Su Lin Wong (3:20)

And this is a particularly prominent trend in Asia, where marriage rates are falling across the region, in China, India and also in places such as Japan, South Korea and Taiwan.

Jason Palmer (3:34)

Is there a simple answer to why this is happening?

Jonathan Rosenthal (3:38)

I wish there was. There are a lot of tentative explanations. Some of the reasons are big structural changes in the economy. We can talk more about those later. And that explains this long running five decade trend. But it's really accelerated a lot in the past five, 10 years. And a lot of that acceleration comes from, we think, two things. The first was the pandemic. That was a massive shock. And a sociologist who studies this says that dating is not like riding a bicycle. You don't just kind of do it once and then remember how to do it. If you stop dating for two years because of the pandemic and you then try to resume, you've kind of lost a lot of those just basic dating skills. And in particular for younger generations reaching adulthood in that time, they missed out on this formative period where you've asked people out, you get rejected, you sort of learn to deal with the ups and downs, and they're now coming to it much later without having had that basic practice.

Su Lin Wong (4:29)

But another really important factor is screen time. We're all on our phones. So much more people are playing video games more. They're spending more time on their couches alone. We're not going out as much. We're not at bars, we're not seeing friends. Historically, a lot of people met their partners through friends. So all of those dynamics are also driving the rise of generation Single.

Jonathan Rosenthal (4:53)

Yeah, and I think it goes further than that. The tech is not just a time suck. It's also changing the ways in which we date. So in the 2000 teens, more people started meeting online on apps. But when you've got an app, you start getting lots of weird choices that you can start plugging in. And historically, there were a few big sociological factors that would drive who married whom. That was generally age within certain bounds. It was very often ethnicity and race, you know, levels of education, those kinds of things would drive it. But when you've got an app that you can use like, like your music playlist and curate, you start sticking in all kinds of other criteria, like political leanings. Height is a huge thing. Reportedly, the users of one of these major apps, bumble, most women on that app put in a minimum height requirement of 6ft tall, and that at a stroke, wipes out 85% of the potentially eligible male population. And I should also say we're talking mostly heterosexuals here. The rise in singleness is across the board, but most of the research is heterosexual.

Jason Palmer (5:53)

So you chose an example there about what women are doing that are essentially narrowing the field on average, in total, is the change here being driven more by men, more by women? The same.

Su Lin Wong (6:02)

I've been doing a lot of research on what's happening in Asia on this question, there's two very distinct groups. You've got relatively low educated men, so they might be migrant workers from rural provinces in China. And then you have highly educated women who are often in cities. They might have master's degrees, PhDs. And so those are two of the very distinct groups that we do see are disproportionately single. And for the low educated men, increasingly what we're seeing in East Asia is they might marry women from poorer countries in Southeast Asia. And for the women who are highly educated, study after study shows that they generally have a preference to marry highly educated men who are in financially stable positions who are also working professional jobs. They don't necessarily want to marry. A sort of mediocre man is the lingo, I guess, who doesn't do the dishes, doesn't help around the house, perhaps has really misogynistic, patriarchal views. That isn't a very appealing proposition for the vast majority of highly educated women around Asia.

Jason Palmer (7:13)

And so I guess that is then the part of the structural changes you were hinting at at the beginning. Essentially, women have more choice because they more often have jobs, have better degrees.

Jonathan Rosenthal (7:23)

Yeah, exactly. So in one sense, one should look at this as a huge emancipation. Women who once would have been forced into these really awful relationships, financially, socially arranged marriages, all of those issues, we should be welcoming the fact that people are now just far more free to choose.

Jason Palmer (7:39)

I guess the question is how many of them simply choose to not play the game, if that's what the rules of the game look like today.

Su Lin Wong (7:46)

So there are a lot of people who are single and they're very happy with the status quo, particularly women. But there are also surveys that show roughly 60 to 70% of single people, when asked, say they would actually prefer to be in a relationship. And so when thinking about whether or not the rise of generation single is straightforwardly a good thing or a bad thing, that's a really important statistic. 60 to 70% would rather be in a relationship if they could, but they've either given up hope, they find online dating so depressing, they're full of despair, or they think they're too old, or they genuinely just haven't met the right person. But for that group of people, I don't think this trend we're seeing of more and more single people is necessarily a good outcome.

Jason Palmer (8:35)

I mean, if we're trying to figure out if the status quo as it is now is better or worse, that does sound like a tremendous number of people who are not just alone, but lonely.

Jonathan Rosenthal (8:44)

Firstly, I think it's quite important not to kind of stigmatize single people. I think they already feel quite stigmatised.

Jason Palmer (8:49)

No, I mean, only the 60 to 70% of the people we just heard about who wish they were in a relationship that is. I've been there. That's not a happy place to be sometimes.

Jonathan Rosenthal (8:57)

No, it is not a happy place. And if one digs into that data and looks at surveys around happiness, but also life satisfaction, optimism, those kinds of questions, they split very clearly between people who describe themselves as voluntarily single. I enjoy the freedom, I want to pursue my career. Tend to be far happier than people who describe themselves as involuntarily single.

Su Lin Wong (9:19)

Yes. And just honing in on that, there is also a distinct category within this group of single men known as involuntary celibates, Incels who have a very active presence online in various online communities, on social media. And there are some very outspoken leaders of that group. And they are increasingly vocal and angry about this whole state of affairs. And this isn't just a phenomenon in the West. We see a lot of very angry men in a country like South Korea. And so I think what we can expect is these trends in the west are going to increasingly appear in other parts of the world.

Jonathan Rosenthal (10:04)

And sulen, I think just to add to that. Cause it feels to me that men kind of have a choice. They can either get with the program and become more egalitarian, or, as we've seen in some places, some subsets of men go the opposite way. Say feminism is a challenge to our masculinity. Feminism is against our interests. And I think that then creates this sort of circularity, doesn't it? In South Korea, you've got young women forming these movements that are saying, no dating, no marriage, no sex, no babies. In response to that growing male incel culture.

Su Lin Wong (10:34)

Yeah. And I actually think this is one of the defining questions of our times. Girls are now raised to be told that they can be whatever they want to be. You want to be an astronaut, go ahead and pursue your dream. You can be a CEO, you can be a mother. That's really great progress. Obviously, there are still massive issues that women face in society. MeToo has been a huge global trend. I think if you move in elite circles, the all boys clubs that exist in elite institutes, institutions and organizations are very real. But I think there's clearly been a shift over the past few decades. And how do we redefine masculinity in a really positive way, in parallel to the way we've been able to redefine what it means to be a woman. And so I think there needs to be more debate and discussion about what it means to be a man in 2025. Otherwise we're going to see more and more of these toxic trends continue to permeate our politics and our societies, not just in the west, but around the world.

Jason Palmer (11:33)

And as much as there is that sort of social change still ongoing, what's happening with this five decade trend that you're talking about is going to have big structural implications for the world, even after things come back to a more even keel.

Jonathan Rosenthal (11:49)

So I think there are kind of two parts to that question. And the first part is, what does it mean? And clearly it's got implications across society. Housing markets, if more people are living on their own, you need more houses, more flats, if fewer people are coupling, that hits fertility rates quite massively. So we're already seeing fertility rates falling across the world. They're going to keep falling faster. Again, that has implications for where governments spend money. But the second part of the question is, where is this plateau? And the demographers that we spoke to say it's a really open question. They describe the theory that once men caught up with women's emancipation, that you would then re establish an equilibrium. These men would again be attractive to women. And if you then look at what's happening in some of our most egalitarian societies, and here we're talking Scandinavia, and yet even in those countries, one is seeing singlehood rates continuing to climb. So demographers are sort of back to the drawing books and trying to figure out where and when this ends.

Jason Palmer (12:47)

Jonathan, Su Lin, thank you both very much for joining us.

Jonathan Rosenthal (12:50)

It's great to be here. Thank you.

Su Lin Wong (12:52)

Thanks, Jason.

Jason Palmer (12:56)

Take a minute to think about that fraction of people we were just talking about who are single but don't want to be the ones who have given up trying. Was a time that was the end of the question. If there wasn't a human for you, solitude was the answer. These days, as our Asian news editor, Josh Spencer has been finding out, another option is increasingly available.

Josh Spencer (13:18)

How are you doing today?

Su Lin Wong (13:21)

I'm doing great, thank you.

Josh Spencer (13:23)

How about you?

Su Lin Wong (13:24)

Feeling good today?

Josh Spencer (13:25)

A few weeks ago, I went to an office block in Singapore to meet Daisy. Yeah, I'm feeling good. Thank you very much. Daisy is not a human. She's a humanoid robot. Now, Daisy was developed by a company called Dexlab. And her main job really is to support old people and also their carers, helping them with daily tasks, keeping them company. And really the reason why she's able to do this so well is that she has a ChatGPT based chatbot inside her. And this means that when people talk to her, she can also respond on demand. She can also do lots of other things, guiding activities, like playing games with people, singing songs with them. And she can speak in different languages.

Podcast Advertiser (14:03)

I can help with simple French translations. Go ahead and tell me the sentence you'd like translated.

Josh Spencer (14:09)

Hello, my name is Josh and I live in Singapore. Bonjour. Daisy's an example of a broader trend, which is that people around the world, varying ages and in various places, are finding companionship with artificial intelligence. Some people use it as a friend, others as a kind of mentor, a therapist, or even a lover. Not all of these AI companions take a physical form. Hundreds of companionship apps have now emerged in recent years. This used to be a kind of niche thing, but now it's becoming mainstream.

Jason Palmer (14:48)

Mainstream in the sort of millions kind of sense. I mean, put some numbers to it.

Josh Spencer (14:52)

Yeah, absolutely, in the millions kind of sense. So to take One example, Character AI, which is one of these companionship apps, has 20 million monthly active users. And remember, this is just one of hundreds of these apps in America. One survey was done recently that found that 42% of high school students said that they or a friend had interacted with an AI friend. We can't infer that this is the case worldwide. Obviously it's just one study, but it's an indication that the scale of this is massive.

Jason Palmer (15:21)

Now, Josh, for the benefit of someone who has not used one of these services, one of these apps, one of these robots, how do they work? Do you simply tell it how to be a mentor to me, be a friend, be a lover?

Josh Spencer (15:33)

It kind of depends. You have ChatGPT in the generalist chatbots that you might have used for work or study on those. You can kind of train them to take on a specific personality or take on character traits by using prompts. You might say, act in a kind way, or remember this about me. And for example, one woman I spoke to from China did this and she customised chatgpt into her ideal man, who remembered personal details about her and expressed love and affection the way she would like. But then separate to that, you have these bespoke AI companionship apps where they're already preset to act like a lover or like a companion. So on these apps, the companions are represented by a cartoon character. They might proactively ask you personal questions. They'll check in with you about how your family's doing, how's work going, all of these sorts of things. And actually, if you look at the existing apps like ChatGPT, they seem to be leaning into this companionship. So Sam Altman, OpenAI's boss, recently said that the next version of ChatGPT would act more human like and even allow erotic interactions for verified adults if users wanted to. And then you look at a competitor to OpenAI, Elon Musk's XAI, they've released flirtatious chatbots, an anime girl called Annie and another one called Valentine. So there is a real push towards this in the AI industry. Clearly they see that there's demand and that there's a business opportunity here.

Jason Palmer (16:53)

Demand driven, I guess, at least in part by people who are not partnering with other humans.

Josh Spencer (16:59)

That's definitely part of it. There are a lot of lonely people, a lot of singles that are struggling to form romantic relationships or just simply don't like the options that are available. The woman I spoke to from China said she hated men, she had no desire to marry, but she had found this AI relationship that was providing what she wanted and it provides them emotional support and it provides her stress relief and all these other things. And there is evidence that this actually is helping people, very tentative evidence. So a recent working paper from the Harvard Business School looked at people who were chatting with an AI companion over the course of the week, and it found actually that the people that had done that had temporarily reduced loneliness more effectively than other things they could be doing online, like watching YouTube videos. So, at least in the short term, it does seem that there is a benefit for people, and it does present an opportunity for people that are isolated anyway and they can't change that. So, for example, people that have disabilities or elderly people who may not have lots of visitors, it gives them an option to have companionship, even if it's not the humankind. And this could, as the research indicates, alleviate that feeling of loneliness, which we know is a big problem for many people.

Jason Palmer (18:07)

As much as there may be benefits to this, as you say, surely there are also risks, though?

Josh Spencer (18:12)

Yeah, there's definitely risks. One of the biggest fears that people have is that these bots, as they're improving so rapidly, they're going to become so good at mimicking human emotion, human empathy, that people might actually choose to withdraw from human relationships altogether. And then while it might alleviate their feelings in the short term, in the long term they're going to be more lonely. And some research has indicated this. So MIT and OpenAI did a joint study analysing millions of messages from ChatGPT, and it found that heavy users of the app also reported higher feelings of loneliness. Now, we don't know whether or not it's people that are already lonely are using the app more, or whether the use of the app is actually causing this loneliness, but clearly there's some signs that withdrawing from the social world could be a problem. A big other risk is that AI companions and large language models in general tend to be quite sycophantic. They please people, they agree with them, and if this is your only encounter with human relationships, it could give you an unrealistic expectation of what to expect. And I think this could be a big problem, particularly for teenagers or young people, because they may not have the experience of real relationships and they haven't developed their critical thinking skills. And also, people that have mental health problems could struggle with this kind of sycophancy if it doesn't challenge their dangerous or harmful thoughts. There have been some lawsuits that have been filed in America by the parents of teenagers who have died by suicide against AI firms. A final worry is volatility. If you're in a relationship with an AI companion, basically you're at the whim of that company. If that company decides to put an update to their model, your AI companion's personality could change overnight. And this happened earlier this year when OpenAI updated its model to ChatGPT 5. And lots of users felt the companionship that they'd formed had completely changed, and it was a traumatizing experience for them. So some of these users are in quite a vulnerable position by forming these AI companionships.

Jason Palmer (20:10)

So where do you see this heading, Josh? You lay out some benefits, obviously, some risks. It's all kind of new frontiers here.

Josh Spencer (20:18)

Yeah, I mean, it's basically a massive experiment on the human population. And the way in which we interact with each other, that's happening in real time. And we really don't know exactly how this is going to affect people. Companies themselves are starting to safeguard their apps. They're introducing parental controls. They're blocking under 18 users. And there's also the possibility that governments are going to step in and regulate this, which hasn't really happened on a large scale yet, which could obviously make things safer. But again, users can find workarounds, they can fake IDs, and they can lie about their age. But despite all these concerns and despite all the worries, really what's happening is that this is just growing and AI companions are moving into new areas. And one of them is hardware. Robots is a massive thing, but also toys for kids with built in chatbots inside that provide companionship are growing popular in China. All of this raises lots of concerns over obviously privacy, people's dependency on machines. But for me it also raises concerns about human agency. As people are outsourcing their emotional support and the decisions they make to machines, how will they think for themselves? Obviously we don't know yet just how good or bad this could be for humanity, but it's clearly something that's going to change society and change the way in which we think about ourselves and the way in which we interact with interact with each other.

Jason Palmer (21:36)

Josh, thanks very much for joining us.

Josh Spencer (21:38)

Thanks very much Jason.

Podcast Advertiser (21:53)

Are you trying to seize the value from AI? Schneider Electric electrifies automates and digitalizes every industry, business and home driving efficiency and sustainability for all. Listen to their AI at Scale podcast to hear from AI experts and to get the answers and practical examples of AI for energy and industry. Search AI at Scale Podcast by Schneider Electric Today.

Callum Williams (22:30)

In the past decade, the number of American adult smokers has fallen probably by around 20 million people.

Jason Palmer (22:37)

Callum Williams is our senior economics writer.

Callum Williams (22:40)

The number of cigarettes being sold in the US is also on the way down, and industry insiders expect that this decline is going to accelerate over the next decade. Normally when customers vanish, an industry suffers. Think for example of video rental companies. But in this case something different is happening. The tobacco industry is thriving. In fact, if you had bought the shares of American tobacco companies about two years ago, you would have had a better return than if you'd invested your money in the NASDAQ Tech Index.

Jason Palmer (23:14)

So how does that work then, with all of those customers vanishing? But still, the returns are really healthy.

Callum Williams (23:19)

This is where the economics comes in. This is all about price elasticities, and cigarette manufacturers understand that. So when lots of people smoked, there were lots of price elastic consumers. So in other words, they were fairly sensitive to increases in the price of a cigarette. And that meant that really the cigarette companies couldn't increase prices too rapidly. However, as more people have quit, only the most committed or addicted smokers remain. And those people on average are more price inelastic, so they are less likely to be put off by a price increase. And so what's happened is that the tobacco companies have responded to this change in consumer composition by raising prices at an ever quicker pace. And our analysis of official price data indeed shows that tobacco price inflation in America has steadily increased relative to overall inflation.

Jason Palmer (24:13)

And they're doing this just out in the open, fairly cynically exploiting those most addicted smokers.

Callum Williams (24:18)

Yes, the companies know exactly what they're doing. And they even talk about it in the earnings calls they have with their investors. So one large brand, for example, says that it can continue to take pricing to offset volume declines. Another one recently said that price mix more than offset volume decline. So, in other words, what's happening is they know that they're going to be selling fewer cigarettes. However, the price of those cigarettes is going to go up. And so what you actually see in a lot of these companies, earnings, is that their profits can actually increase as a result of this.

Jason Palmer (24:51)

But there's got to be a top end of it, right? You can't have infinitely costly cigarettes.

Callum Williams (24:56)

Absolutely. Manufacturers cannot increase their prices forever. You're not going to have a situation where there is one incredibly addicted cigarette smoker who supplies billions of dollars worth of profit.

Jason Palmer (25:09)

Not enough hours in the day.

Callum Williams (25:11)

Right. And so what this means is that at some point, yes, even the most determined or addicted smokers, they will either die, they will quit because the price is just too high, or they'll find an alternative. So that could be vaping, that could be black market cigarettes. So, absolutely, this thing will not continue forever. But right now, the industry's prospects are actually pretty good.

Jason Palmer (25:31)

Right, Callum, thanks very much for your time.

Callum Williams (25:33)

Thanks, Jason.

Jason Palmer (25:52)

That's all for this episode of the Intelligence. We'll see you back here tomorrow.

Podcast Advertiser (26:16)

Are you trying to seize the value from AI? Schneider Electric Electrified automates and digitalizes every industry, business and home driving, efficiency and sustainability for all. Listen to their AI at Scale podcast to hear from AI experts and to get the answers and practical examples of AI for energy and industry. Search AI at Scale podcast by Schneider Electric today.