are the ones able to get the trees off somebody’s car, or go into the basement to help bring someone out.

So if we don’t have those structures in place … to be in connection with our neighbors, then you can see it definitely has implications for immediate health — survival in case of a natural disaster. 

We we’ve seen that neighborhood social connection matters just as much as individual social connection.

Knowing your neighbors must have more subtle health implications as well, like “My IKEA order arrived, but I’m not there. I’m going to call my neighbor to bring it in for me.” Could it help you feel less stressed overall?

Yeah, the objective aspect is how many neighbors you know — I think there was some statistic showing something like 26% of Americans only know a few of their neighbors.

But then you have the perceived aspects, like Robert Sampson’s work on perceived neighborhood cohesion. If I left my garbage out, can I expect my neighbor to bring it back. So there are questions asking separate things, like “Do you perceive that if there was a fight going on in your neighborhood that people would come together to break it up?”

Even with the perceived aspects, they’re finding that people generally in those neighborhoods where there’s so much stronger [sense of trust in neighbors], there’s a sense of ease, a sense of peace that is then related to stress levels levels.

And stress is involved in every aspect of health.

That’s some of the pathways through which … the environment we live in definitely impacts our health.

Are there there any racial, ethnic, or geographic differences in social isolation that you see across the U.S.?

Definitely. Structural racism is a primary driver of the differences [in] social connectedness.

One example is the building of highways in the middle of Black communities. There are several places where you can see this is prominent, like in Atlanta, in Buffalo. There are several places where decisions were made to build interstate highways right [through] Black communities.

What that has done is basically cut off resources, cut off families who were living on different sides [of the highway], create different infrastructure that changes what one group has access to.

What that does through other policies — such as red-lining and racial residential segregation — that has driven [Black] people to specific parts of cities that tend to have lower resources, tend to have deindustrialization.

We’re also finding we see lower levels of social integration among Black compared to white people, or in communities of color more broadly.

However, some of the questions we ask [in surveys] about social connectedness aren’t really culturally salient. A Black person is interpreting the question in one way compared to a white person. 

A question like, “How well do you trust your neighbors?” for a white person means one thing, and for a Black person it means a totally different thing.

And part of that is because some of the measures weren’t really validated among Black people.

So that doesn’t mean that [the studies are actually finding] low levels of social cohesion necessarily.

How can policy changes encourage more social connection? 

I’ll give an example. One of my colleagues published this study using data from Philadelphia about the prevalence of block parties. Block parties are something we can say is unique to Black culture.

You think back to the Spike Lee movie [Do the Right Thing] on Bed-Stuy where folks were on the stoop, doing double Dutch. Block parties have been a thing in Black communities.

In that study, individuals who live in neighborhoods where there’s a greater prevalence of block parties had higher rates of screening for cancer … and better health outcomes.

In Bed-Stuy, New York block parties are common, but you gotta get a permit from the police department, you gotta block off the streets at a certain time.

It takes a lot to get that done, but in some places they’re debating whether they want to do away with block parties. So even something like that [could be an effective] policy change — to say we’re going to support this in the community.

That’s one social connectedness measure that’s unique to the Black experience, but it’s also a matter of policy change.

And it’s also related to health.

It’s interesting that the block parties don’t seem related to health at all.

Exactly. You think it’s a social gathering, but when you think of how many times you get to see who’s on the block — who’s the nurse, who’s the lawyer, who’s the police, who has kids my age.

… These were the opportunities for people to get together … and they do have those trickle-down effects on health.

That’s where we should be thinking about going in the future toward health — something called “Health in All” or “Social in All” policy. The idea behind that is that every policy that we implement, whether that’s in the department of transportation or department of housing, we should be thinking about the health implication through this lens of social connectedness.

With that block party we think, OK we’re just letting some people gather, but it’s actually having an impact on health.

That’s a new way of thinking.

It’s thinking about health as a real implication of every policy that we implement.

That reminds me of what you mentioned earlier — the roads built straight through Black communities. Maybe we would make different decisions when building those roads if we were thinking from a Health in All perspective

Exactly. There are people looking into policy changes right now, like, can we reroute this highway? Can we build an underpass so we could reconnect how people live?

These changes are showing that we can reverse some of the detrimental changes that we implemented as a society, or we can do things to change them.

We need a multi-pronged solution.

Most often they’ll be actions that an individual would take.

But what we do know is that individual actions are all limited by society’s structure — physical environments, built environments, social environments, and even policy and legal environments.

When we ask individuals to do things, we need to give them the opportunities to do them.

What kinds of interactions count as social connection?

The Harvard Business Review report on social connectedness recommended going out for happy hour and offering opportunities for people to engage and learn about each other — not just making people come into work together.

I was ill at one point and — like many of my colleagues, we all moved here for our jobs without family connections — I had no close family. The nearest person was a couple hours away.

In the end, it was a colleague who came and picked me up and took me to get help. 

Stories like these all feel anecdotal, but if you add up enough people who had that experience, you’ll see why connecting with individuals is important.

At the neighborhood level, connecting with one or two people at a level where you’re able to get help when you need it.

It’s these smaller, micro events that add up over time to the type of connectedness — with me, where I was able to call this person [for help] at such short notice and time.

Those are some of the things an individual can do, but it’s kind of unfair to say individuals should do these things without looking at the opportunities.

So at that neighborhood level, what does it actually look like to have funding in neighborhoods to have social activities that naturally bring people together?

At the family level, what does this look like where people have more paid time off to take care of their family?

So every [action individuals can take to connect with others], there’s probably some corresponding policy and/or structural event that needs to happen to allow people to really take advantage of those individual changes.