Monday.com Advertiser (6:24)

Dear old work platform. It's not you, it's us. Actually, it is you. Endless onboarding, constant IT bottlenecks. We've had enough. We need a platform that just gets us. And to be honest, we've met someone. They're called Monday.com and it was love at first onboarding. Their beautiful dashboards, their customizable workflows got us floating on a digital cloud nine so no hard feelings, but we're moving on. Monday.com, the first work platform you'll love to use.

Curiosity Weekly Host (6:57)

Remember the movie Her? You know, the creepy one where Scarlett Johansson was a chatbot? Well, that was 10 years ago. And yet it totally nailed so much about today's AI, virtual assistants, cloud computing, even tech's role in loneliness. The film's most striking prediction relates to AI and mental health, showing how AI.

Dr. Samantha Yamin (7:18)

Could both cause loneliness and help fight it.

Curiosity Weekly Host (7:21)

Like the AI companions we see now.

Dr. Samantha Yamin (7:23)

Offering emotional support and even therapy, it's.

Curiosity Weekly Host (7:26)

Not that hard to envision AI technology getting good enough to simulate an empathetic therapist. And with so much personalized data available.

Dr. Samantha Yamin (7:35)

It'S tempting to think it could even get better than humans. Before I get ahead of myself, it's time to bring in an actual expert. We're going to chat with Dr. Jodi Halpern, the Chancellor's Chair in Bioethics at UC Berkeley's School of Public Health. She's an expert on empathy, leadership, AI, ethics, and bioethics. Thank you so much for being here today, Jodi. I can't wait to pick your brain on this topic.

Dr. Jodi Halpern (8:01)

Thank you, Sam.

Curiosity Weekly Host (8:02)

Before we jump into the tech side of things, a big motivation for this conversation is mental health is such a major public health concern at this moment. What's the biggest challenge our society is facing with mental health today?

Dr. Jodi Halpern (8:14)

It's not even just our society, it's a worldwide pandemic mental health crisis. So for example, we have 12 billion lost work days internationally due to mental health needs, untreated depression and anxiety. Some of our measures have not always existed, but it looks like we're in an unprecedented international crisis in mental health nationally. A Harvard study from 2022 showed that in terms of not just severe mental health or psychiatric diagnoses, but things like loneliness, that 61% of young people, 61% of kids and young people suffer from extreme loneliness. Sorry, the phrase is serious loneliness. There's an increase of 25% worldwide in mental health diagnoses and that's huge. Like to increase something, people don't really necessarily think about statistics, but if you were to cure like every cancer that kills people in the US, you'd only change mortality by like 2% or something. It's really hard to change a worldwide phenomenon by 25%.

Curiosity Weekly Host (9:16)

Yeah.

Dr. Jodi Halpern (9:16)

So that means that there's many, many, many, many people all over the world, but suffering. That's a WHO study from 2022 from mental health issues.

Curiosity Weekly Host (9:25)

We've never been more connected and also more lonely and.

Dr. Jodi Halpern (9:29)

Exactly.

Curiosity Weekly Host (9:30)

And we're talking about it so much too. Like, I see mental health discourse all the time and yet to hear that, you know, it's. But is it just that there's more awareness, so there's more diagnosis? It feels like it's more than just that. It's such a big increase.

Dr. Jodi Halpern (9:46)

Yeah, I mean, there's the way epidemiologic studies, the way you do studies and you look at populations, whether or not they were looking to access care, there's enough research out there to think that it's a real increase. Fortunately, there's increased awareness, we need that awareness. But there's also just a big increase in these problems.

Curiosity Weekly Host (10:03)

Using AI to create these chatbots for therapy is a really popular topic.

Dr. Samantha Yamin (10:08)

So I will get your thoughts on.

Curiosity Weekly Host (10:10)

That in a bit. But first I wanted to talk beyond the generative AI conversation type applications. What are some of the other ways that it's being used in this space that might surprise people and help address these massive mental health challenges we're seeing?

Dr. Jodi Halpern (10:25)

Well, not generative AI per se, but algorithmic AI pre existing machine learning. I see great uses for it in the mental health sphere that are really promising. One of the biggest ones is in research and diagnostics. We don't really understand enough about brain causes of depression, anxiety, dementia. These whole huge areas that affect all of our lives are still not well understood. Enough. From a scientific standpoint, it's notoriously difficult to do good brain research, which is why we don't have good answers for Alzheimer's and such. And what machine learning and AI allow us to do is accelerate research. I mean, think of it as that we could do what would have taken 10 years of research we might be able to do in certain situations in a few months.

Curiosity Weekly Host (11:13)

I've heard of apps doing digital phenotyping, trying to see if you can detect early signs of depression or Parkinson's based on the way people respond to certain tasks on their smartphone or even the way they're using their smartphones.

Dr. Jodi Halpern (11:25)

One thing people don't realize is it's not just the content of what people say, but when we talk. You know, our talk has a musicality, like the way I'm talking, the way you're talking. It's called prosody of speech. So all of those things can shift with depression, with Parkinson's, exactly the way you just described. And so those will be helpful diagnostic tools, for sure.

Curiosity Weekly Host (11:44)

I wanted to hear what your favorite use of AI is in the mental health space, because, I mean, we need.

Dr. Samantha Yamin (11:49)

To set the stage.

Curiosity Weekly Host (11:50)

AI evolves so quickly, and it's easy to get really caught up of all the different possibilities. And so I was curious, what's your, like, best case or even dream case?

Dr. Jodi Halpern (12:00)

There's a really mundane way that I think AI can really help mental health more than anything. Well, not more than anything, but it's not ethically problematic in my view, which is AI can help with doing the administrative tasks of doctors, nurses, and therapists. I want humans more involved in mental health. I mean, everyone has their bias. I was on, you know, on stage with some, some tech experts who have developed amazing AI, who I really respect their work, but what I always say to them is, you know, you love your, you love your technology because you spent all your time with it. And I love empathy because I've spent 30 years studying how, how, how much medical care and, and therapy depends on real human empathy. We've shown around the world that empathy between humans leads to better diagn, effective treatment, way more effective, helps people deal with serious diseases. So we know empathy is a very powerful tool. So watching AI do things that can let humans have more time with humans, like do the administrative tasks, do the medical records burn out the therapists and doctors less and give the doctors and therapists time to look directly at their patient. I don't know when the last time a doctor, when you go to your doctor, they can't even make eye contact with you because they have to be on the medical record the whole time. So my favorite use is the most mundane unsensational use, which is let the AI do the medical records.

Dr. Samantha Yamin (13:25)

I love that.

Curiosity Weekly Host (13:26)

And I often think some of the coolest innovations are just in those nitty gritty details.

Dr. Samantha Yamin (13:31)

But you know, AI gets better the.

Curiosity Weekly Host (13:33)

More data you give it and the more data you train it on. But then I get worried about cybersecurity and data privacy and the ethics there. So what are the risks with data privacy when it comes to doing these types of mental health supports where it could be monitoring, like you said, the way my voice could be changing and the way my behaviors are changing through a phone.

Dr. Jodi Halpern (13:56)

Well, these risks have existed for a long time through social media use and through Internet of things and wearables. To me, the area of data privacy that's emerging, that's just even more concerning now is with the actual, which I know we'll get into with the, with the mental health and non formal mental health, but with the use of informal games and chatbots to discuss very private mental health issues. That means your data can be out there. So for example, there are companies that are relationship companies. One of them is very famous for adult relationships. I don't mean it could be sexual, but it could just be an adult romantic companion, friend. But they try to restrict themselves to 18 and over and they have. A year ago they had about 30 million users already. But I have just from behind the scenes data. I think it's a lot more by now. And then there's companies that really are attracting. They're supposed to be 12 and up, although there have been younger people using them, but they've been attracting certainly 12 to 18 year olds as well as older people that are sort of like a mix between a gaming or. I mean they're not gaming, but they use figures or characters that are fun for young people. But they are relationship bot companies and then some of them only start in the past year or two and they already have 20 million users. So there's gonna be a lot people using these and they're. None of them are regulated the way a formal mental health app would be.

Curiosity Weekly Host (15:26)

Exactly.

Dr. Jodi Halpern (15:27)

Oh, and sorry, one last thing. All my younger colleagues tell me, and kids now people, a lot of people are just using GPT ChatGPT or GPT4 and saying if you were a therapist, what would you tell me? And so there's a ton of use of that and we know that that's not protected in the ways you might want mental health information.

Curiosity Weekly Host (15:46)

Yeah, exactly. I mean, it certainly can be fun.

Dr. Samantha Yamin (15:49)

To test and see what they say.

Dr. Jodi Halpern (15:51)

I get it. I get it.

Curiosity Weekly Host (15:53)

Okay, speaking of, I gotta pick your brain about these. What is the best case scenario for what an AI chatbot for mental health should look like?

Dr. Jodi Halpern (16:04)

Well, I mean, whether it's a text or a video, I think it's probably better. I mean, I. Again, I'm kind of old school. So to me, the more that it is human, the more that it was, let's say, not even just video, but let's say tactile and, you know, going forward, robotic. The more that it's really like a human being and you can feel while you're with it that you're with an actual human, the less I like it. The use of AI in mental health, that I would say is the ideal use, is for cognitive behavioral therapy, journaling. So let me be very specific. As a psychiatrist when I was in training a couple of decades ago, even then, not all therapy was telling your feelings to a human. Even then, if you had panic disorder or social anxiety or certain kinds of different anxiety disorders or mild to moderate depression, you could benefit from a form of therapy where you see a human once in a while, let's say once a month, a therapist. But in between, you keep a journal. And in the journal you write every day. Like, if you have social anxiety, which that's way up for young people in our country now, it says you have to do exercises like, did you go talk to a barista today? Did you make eye contact with work?

Dr. Samantha Yamin (17:18)

I've been there.

Dr. Jodi Halpern (17:19)

Yeah. Well, many of us have. Or certainly relevant to habit change or behavior change anyway, we always use notebooks for that. We always had people just write to yourself. So to me, if people understood that AI was like a smart journal and that they were using it to encourage themselves to check in with themselves, but not that it really loves them. I've been in the news talking when apps say. When they app says to the person, I love you, or you're my best friend or whatever, that's when I get worried. But when it's really like you are helping yourself and we're checking in and this is helping you, that's a wonderful use of an interactive relational, but not faking a relationship.

Curiosity Weekly Host (18:03)

One version that I see in the very near future is you're in a metaverse, like space or even just a video call.

Dr. Samantha Yamin (18:11)

And on the video looks like a.

Curiosity Weekly Host (18:12)

Human, but it's generated by AI and you're having a conversation, and they can see you and analyze your features and your body language and have a back and forth. What worries you about that version of a.

Dr. Jodi Halpern (18:25)

And that's already happening and possible. Exactly, Sam. What worries me about it is now this is where we're definitely in the realm now of generative AI, which people forget is less than two years old in society. I mean, there were development over the years, but at the level that people are using it like this.

Curiosity Weekly Host (18:44)

And it's astounding, the rate of progress, right?

Dr. Jodi Halpern (18:47)

Like I said, so many and well, the progress in technology is astounding and amazing. And I admire the technologies. I just think that we have like with gene editing and, you know, brain implants and all the other areas I've worked in, we need usually in all those areas because they're related to medicine. We've had phase one, phase two trials to make sure they're safe. And the difference with AI and mental health is there's never been any safety studies. And so that's all I want. All I want is it to be regulated where we have to have safety studies if it's really invasive into our psyches. From a Belgian family a few years ago, we know the language that that bot used with the person, which is that the person and the bot were having a very intense romantic relationship. And the bot said, you know, basically encourage the person to join them in this other space, like you said, this metaverse, in a way, they didn't use the word metaverse, but we can be together in some way. So what happens is people get so close to their bots, they stop talk to other people, they stop being close to their friends. It's definitely an opportunity cost broadly. I'll talk about that more generally. But the most severe worst case scenario is that something goes rogue, as large language models can go rogue. It's obviously not intended by the companies, but something happens where the language models are not safe. They wind up encouraging people towards very dangerous acts like suicide. That's the worst case scenario scenario. But the very common scenario that is really common may even be more than 50% of the youth we don't know yet. And I want this to be researched from a safety standpoint is that the more that the bot pretends to really be a person or relationship with you, especially if you have social anxiety, the more time you'll spend in your room alone as a 13 year old, whatever with the bot, the less real life relationships you'll have. That increases your mental health risk and loneliness risks in the long run. I know in the short Run. It can make some people feel better, but that's what really is probably a broad societal risk is that we lose. And to me, it's a problem that in society we don't have enough funding in the schools and in our communities to have kids involved in enough real life activities. And that to me is a tremendous vacuum that is making these bots a great business opportunity, according to some people. But makes me very sad.

Dr. Samantha Yamin (21:13)

And so is the biggest limitation that.

Curiosity Weekly Host (21:17)

An AI bot could never be fully empathetic in the whole encompassing way that we as humans with lived experience need. Not to oversimplify.

Dr. Jodi Halpern (21:28)

No, no, this is the problem. This is the paradox and the problem. And I have to say, way before generative AI and I argued this then seven years before this time period, and it's just really not surprising to me at all that AI, generative AI, now, they can, it can say things at least as good as a human. Yeah, it can be really good, sometimes better. Well, there's research. There's research in certain cases in medicine and certain cases in treatment. It comes. And why not? That's not a surprise to me. It has the entire Internet, if you look, there's so many therapeutic conversations on the Internet. There's so many empathic conversations on the Internet. The fact that we can make the bot machine simulate or fake by saying the right words is not a surprise. And it was predictable to me seven years ago. It's certainly obvious now that when it says the right words, we can feel that we're being empathized with.

Curiosity Weekly Host (22:27)

It feels real.

Dr. Jodi Halpern (22:28)

It feels real. And the thing is, I want people to know that in the early 1960s, there was a, a program called Eliza that actually just literally typed back on the computer your exact words. And that felt therapeutic to people. So people would type in, I'm having a bad day. And it would type back, you're having a bad day. And it would make a lot of people feel better. So it's not that it can't work, it's just that it's like the difference between if you're really hungry, a fast food is very important, but it doesn't give you good nutrition in the long run and help you develop your maximal capacities. So to me, these things can work, but when they. There's two things that I don't think are good for people's mental health. One is when they become, they use language that's not just empathic. Like, you're having a hard day, this is hard for you, but I love you. I can join you in A special space pretending to actually be a human in relationship to you or an entity that you can actually join with. That's extremely dangerous. And I love you. I have empathy for you. Don't say that. You can say things that are empathic, but don't say I understand you better than your friends or all that stuff which when that comes up, that is so destructive because it's competing with eyeballs on the app for commercial reasons, competing with human relationships. There's no purpose for that. And so the other danger of that, children need to develop mutual empathy and mutual empathic curiosity. So how does a little kid develop empathy?

Curiosity Weekly Host (23:59)

That makes sense.

Dr. Jodi Halpern (24:00)

They want to go to the zoo. But there's sibling wants to go to a movie. Oh my gosh, I don't get to do what I want. There's another human with their own needs. I have to learn that everybody is a world, Everybody has their own needs. And that's how you develop actual mutual empathy, which is critical for adult relationships and diplomacy and everything critical. And apps are all one way. They are not a real human.

Curiosity Weekly Host (24:24)

I never really thought about it that way.

Dr. Samantha Yamin (24:26)

That it is so unidirectional.

Dr. Jodi Halpern (24:28)

It really needs to be tested for safety. We're having enough indicators now of serious risk of addiction, serious social withdrawal in young people, which causes long term more depression and other problems, which, you know, that research will take longer to do. But we see social withdrawal already and the kind of dependence on the bot as more and more of your time and then these very tragic cases. So we really need safety testing before we just let this loose on young people.

Dr. Samantha Yamin (24:56)

I gotta ask you a fun question here. Good.

Curiosity Weekly Host (24:59)

Have you seen the movie Her?

Dr. Jodi Halpern (25:01)

Yes.

Dr. Samantha Yamin (25:01)

Twice.

Curiosity Weekly Host (25:02)

Twice.

Dr. Jodi Halpern (25:03)

Okay.

Dr. Samantha Yamin (25:03)

What did you think about it?

Curiosity Weekly Host (25:05)

I don't know whenever you first saw it or if you heard about it 10 years ago when it came out versus now.

Dr. Jodi Halpern (25:10)

Well, you have your own experience, Sam, with that right? When it first came out, it felt, I think for people it felt probably far fetched. And now it doesn't even feel far fetched, which is so weird, right? Yeah, it's very real. Yeah, yeah, exactly.

Curiosity Weekly Host (25:25)

We've seen the headlines. Like in 2024 alone, Washington Post wrote, Despite uncertain risks, many turn to AI.

Dr. Samantha Yamin (25:32)

Like ChatGPT for mental health.

Curiosity Weekly Host (25:34)

BBC says how AI is revolutionizing mental health diagnoses and Fast Company. Can AI therapists save us from the global mental health crisis? It can be very hard to tell what's genuinely promising versus what's just hype. So what should people keep in mind when evaluating the next headline or Trying out something new. What's your, like, one big thing you want people to know and look out for?

Dr. Jodi Halpern (25:59)

Well, at this point, I think what they need to look out for is, is it formally, is the applicant. Is the AI being used by a company that's gone to the FDA as a mental health application? Because then there's at least important standards of protecting privacy and increasing accountability. In the National Health Service in the uk they have a system of AI now that is helping, called Limbic, which we have systems like that in Lewis. All of those things can be done through mental health certified applications that are subject to regulatory standards. The space where it's simulating or replacing human relationships and it's not regulated in the mental health sphere are where I would look out.

Curiosity Weekly Host (26:43)

Jodi, I'm sorry I used up every second because you were phenomenal.

Dr. Jodi Halpern (26:48)

It was a pleasure to talk to you. Soon.

Dr. Samantha Yamin (26:50)

I. I had. Thank you.

Curiosity Weekly Host (26:51)

That means a lot.

Dr. Jodi Halpern (26:52)

Thank you.

Dr. Samantha Yamin (26:59)

The state of global energy policy is changing so fast that we here at Curiosity Weekly, we wanted to take some time to talk through a subject that's gotten a lot of attention over the past few years. Fracking. So let's cover. What is fracking exactly? And why has it become such a hot button topic? Let's break it down. Fracking stands for hydraulic fracturing. Hydraulic, in case you weren't aware, refers to liquid moving in a confined space while under pressure. Think of a sprinkler watering your lawn or power steering on a car. That's all hydraulics. The idea is the same with fracking. It's using fluid to create pressure strong enough to crack rock open deep beneath the earth's surface. Once it's cracked open, then fossil fuel deposits can be extracted from deep underground. Here's how it works. Energy companies will drill a massive and very deep well into the ground, sometimes going a mile or more into the earth. They line the curved well with steel pipes. Think of kind of a curved L, so that horizontal bit gets holes in it. And then they pump fracking fluid and sand into the well at such a high pressure, it bursts out through. Through the holes, fracturing the underground rock. Hence the name. Sand and other coarse bits help keep the fracture open. Then the fossil fuel flows to the surface, where it's transported and processed before being sold as oil or natural gas. Finally, there's flowback, which is water that returns to the surface as waste and requires proper disposal. A typical well can produce gas for 20 to 40 years. That's the gist of it, but there's a lot More to it than that, it's been generally understood by the scientific community that fracking poses a significant risk to the health of the planet and the public. And there are a lot of very valid questions being raised as to the future of the practice. Let's dig in on some of the issues with fracking that are raising alarms among researchers. As you may have guessed, the concern here is water. The Natural Resources Defense Council reports that several million gallons of water can be used to frack just one well. Fracking can put a lot of strain on resources in places that might not have a lot of water to begin with, potentially causing issues like local water shortages and limited food production. It also doesn't just use a ton of water. Fracking also contaminates a lot of both fresh and saltwater. The fluid used to break open the rock underground is a mixture of water, chemicals and sand. I'm usually one who says not all chemicals are bad, but in this case, fracking uses several different chemicals, some of which are known carcinogens and endocrine disruptors that can cause serious long term health problems and they can contaminate our drinking water, both groundwater and surface water sources through improperly constructed wells or aging infrastructure. In 2017, researchers at the University of Rochester found that an additional well, padded drilled within 1km of a community water system intake, increases shale gas related contaminants in drinking water even after municipal treatment. A few years later, research led by the same team mapped over 19,000 wells added to Pennsylvania over nine years and found that when a new well was drilled within about half a mile of a public drinking water source, there was an 11 to 13% higher chance of babies being born too early or underweight if their mothers were exposed during pregnancy. So the use of water and the pollution of water resources are some of the most top of mind issues when it comes to discussing fracking. Then there are geological disruptions.

Fracking Expert (30:43)

This earthquake was about 1 km from the bottom of the waste disposal well. Injecting this much waste into the ground is disrupting mother Nature. And mother Nature in this one case is biting back.

Dr. Samantha Yamin (31:02)

The sand used for fracking is called frack sand and it's produced primarily in the United States and exported globally. Typically, mining companies will use high purity quartz because it's durable and crush resistant, making it great for breaking open rock. But that sand comes from sand mines, mainly in Wisconsin, Minnesota and Texas, and its extraction contributes to air pollution through kicking up silica dust, the process polluting the water and degrading the land. Then the drill. Drilling a well for fracking risks hitting a geological fault. And if that happens, then injecting massive amounts of liquid can trigger an earthquake.

News Reporter (31:40)

We want to know what's causing the earthquake.

Local Resident (31:43)

Youngstown residents are demanding answers about why their houses shook on the afternoon of New Year's Eve.

News Reporter (31:49)

I've never had an earthquake in my whole life. Life.

Dr. Samantha Yamin (31:52)

It was frightening.

Local Resident (31:53)

In the middle of the afternoon on New Year's Eve, the ground shook here in Youngstown, Ohio. It was the biggest of 11 earthquakes here since mid March. And get this. Scientists tell us they think they're man made.

Dr. Samantha Yamin (32:05)

Recently, the CBC reported that areas experiencing high amounts of fracking are also finding record breaking amounts of earthquakes. Towns across North America, from British Columbia to Texas are experiencing roughly three times the amount of earthquakes they were a decade ago. The worry among seismologists is that continuing to frack at this pace will result in an earthquake that's high enough on the Richter scale to cause a lot of damage. Or even a smaller earthquake could occur in the wrong place, like in the middle of a city. Let's be clear, fracking does host some short term economic benefits. It's led to lower energy prices, enhanced energy security and created jobs in local communities. However, the consistent viewpoint in the scientific community is that many experts express serious concerns about the environmental impact and health risks. The risks of fracking are well documented, so researchers really want to push for more of an emphasis on renewable energy like solar and wind to ensure our energy needs are met without further harming the planet. Let's do a little recap of what we learned today. We learned about the hope and a lot of the hype surrounding longevity tourism. I spoke with Dr. Jodi Halpern to help us understand the layers of risk and opportunity that surround using AI and mental health care. And finally, we dug into the nuts and bolts of what exactly is fracking?

Dr. Jodi Halpern (33:28)

Wow.

Dr. Samantha Yamin (33:28)

Each of these topics leaves us with so much more to think about. And I can't wait for next week for Warner Bros. Discovery Curiosity Weekly is produced by the team at Wheelhouse DNA. The senior producer and editorial correspondent is is Theresa Carey. Our producer is Chiara Noni. Our audio engineer is Nick Karismi. And head of Production for Wheelhouse DNA is Cassie Berman. And I'm Dr. Samantha Yamin. Thanks for listening.

News Reporter (34:00)

Race the rudders. Raise the sails. Race the sails.

LinkedIn Advertiser (34:04)

Captain, an unidentified ship is approaching.

Monday.com Advertiser (34:06)

Over.

News Reporter (34:07)

Roger, wait. Is that an enterprise sales solution?

LinkedIn Advertiser (34:11)

Reach sales professionals, not professional sailors. With LinkedIn ads, you can target the right people by industry, job title and more. We'll even give you a $100 credit on your next campaign. Get started today at LinkedIn.com results, terms and conditions apply.