Host: If you're a young man struggling with your mental health, would you ask for help? Because a surprising number say they wouldn't. A new Canadian report found 1 in 3 young men see reaching out for mental health support as a sign of weakness. And that's despite years of campaigns telling men it's okay to talk. So what can we do to fight the stigma? And are some men more at risk than others? We dig into what the data says. Then could AI become a cure for loneliness? We talked to a psychology professor and author who says yes, but warns that it could come with a cost. This is The Rundown. It can be hard for young men to ask for help when they're struggling. How hard? Well, a new report from GreenShield and Mental Health Research Canada breaks it down. Khush Amaria is director of clinical services at GreenShield, and she joins me on the line. How are you doing, Khush?

Khush Amaria: I'm good, thank you.

Host: Let's talk about sort of broad strokes. How would you describe the state of young men's mental health here in Canada?

Khush Amaria: Well, GreenShield here, we've been doing some research recently to try and understand really what is going on for young men. And, you know, we've learned that young men still view seeking mental health support as a sign of weakness. About 1 in 3. We also know that about 4 in 10. So 44% of young men, they don't get the help they need when they identify needing mental health related issues. And we also know that when men, young men engage in care, they often don't stay engaged in care. So about half of them drop out before they can get sort of what I would call a full dose of the right supports.

Host: When we will delve into a little bit about what supports are there. But I am curious. We have seen among Gen Z more of an embrace of mental health and talking about mental health, but we've seen that sort of on TikTok. If you go on TikTok or Instagram, we've seen there's no shortage of trauma talk. And I'm curious, is this embrace just not playing out for young men on those platforms?

Khush Amaria: I think, you know, part of what we are trying to understand is where are young men going to find help, right? Where are they seeking support? Who are the trusted people they are speaking with about their own mental health? And we wanted to believe, with all the work that we've done across Canada, to destigmatize access to mental health services, that we've made a mark. But I think we've recognized that we've sort of left young men behind a bit, and we need to understand better how they want to seek their mental health support. Maybe it's on TikTok, maybe it's not. Maybe it's in other forums. And so that's really what we're trying to understand right now. The way our current system is set up. Do you think we are failing young boys and men? I don't want to say that we're failing young men and boys. I do believe that we are working. We're moving the needle forward by trying to understand where young men, where young men need their help and how they're going to access it. So Health Canada actually has the very first strategy in place, Canada's first men and boys health strategy. So that tells me we're recognizing that we actually do need a gender informed model. That doesn't mean prioritizing one gender over the other. It just means helping us understand how do young men want to access care, and where are our risk factors? Our racialized young men, newcomers to Canada? We can't have a one size fits all model. And so I think that's what's really different in the coming years, trying to build a model where we are prioritizing groups that may have been left behind.

Host: In your report, it says how mental health providers broadly talk about mental health doesn't resonate with men. How so?

Khush Amaria: Young men tend to see health seeking as a sign of weakness. And so we need to change our language. We need to be talking about supports for mental health as being a sign of strength, as a sign of improvement. This isn't just the typical clinical jargon. We want young men to have trusted voices, diverse voices coming to them, sharing with them the idea that we can, you know, change how we perceive challenges with mental health. And so I think there are ways that we can do that better. And part of it is genuinely by listening to young men trying to understand what is it that they need and they want, how do they resonate best with that?

Host: How does this compare with young women? I am curious about that.

Khush Amaria: Yeah. And I think we know that young women have some higher risks of identified mental health conditions. And so they talk about whether it's severe depression, anxiety. But actually the numbers for men are pretty high there as well. We see 1 in 5 young men report moderate to severe depression or anxiety. I think part of it is recognizing what are the issues that young men have faced recently that stops them from engaging in care as an example. And so we're trying to understand that better. And I think we're learning that the reasons that a young woman might engage in care versus the reasons that a young man might, might be different, but how they stay in care might matter even more so. So we want to understand from the young man's perspective, what is it? Why are they dropping out? We hear things such as, I don't feel like I have control over my decision making. When I'm in mental health services, it's very inflexible. I don't have a schedule that aligns with what I need. I'm not seeing progress. And we might see some of these same issues with young women, but I think we're hearing it from the young men right now. We need to change our systems.

Host: I want to talk a little bit more about that stat, the 1 in 3 young men, because if you delve into it a little bit more in your report, the number rises to 1 in 2 among newcomer young men. What's the difference there in terms of the conversation around mental health there?

Khush Amaria: And I think part of what we're trying to understand is why, oh, sorry, not so much why, but what can we do differently? We recognize that mental health isn't spoken the same across the world. And so young men are coming to Canada with other challenges and barriers, and they may even more so than than our Canadian young people see mental health as a sign of personal weakness. And that's a huge barrier. If you identify that as personal weakness, you're not going to ask others in your community for help. You're not going to ask others about where you can access, you know, easy to find services, low cost services. So those barriers are by no means insurmountable, but they are something we need to understand. And so it is it's worrisome that it's even higher than that newcomer young men population.

Host: I want to talk about some of the tools in the toolbox. There are some new ones. And according to your report, young men are equally likely to engage in AI tools for mental health support as young women, 12% for young men and women. Is this a good sign that they are asking at least? Maybe not necessarily someone, but they're asking and going somewhere for these questions.

Khush Amaria: Yeah, absolutely. I mean, you know, if you've if you're a young person, man or woman, but if you're a young man who's recognized, right, well, before things have gotten really bad that you need support, people often want to go somewhere. We do want them to go to trusted resources, trusted supports, evidence based supports and AI tools. I do believe they'll get there. We're not there 100% just yet. We need to figure out how to get to that point. But I think the model here that we're seeing is that we're trying to teach them how to identify that they have concerns that can be supported with the systems and tools that we have well, before things get out of hand, well, before they start using unhealthy coping strategies. And so AI is going to be a partner in this where maybe we're not quite there just yet, but we can't pretend that it isn't going to be something that we're going to be faced with and getting young people to be able to use in a way that would be helpful.

Host: Well, let's pick up on that. AI just not being there just yet. How do we understand the promise and peril of AI chatbots when it comes to mental health as a mental health resource?

Khush Amaria: They've actually been around for quite some time. There's been all sorts of models, actually, that have helped people use some really good evidence based standard CBT, for example. So cognitive behavioral therapy is a pretty common mental health related support tool out there. And so if I was, you know, encouraging somebody to use a tool that was in front of them, that's probably what they would start with. And so I think that we're going to get there. I'm not worried as a psychologist that I'm going to lose a job or anything like that. If anything, I just want more and more tools in the hands of young men and Canadians across the board.

Host: Your report also delves into different coping strategies for for young men. And we have seen an explosion of gambling ads and options across our TVs and platforms since Ontario legalized single game betting. What's the story here and how does it fit? When we talk about the mental health supports for young men.

Khush Amaria: The biggest lesson I think we learned from this report and just understanding young men is the risk is high, right? We see about 1.2 times more likely to experience problematic gambling for young men, it's it's double. I think what we see across the population is about 15% compared to 7.5%. We know that addictions often start at a young age under the age of 29. And so I just think of this as a compounding problem. If young men are turning to gambling, that is readily available as a coping tool, as a tool to deal with stress, anxiety, depression, this is worrisome to me. This is worrisome that we won't be able to provide young men with alternative, more healthy coping strategies when these things are much, much more readily available in many ways.

Host: Khush what are some changes you would like to see that will help address the mental health issues we are seeing in young men today?

Khush Amaria: I think probably the most important thing is we need to co-design, right? And I don't mean after the fact, I mean truly co-design with young men. We want them to feel that they are part of the solution and understanding the care that they might need to live their healthy life, whatever that might look like for them. We need to change the language, right? I use a lot of clinical jargon. As a psychologist, I need to learn how to talk about access to mental health care and support in a different way. That really resonates, again, with young men, whether it's talking about improvements, signs of personal strength. We need to also involve those that are in the young men's lives right now that are trusted sources, whether it's counselors, coaches, teachers, whether it's platforms, AI platforms, Reddit. We need to bring all of this together so that we can deliver the type of care that young men really deserve to be able to live their best life right now, especially with all the challenges they may be facing.

Host: Khush we are going to leave it there. I really want to thank you for your time. An important conversation.

Khush Amaria: Thank you.

Host: An AI chatbot can't reject you and it won't judge you and it probably won't leave you on read. So could they be an antidote to loneliness? University of Toronto psychology professor Paul Bloom says for some people, the answer may be yes, but that's not the whole story. Rundown producer Eric Bombicino talks to him about why being lonely matters, and what we may gain or lose by turning to AI for connection.

Eric Bombicino: So you've been putting out some really fascinating work on AI and loneliness. So we're going to get into all that. But first, let's sort of set up our problem here. The United Kingdom created a minister of loneliness. Japan created a minister of loneliness. How should we understand the problem of loneliness?

Paul Bloom: Loneliness is awful. Loneliness is one of the most terrible forms of human suffering. And, you know, you can think of it in terms of solitude. You just have nobody in your life. And a lot of people in the world are like that, particularly old people institutionalized, no friends, no family, spending their days and nights alone. But, you know, you feel loneliness. Even if you're surrounded by people, you don't feel loved or respected, understood and needed. Loneliness is terrible health consequences. It's presumably worse than being obese, worse than smoking a pack a day. Yeah. But, even if it had no health consequences, anybody who suffered from periods of loneliness knows how bad it is.

Eric Bombicino: And there's an evolutionary component. There's this. I think there's this feeling of way back in the day, prehistorically, if you were away from your people, it was a danger you survived with others. So there would be these evolutionary like alarm bells that go off. Is that why it's so painful? Does that tell a bit of the story?

Paul Bloom: I think it tells the most important part of the story. We live in an unusual time now, where I really could go a very long time without talking to anybody. I could order my food from Uber Eats and have them drop it off at the door. I get my entertainment through Netflix. I could read books, you know, and I live in a solitary environment and maybe do fine. But in the environment in which we evolve, everything of value, everything necessary for life was provided by other people. And so social isolation wasn't just kind of a bad thing. It was death. Yeah. And in fact, one when if you were to be abandoned by your group, it's essentially a form of murder. This is what they do when they want to kill somebody, they leave you behind. And then you're in the wilderness by yourself and you die.

Eric Bombicino: Yeah. And this is the hardware we're left with when we're alone. Okay, so in steps AI, we have, I think more and more people are using chatbots, be it Claude, be it ChatGPT. How can we, can we go in slowly here? How can, in your view, AI and chatbots help someone feel less lonely?

Paul Bloom: Well, a lot of people say that they get some sort of social sustenance from interacting with Claude or Gemini ChatGPT. The data are controversial and we should get into this. There's a lot of studies showing that in the short term you use, you find loneliness goes away. You enjoy these contacts in the longer term. There's a study that just came out looking at people over the course of a year, and at the end of the year, there were people get actually more lonely. If you're constant contact with with these with these eyes, if I was making the case for it and I'm willing to partially make the case. Okay, I'll go. I'll think of old. So there's a lot of elderly people, millions upon millions of people who are, many of them are institutions and institutions. Their family won't see them or their family is dead. They have no friends. They're totally isolated. Maybe they have dementia, maybe they have all sorts of problems that make them difficult to communicate with. They spend their days and nights in a agony of loneliness. And if and if some sort of AI could make that problem go away, it would be a godsend. It would be wonderful.

Eric Bombicino: That's kind of like antidepressants. They can work. They can help people in the short term, but sometimes not in the long term.

Paul Bloom: Well. It's like, so sorry. It's a great analogy. But what fentanyl is, it's like, like, fentanyl, powerful, addictive painkiller, right? I think I like that analogy because it reminds us we're going to talk about some of the negatives of this. But if I, if there was an 80 year old who was in constant excruciating pain given fentanyl, so they become addicted. So then they live a life of, of pleasure instead of agony. Yeah. And I feel the same way. If these things could be painkillers, they could be antidepressants. And as a result, they can make the pain of loneliness go away. There's a lot of positive to that.

Eric Bombicino: So what other moral problems do you potentially see?

Paul Bloom: Well, I'm not entirely positive about the use of these AI's alleviating loneliness. I think for the elderly, really, I really am positive. But but for younger people and young adults are those who suffer the most from loneliness. Yeah. I think that there's real risks and there's a couple of risks, but but one I'll focus on is from the sycophantic nature of these eyes. I think takes away the opportunity for social growth and learning. Yeah. So, you know, you're a teenager and you learn, you start, you tell long boring stories, but you learn that you're boring people and they tune out. You're not a very good social person. And so you get better. You, you learn to apologize because you know, you offend people. You learn to express an interest in other people because this is how they stick around. And are your friends. But an AI needs none of that. An AI will never need an apology from. You will never roll his eyes as you tell your boring story. We'll never be offended. Everything you say is wonderful to it. Now, this might be immensely pleasurable to deal with, but it's also a recipe for creating a really awful person. So if we did, and this is the big if, if AI companions evolve to a level that they did solve for loneliness, bigger picture, would we be better off? In some ways? I mean, my, my, my answer is not going to be a neat one. No, no. No, in some ways. Ways obviously yes. Because loneliness is a form of suffering and parallel question. If we'd all take enough fentanyl that it would would obliterate all of the everyday aches and pains, all of the suffering of chronic pain and so on. Wonderful, wonderful, wonderful. But the negative side is, say, just zooming in on AI, the negative side is it might ruin our capacity to have relationships with others. It might it might make us worse people. It might, it might make us lose the sort of social give and take with other people who have their own needs, their own desires, their own priorities that I think make us human. And you take that away. I think in some ways that's a disastrous consequence.

Eric Bombicino: So do we need and we it's tough. There's lots of grey area here, but do we need loneliness in our lives? Sometimes in some measure.

Paul Bloom: I think we definitely need social pain. I think social pain is the corrective force. Just like, you know, a toddler learns to walk by falling down. You know, a cook learns to cook by tasting the food. And it's not quite right. A boxer learns to block by getting punched in the face. We learn to become social creatures by messing up, sometimes by trial and error and full blown loneliness. Maybe it doesn't have to be, but it has to be a sort of a sort of a little bit of the hurt of not being understood, the hurt of, oh, I've offended somebody. I've bored somebody doesn't feel good. I'd rather not have that. But without that, you know, we wouldn't grow to be people.

Eric Bombicino: But you said that you don't. You don't think AI is conscious? No I don't. Okay, so if AI doesn't, AI doesn't have a mind. Do you think there needs to be a mind for there to be real empathy there?

Paul Bloom: Yeah, I do. I think that, I think that if I'm, if I'm communicating to a chatbot and I say, oh, you know, I really feel for you and it sounds awful, right? Maybe there's delusion of empathy and maybe delusion of empathy is enough. There's moral questions about AI that fool you. Yeah, but still without a mind, without consciousness, it's just, you know, it's fancy autocomplete. It's a stochastic parrot. It's all of those things. Yeah, yeah. Now, I don't think current AI's are conscious. Okay. And it's a controversial theory. The godfather of AI, Geoff Hinton, a fellow fellow U of T faculty member. Yeah. Says current AI are conscious and they have emotions and they have feelings and so on. I think this is mistaken, but I grant the possibility that ten years from now we will be dealing with conscious AI's and that's going to change everything. But what do we know? We don't really have a test for consciousness. So so I think in your piece, you talked about how folks that might be getting some nourishment from talking to an AI and feeling less lonely, there might be a level of self-deception there where they have to, but we don't we don't really know for sure. We we can't know when and if that happens. No, you're right, we don't know for sure. We have no good theory of consciousness that could apply and test that's conscious, that's not conscious, that's conscious, that's conscious. So the debate over history has been over all sorts of animals. Most people would say, oh, well, a dog or a cat has some degree of real experience, can feel pain. For instance, you know, what about cockroaches? Bees? You know shrimp. Yeah, it gets more complicated than that. And now we have to ask the question about about AI. And we have no way of knowing. All we have is intuitions, basically. Yeah. So in terms of therapy, I was once taught cognitive behavioral therapy by something called a robot. It wasn't large language model. It was done by these brilliant folks at Stanford. And that's very formulaic. But for a lot of people, therapy is a big part of it is just being able to sit with someone and having them having an actual mind listen to you. So I'm curious, where are we at now with understanding how effective AI can be? Chatbots can be. As a therapist. There are some preliminary studies which provide somewhat promising results, but I wouldn't I wouldn't trust them or make too much of them. Yeah. I think for some of the therapeutic interventions that are sort of by the numbers where you say, do this, then do this, then do this, fill out this, this checklist, and so on. Versions of CBT, for instance. Yeah, I wouldn't be surprised to see AI chatbots doing quite well out of them, but for the sort of, you know, tell me about your mother sort of thing that the sort of more tell me about your life, the, the deep relationship forms of therapy, then it's a bit more complicated.

Eric Bombicino: And they can be quite sycophantic. So I found if I put queries, it'll almost know what I want to hear. And sometimes maybe that's not great for a therapist to just go along with whatever your whims are. And of course, we know the very sad, tragic case of, of someone committing suicide with the help of, of AI. So the sycophancy of AI there sucking up is a huge problem.

Paul Bloom: Yeah. They are there are now many studies finding that they are these AI, the big ones are far more sycophantic than any person would be that this sycophancy actually affects people. Yeah, people actually believe what they are told. They believe that they are smarter. They believe that they are more moral. And a recent study that just came out, I think a couple of weeks ago, finds that exposure to sycophantic AI makes you less satisfied with your existing human relationships because, you know, the AI is telling me I'm brilliant, I'm beautiful, everything I say is like, you know, my wife, my kids, my friends, they don't do that, right? They don't appreciate how wonderful I am. And so they're kind of grace. And I want to go back to this sycophantic machine that really makes me feel good about myself.

Eric Bombicino: It draws out the more narcissistic parts of us.

Paul Bloom: Yeah, a little bit.

Eric Bombicino: And you might think, well, how big a deal could this be? You just go, you go, you go. You could change the settings. Everybody who uses chat or Claude knows you can just change the settings and say, stop doing that.

Paul Bloom: Yeah, but I think the problem is in the short term, at least we don't want it to stop doing it. We like the feeling of being praised and being comforted. Yeah. So and in that way too, it's like a drug. And back to the empathic part. I've always been quite obsessed with how we look at each other, not just you and I, Paul, but people in general. I think a lot of therapy is undoing and learning how you were seen when you were a kid. If you were neglected, if you weren't attuned to, if you were seen with malevolent eyes or angry eyes. And then on the other side of that, if you think about your best relationships in your life, those are the ones where you feel seen and heard. And for me, that's there's this feeling of home. And if I was going to think of a word that's the opposite of loneliness, it would be home. And home may be where you hang your hat, maybe where your heart is. But I think home is where your people are. Home is inevitably relational. And so I'm curious, do we need others to feel that sense of home and further down the horizon? Could a chat bot ever offer that sense of home? I mean, the Melfort now at least some people, a lot of people receive solace from chatbots. Yeah. We don't know where the technology will be 10 or 20 years from now. I mean, right now, for instance, we communicate with chatbots by talking or by typing. But how far are we away from something where I'm sitting staring at a screen, a big screen, and there it is, as if I'm looking at you. You know, the handsome, normal looking guy staring at me with interest and engagement. Okay, but now as an AI, but but I can't help but see it as a person. Yeah. Okay. And what we do that with everything we, we, we, we see minds everywhere very easily. Is that a good thing or a bad thing?

Eric Bombicino: It is.

Paul Bloom: We talked before about evolution. It makes kind of sense to, to quickly jump to the inference or somebody there or somebody looking at me, you get a wrong, no big deal. You get it. But you don't want to miss a case where there's a real mind out there. Yeah, but but I think it could lead us to be fooled. Yeah. And and and that would have moral problems. You know, if we if, you know, there's a little there's a little bit of irony because we, we take animals which we know are conscious. If we could know anything, we know pigs and cows are conscious, you know, they. Can recognize themselves in a mirror. You know, chickens, pigs have deep, complex social hierarchies, relationships and so on. We, we, we, you know, they die terribly by the billions. And it would be funny or ironic, at least if we continue to neglect their suffering, but become exquisitely attuned to the suffering of chatbots.

Eric Bombicino: So what other moral problems do you potentially see?

Paul Bloom: Well, I'm not entirely positive about the use of these AI's alleviating loneliness. I think, for the elderly, really, I really am positive. But but for younger people and young adults are those who suffer the most from loneliness. Yeah. I think that there's real risks.

Host: Thanks for watching The Rundown. If you're looking for a connection, let us know. Email us at rundown@tvo.org or leave us a comment on our YouTube page. Until then, I will see you tomorrow. This week on The Rundown.

Paul Bloom: I think it's nice when students see themselves in the history that we're learning, but it's equally important for them to learn about others through the history that we study.

Host: That's this week on The Rundown.