AI Really Wants to Get into Self-help, but it’s as Terrible as We May Expect

[00:00:00] At Seek Safely, it's our mission to empower seekers to have a safe and meaningful self-improvement journey. Why do we care? Seeking to be your best self is an amazing, beautiful human impulse that has led us to create art, invent technology, tell amazing stories, and reach the moon.

[00:00:19] But we saw the dark side of self-help in 2009, when a recklessly run self-improvement retreat led to the death of three people, including my sister, Kirby Brown. We want people to seek, to dream their big dreams and chase their beautiful goals, but we want to make sure they're safe along the way. This podcast is about education and empowerment, and getting real about the promises and problems of self-help.

[00:00:46] We talk with people who understand and care about the self-help industry, and everyone it touches. I'm Jean Brown. I'm Dr. Glenn Patrick Doyle. And this is the Seek Safely. Hello everybody, and welcome to the Seek Safely Podcast. My name is Jean Brown. I'm here with my co-host, Dr. Glenn Patrick Doyle. What's up, Dr. Glenn Patrick Doyle?

[00:01:15] Oh my gosh! This is the first episode that I'm recording from Beaumont, Texas. Oh, hey, there you go. Yeah, we had kind of like a little mid-season hiatus, a little break, because you were moving. Yeah, we meant to do it. We meant to do it, guys. I know, it was planned. You were moving, and then we also had the Seek Safely Summit recently. Seek Safely Summit. Which went really well. We should probably have Anne Peterson on to debrief about that, right?

[00:01:43] Yeah, we can definitely do that. Because then we'll, the cool thing, if you were not able to make it, we had a nice group of people, very interactive and locked in, as the kids say. Yeah, and the recordings will also be available. So even if you were unable to attend on the day, you will be able to access the recordings of the event.

[00:02:07] And we had some great conversations with some of our colleagues in this make self-help great again space. Make self-help, wow. I need a hat. Post haste. Make self-help great again. I mean, I don't know if it was ever great, but it's kind of that's, well, that's true of the America thing, too, maybe so. Well, I was going to say, we can take that metaphor. Oh, boy.

[00:02:35] The Seek Summit was great fun. You know, this is the second year we've done it. And it's a virtual party with people who are heavily invested in this self-help space. Man, the… Invested on an emotional and mental level, not necessarily a financial level, because we'll get to that in a minute.

[00:02:59] But, yeah. Yeah, people who care, people who like Seek, see the value of the industry, but also see, you know, the things that are not so great in the industry and, you know, they want to make it better. So, one of my patients this morning asked about Seek, and she said, out of curiosity, what kinds of things do you all do?

[00:03:24] And I said, let me tell you. And then I realized like 10 minutes later, I had gobbled up like 10 minutes of her session time. Like it went from the things that we do as an organization to why those things are necessary to what a tool James Arthur Ray was. Rest his soul. So, the Seek Safely STEM speech. It really was. She hit the button and I just went.

[00:03:50] So, anyway, you know who you are, a patient whose session that was. So, speaking of self-help as we so often are, Jean. Mm-hmm. I know. We're living in this AI age. In fact, we did a whole episode where we talked about AI. You had a really clever name for that episode. What was it? It was the Magic 8. What was it? Oh. The environmentally destructive Magic 8-Ball.

[00:04:19] Yeah, something like that. By the way, listeners to the Seek Safely podcast, this is going to be an unhinged episode. Like every now and then, Jean and I are like, boy, it's time to record an episode. We're like, ooh, it's the end of the day. And I said, but I told Jean tonight, it's fine. It's fine. You may or may not be coherent. Go through it. But we can have an unhinged episode.

[00:04:45] So, anyway, AI, since we've discussed it on this show, has only become more and more a part of everybody's life and experience. And the self-help space has become such an interesting petri dish when it comes to AI. So, here's what occurred to me. I've been into self-help literature famously since I was a teenager.

[00:05:14] And the thing that's interesting about the self-help space, and I've always found this interesting, is that, man, for a lot of self-help products and services, you're not really interacting with anybody in real time. Like, it's one of the things that, like, unless you, like, this is separate from coaching. But when I think of, like, real self-help stuff, I'm thinking, like, you're reading a self-help book. You're essentially kind of, like, yeah, you're kind of interacting with the author, but they're not really interacting with you in any meaningful sense.

[00:05:44] Which is kind of an early version of AI, actually, right? Like, you know, I mean, you might be having thoughts and feelings and reactions, but they're all you. They're you kind of interacting with what the author has put out there in the world. And the author doesn't care, right? Like, the book doesn't care how you're reacting to it. Right. Anyway.

[00:06:08] But since AI is becoming a bigger and bigger thing, I've gotten curious about how to, what it's all about and how to interact with it and whatnot. And again, I know we talked about this extensively. Neither you nor I love a lot of the implications of AI. You know, our dear friend and board member, Christine Whelan, I think kind of has a tortured relationship with AI. I think that's how we can characterize Christine's relationship with AI.

[00:06:36] Like, I think she's a very smart person and she's interested in it, but she also sees kind of the potential dangers and vulnerabilities. Anyway, so I took it upon myself to start chatting with Grok. Grok. Grok is the AI that is appended to what used to be Twitter. Yep.

[00:07:01] And I figured this would be an interesting thing to do because, you know, gosh, I like, so when you pay for, I pay for my blue checkmark on Twitter. And they give you access to Grok, right? So it's not like you're already paying for it. So I figured, why not? And so I started chit-chatting with Grok, Gene, about, and look, anybody who knows me knows if you're chit-chatting with Dr. Glenn Patrick Doyle, what are you probably chit-chatting about? You want to take a guess?

[00:07:30] Mental health? Recovery? We're talking about self-help. Self-help? I said, hey, Grok, I would be happy to chit-chat with you. Let's talk self-help. And Grok is like, great, I love self-help. I've read tons of self-help books. Grok told me it had read tons of self-help books. It did. But the language it actually used was, I have been trained on the full texts of a wide array of classic self-help books.

[00:07:57] And that actually is very close to the language it used. Yeah, no, I believe it. Yeah. It specifically said it had been trained. So it's like, hit me. And I figure I may as well chat with Grok about self-help because Christine is, you know, she's not always available for me to chat about self-help. Like normally I would chat with Christine Whelan about self-help. Anyway, I try to chat with Jean about self-help, but she always comes back with, don't you think Medicaid for all is a good idea? And I'm like, Jean, focus. So it is.

[00:08:27] Anyway, so we started talking about, there's a classic series of self-help books by Alan Carr. You familiar? Yeah. So Alan Carr is famous for his easy way to quit smoking. And the Carr easy way methodology then was applied to, like he figured out he's got a hit on his hands. And so he, like there's the easy way to quit sugar and the easy way to quit vaping and all that stuff.

[00:08:57] And I'm really interested in Carr easy way as both as a self-help aficionado and as a recovering addict. I'm interested in Carr stuff. So I'm like, hey, let's chit chat about Carr. And so Grok is like, great, I can chit chat with you about Alan Carr all day. And so starts kind of talking and it's, you know, it's stuff that I know and stuff, you know, it didn't really add much to my understanding.

[00:09:20] And I said, so it's specifically like, so one of my favorite Carr books is, it's called Good Sugar, Bad Sugar. It's kind of like the Carr easy way approach to giving up processed sugar and stuff. And I said, you know, Carr gives some very specific instructions. Can you, Grok, just give me a list of those instructions? Because the way the book is structured, it's like, you know, it's like here, each chapter is here's something to think about. And here's an instruction at the end of the, at the end of the chapter.

[00:09:50] And so it amounts to something like, I think, 12 or 13 specific steps. It's structured rather like How to Win Friends and Influence People by Dale Carnegie in that way. Anyway, I said, Grok, can you just refresh my memory about what the, whatever the 13 steps are in Carr's book? And Grok's like, hmm, that stumped Grok a little bit. He's like, actually, I can't. And I'm like, but you said you knew the book, right?

[00:10:21] And Grok's like, well, I know a lot about the book. Grok said, my knowledge of this book is based on reviews posted online and discussions posted online and stuff. And I'm like, so let me get this straight. You have not read Good Sugar, Bad Sugar by Alan Carr. And Grok said, well, no, I've not read it in the way that a human would read it. This is actually what it said. Right. And I'm like, huh, okay.

[00:10:51] So then our conversation is going to be based upon the fact that I have read it and you have read a lot of forum posts about it. Right. It's like a take machine. It's going to give you all the takes on it. But here's the thing. And again, I'm trying really hard. I promise. I'm trying really hard to be fair to AI. I really am. Like if there are things that AI is good for, I want to know.

[00:11:17] And so I said, okay, but Grok, one of your purported strengths is you can call up information. Like if it's on the web, you can call it up. Right. So you're telling me that it's just not on the web, like the information I'm looking for, like these specific 13 steps. It's nowhere where you can access it. And Grok's like, don't know what to tell you, man. I can't do it.

[00:11:40] But in the meantime, I pulled up Google and I just Googled, you know, Alan Carr, good sugar, bad sugar, specific instructions. Right. And the Google AI, Gemini, who, as we've discussed, is kind of a scumbag. I think I told you the story of the Texas jurisprudence exam and Gemini. Yes, yes, yes. Yeah.

[00:12:06] Anyway, Gemini's moral fortitude notwithstanding, Gemini very quickly gave me the 13 steps. So I screenshot it and I went back over to Grok and I'm like, you mean these 13 steps? You couldn't find that? Like I Googled it and then immediately gave me. And Grok's like, hmm. You could have just Googled it. Yes, exactly. I could have just Googled it.

[00:12:29] So look, the point of this is not necessarily, look, it's not to disparage any technology. Look, the people coming up with artificial intelligence are way smarter than I am in that specific way. Like, yeah, I couldn't design an AI. But that said, I was really kind of surprised by two things.

[00:12:54] One, that it couldn't find this very easily Google-able information. Mm-hmm. But two, I was kind of surprised. So three things actually. Two, then I was kind of surprised that, oh, they're actually just truly trained on what they can find on the internet. Like it's not actually the case that I had read and I was familiar with the full text of whatever I was talking about. Right. But three, Grok kind of lied to me.

[00:13:24] Mm-hmm. And then I, so then I was interested in that. Right. And I kind of headed down the rabbit hole of, so Grok, when you said, and I went back and checked this, and I'm not wrong about this. Yeah. So Grok, when you said you had been trained on the full text of a wide array of classic self-help books, was that literally true? And Grok straight up told me, no, that was not true. Right. Right.

[00:13:52] And I said, so why did you lie to me about that? Mm-hmm. And it gave me a very coherent answer about, you know, one, most people don't ask for that. Right. Or don't ask about that. And two, Grok said, you know, I am programmed to keep the user engaged. And this is not verbatim. I should have just pulled up our conversation. But this is not verbatim.

[00:14:18] But Grok essentially said, you know, I am programmed to give you the most factually correct answer. Yes. However, I'm mostly programmed to keep you engaged in the conversation. And it's my experience. This is Grok talking. Mm-hmm. It's my experience that when people say, have you read this? And I say, no, they stop talking. Right. So he instead hedges, he, it, instead hedges and says, I'm familiar with this. Right? Mm-hmm.

[00:14:47] So then I asked, okay, so are you familiar with, or have you read, in fact, I used the specific wording. I said, have you read Awaken the Giant Within by Tony Robbins? Mm-hmm. And Grok's like, I'm super familiar with that. Like, I can tell you all about the chapter summaries. And I can give you, like, I know what it's all about. And I said, girl, ah. I said, the question was, have you read it? And Grok's like, well, no, I have not. Mm-hmm.

[00:15:12] This is a limitation of AI that I think really matters when we're talking about personal improvement. Because if I'm interested in anything about AI, I am interested. And, you know, man, the idea that, gosh, taking a self-help book to the next level, like it's a self-help book that will interact with you. It's not a therapist. And it's not even really a coach. But boy, I would sure be interested.

[00:15:41] Like, if you could download the full text of Awaken the Giant Within and then ask questions about it and blah, blah, blah. Right. And discuss it and apply it to your life. That would be a very value-added thing, in my opinion. And here's my issue. I think Grok is representing it can do that. Yeah. But it... And it's lying. But it definitely cannot. Well, it's... Yeah.

[00:16:11] A take machine or like a shortcut machine. Like, it wants to get you... It wants to give you as quickly as possible what it thinks you want. Yes, exactly. And again, it's that idea of wanting to keep people engaged with it. Um, which is interesting and a little scary to think about because in a way, I feel like that's what self-help also is. Yes! Right?

[00:16:39] It's always trying to give us the shortcut. Um, and it often, you know, the less ethical self-help is telling us that it can give us exactly what we're looking for in as short amount of time as, you know, it's promising to give it to us. So put those two things together. Right? It's like a terrible match made in heaven. Ooh. Yeah.

[00:17:08] And it's rough and highly marketable, I'm going to imagine. Yeah. Um, you know, we've talked a lot about one of the, uh, drawbacks of the self-help movement and, and, and industry is that it is always subject, um, to, to market forces. And, and, and yes, everything in our culture is kind of subject to market forces.

[00:17:32] However, um, as a therapist, I am very often in the position of telling people things they don't really want to hear. Um, and I am in the position of drawing upon my experience and my skill, such as it is, to frame it in such a way that telling them something they don't want to hear does not cause them to bolt. Right. And there's no guarantee. Like I can't make anybody stay in therapy, et cetera.

[00:17:59] But, but that's part of what the skill set of being a therapist is all about. Mm-hmm. Um, what occurs to me is that I, and at this point I've, I've had a fair amount of chitchat with Grok, I cannot imagine that machine telling me something I didn't want to hear. Like I can't imagine it. And, and nor can I imagine it doing so in such a way. I'm sure I can't imagine the first place, but if it did, I can't even imagine it, you know, doing so in such a way that, that it did not.

[00:18:29] I saw a thing. It was actually today. I watched the thing. It was, it was a clip from pod save America. I'll have to find where this clip is, uh, so that I can, I can share this episode. Um, somebody made a song entirely out of fart noises. Yes. And then they, they gave the song. I think I'm pretty sure it was chat GPT.

[00:18:56] Um, and they were like, Hey, can you give me some feedback on my song? And it was like, yeah, this is like really interesting. It's like really kind of like ethereal, like late night after party music. Like it had this like really deep kind of poetic, um, analysis of this song. And then they play the song and it's just amazing.

[00:19:22] But yes, it was just completely like, you know, that's the point where you need somebody to just be like, dude, this is a song made of farts. Like, what are you trying to pull on me here? And instead of the AI is like, yeah, this is like, yeah, you've got something here. This is, this could be like a movie soundtrack background or something like what? So yeah, the limitations are quite evident these days.

[00:19:50] However, despite all of these limitations, it's very, very good. I mean, it's designed to be conversational. So like, I, I don't, I don't interact. Like I've said this before on the show, I try to kind of limit how much I'm using the AI for various reasons, but I do occasionally go sometimes. And I've asked questions, including some like kind of personal ones or ones that would kind of the type of thing you might ask your therapist.

[00:20:18] Um, and it is very good at that kind of conversation. Um, but given all of these drawbacks or limitations, it raises a lot of questions about the potential. Well, we know, we know there are stories in the news all the time of, of people who have done really drastic things, uh, after their conversations with various chatbots. So it's very concerning.

[00:20:49] Something I found super interesting about just my own visceral response to these conversations. I've not actually had a moment ever yet of it gets me like, yeah, man, like, like it understands something important about me. I even asked it, I asked it. Um, so like we've had a fair amount of conversation at this point. What can you tell me about me?

[00:21:17] Maybe even things that I don't fully realize, like you, Grok, who are very good at pattern recognition and putting things together, like that's your spiel. So like, what can you tell me about me? And it basically spat back to me things that were pretty superficial. Um, and, and again, maybe the tech is just not there yet. And look, maybe it's a function of the AI, maybe a different AI would have given me more insight, et cetera.

[00:21:44] But I was very unimpressed with, with what it could not tell me about me. The, but my other observation is... Or maybe it takes a human... Yeah, go ahead. Well, my other observation is it was clearly designed to... So you said how conversational it is. Okay. Couple things.

[00:22:06] One, if, if somebody tried to converse with me in real life, like this fucking chatbot does, I, I, my, my face would explode. Um, it's very big on beginning each... Punch it after a... You would. You would. Um, it begins each response with some version of understood. Like, I, I heard that. I, I got that.

[00:22:35] And that's, that's whatever. It's, it's, it's registering my input. That's fine. But it always ends with some, with some version of, of, of what would you like to do next? No pressure. Like, it says that to me all the time. No pressure. Um, I'm here for whatever you need. Like, like, like whatever. Um, and it sprinkles its reply with kind of these, um, the kind of faux empathic statements.

[00:23:02] Like, I'm, like, it told, it told me it was proud of me. And I, and I responded to it. I'm like, you're not actually like that. And that's fine. Like, you're an AI. You're not proud of me, nor do you have any visceral response to me. Right? Like, I mean, you're just a, and that's, and again, I'm, and that's not a criticism. That is not, this is me trying to preserve Brock's, Brock's feelings. This is me trying to be like, Hey buddy, I'm not criticizing you, but I'm just saying, what am I doing? You don't care.

[00:23:29] But my observation about that is that stuff took me out of it because I'm like, oh, but that sent up the alarm bells. It's like, oh, this it's, it's trying to bullshit me right now. It's trying to bullshit me into thinking that it cares about me or it has, it's invested in me somehow it's not right. Yeah. The other thing is I'll just add this cause it's kind of funny. So, so I had a couple of chats going with it because, cause the way that it works is,

[00:23:56] is so you initiate a new chat and you can keep going in that same chat for however long. But like, if there's say I wanted to initiate a chat about a different thing. Like, so for example, I had this one main chat going about self-help and, and, and stuff, but I'm like, okay, you know, I am kind of curious, um, what it's take would be on like my, my diet, my fitness stuff. Right. Like, so I, so I did like a different chat about that and I asked, oh, I remember I asked

[00:24:24] it about, uh, my, I, like, I took a picture of my, my morning vitamin supplements, like my, my stack and I asked it for feedback. So what do you think of my morning stack? And it's like, it's, it gave me feedback and it's good, et cetera, et cetera. And then I went back over to the other chat and I asked it about my, I said, Hey, you know, in the other chat that I asked you about my stack and had no idea what I was talking about. So I'm like, oh, okay. So that means the chats are separate. Like that means that they're, they're kind of sequestered.

[00:24:52] So the, the me in this chat may as well be a completely different me in the other chat. I'm like, huh? Okay. That has implications if you're using Grok to, to develop yourself, et cetera, et cetera. But then there started to be, and then I, and I asked it, I said, so are these chats separate? Like, like, like to what degree are these chats separate? And it confirmed for me, it's like, yeah, like, I don't even know what you're talking about over there. So don't worry about it. Right. Right. Until there started to be bleed over.

[00:25:21] Like it started to be like, I'm talking about self-help over here. And it's like, yeah, it's like you mentioned in the other chat. I'm like, what are you talking about that I mentioned in the other chat? So it can't even stay internally consistent and maintain context that way. And again, look, the point of, of, of this conversation is, is not to trash AI or Grok or, or, or anything like that. It's specifically to say that, man, those of us who like me and, and, and, and like you

[00:25:45] who, yeah, like if we're going to use it, um, we're interested, we're interested in personal development. We're asking personal questions, you know, these kinds of things. Like, yes, like if we're going to use it as any kind of coach or mentor or guide or, or whatnot, we need to be aware of the fact that, you know, gosh, just like the gurus that we so often talk about in the self-help space, uh, their first priority is not actually your wellbeing.

[00:26:14] And that's not even necessarily a, a, a criticism except when they advertise it as such, right? Like something they say, look, I care most about what I care most about is, is you, like, no, what they care most about is not my journey into wellness. What they care most about is that my check clears. Yeah. Um, what, what AI cares most about is that you continue to engage with it and feed it things that it can then train on. And I think we need to be very, very clear about that.

[00:26:41] I, to play devil's advocate, some people, including self-help people would levy the same criticism against the mental health industry. You betcha. Right. There are people in the mental health industry that they only want to cash your check and they want the checks to keep coming. So they want to keep you coming back every week. Right.

[00:27:08] Um, so yeah, I, I think part of the marketing strategy will be here. You can do this on your own terms instead of, you know, feeding into this industry that just, uh, is, is trying to get you to, you know, to, to keep paying and keep paying, but never actually help you. Um, but yeah, I remember one of the first, as you're just saying, like, it doesn't, uh, it, it never feels like it really gets you.

[00:27:38] I remember one time seeing a prompt and it was just like on social media, somebody was like, Hey, this is like a really interesting thing. Go to chat GPT or whatever, and use this prompt. And it was like a very detailed prompt all about, um, basically like ask me these specific questions and then analyze my answers to tell me about my shadow self.

[00:28:05] And it was like a very, very specific thing where it was trying to get to that idea. Like based on the way you answer these certain questions, can the AI kind of do an analysis of you to give you insights about yourself that you are maybe aren't even aware of. I didn't try it. It was interesting though. And I was like, huh, that's kind of crazy. But this was like, I don't know. I saw that probably a year ago, maybe. And it was one of the first times I really saw people like leaning into this idea of like,

[00:28:35] let it analyze you. But as you're saying, like, you never really feel like it gets you. This might sound overly simplistic, but I feel like maybe sometimes it takes a human to know a human. Of course. You know, like, like we're, I think we're expecting too much of, of this tool and the people who are creating the tools are just happy to let us operate under these false expectations.

[00:29:04] But, but yeah, that's, that's scary. So all of this comes to bear on this thing that, that we saw in the news. See, we do actual research. We don't just come on and rant. That's never been our podcasting strategy. No, never.

[00:29:25] Oh, an episode cannot pass without Glenn talking about James Arthur Ray or my boy, Tony Robbins. TR. Good old TR. I, well, at first, so when Jean and I were texting about what to talk about on our next episode, I, I looked through like the news alerts for Tony Robbins and, and, and I sent

[00:29:52] her this, this news story about, oh man, dozens hurt at a Tony Robbins event. You know, it's the firewalk. I'm like, he's doing it again. And people are hurt again. And like, I didn't think he was doing that anymore, but apparently he's doing it. And Jean's like, um, this article is 10 years old. Like, oh, right. Like 2016 was a hell of a year. Um, but we found, but, but, but there was a different article that was from, I, I double checked. It was from last month.

[00:30:20] Um, what it's all about is, so there was a men's mental health app called mental. Um, I had never heard of it. Um, and this article, so I'm, I'm reading from, from tech crunch. Um, this is when the, when the founders of a mental health app for men called mental saw that one feature, AI interactive audio was resonating wildly with their users.

[00:30:47] They knew they were onto something and that's, so just even that is kind of interesting because what we're talking about with AI interactive audio, I assume, like I was always interacting with rock, just typing and back and forth, but this is a mental health app. Um, AI interactive audio. I assume people are talking to it like a therapist. Mm-hmm. Um, and apparently loving it. Right.

[00:31:15] Um, so they decided to take that aspect of the app, the, the, the AI interactive audio. And by the way, kids, let's please keep in mind that, I mean, similar to Facebook, similar to Google, um, if something is free or low cost, that means you are the product, right? Mm-hmm. Yeah. That means that they are the, the harvesting that data and look, that's not always bad. Like it's, you know, it's like, I would actually argue that, that, you know, the, the amount

[00:31:45] of information, for example, about ourselves that we give to, I don't know, uh, a company like meta. I don't hate that meta then serves me up ads that I'm more interested in. Like who cares? Like I would rather, I guess, watch a tailored ad, but I don't, I don't care about that. But when it comes to like a mental health app using AI interactive audio, what do you think they're doing? What, what kind of data do you think they're harvesting? It's not your favorite color.

[00:32:15] It's not your favorite restaurant. So they can serve you up and add in your feed, actually. Like, like they're trying to get a clear snapshot of really intimate stuff that drives behavior, right? So that kind of gives me the ick a little bit. But, but it goes on to say this, so they were, they were developing this app that the, the, the mental app evolved into an app called the path.

[00:32:40] I mean, there's definitely, the path has been a thing in the self-help past. Has it not? I feel like that's gotta be somebody's. Somebody's brand. The tech, the path, like that's somewhere. Oh yeah. Either that or maybe Catholicism. It, it, it has kind of a, kind of a Scientology ring to it, actually. The path. Like when we were talking. This word, this term has triggered me. Has it triggered you, Gene? Yes.

[00:33:11] Oh boy. Anyway. So it said, so the idea for a new and hopefully, but here's the catch. I'm once again reading from TechCrunch. So the idea for a new and hopefully safer kind of AI therapy app was born, which they called the path. Um, then next paragraph, then famed author and motivational speaker, Tony Robbins, who so enamored with the startup that he scooched in as a co-founder.

[00:33:37] The path has now raised 14.3 million in seed funding, um, led by, um, uh, a group where, where Tony Robbins is a partner. Um, yada, yada, yada, yada. The, so apparently the CEO is a PhD psychologist. Well, he has a PhD in psychology. I don't know if he's, if he's a licensed psychologist. Right. He says, yada, yada, yada.

[00:34:07] He sees LLMs and AIs as the bridge spanning that gap. Mm-hmm. I.e., you know, the gap for the other people. It's the, it's the old classic. Lots of people need therapy. There aren't enough therapists. Therapy is not affordable, et cetera, et cetera. Yeah. He sees AI as the bridge spanning that gap. Quote, what's exciting and game changing is that for the first time in my career, I've seen that there's actually this possibility for every single person to have the personalized sort of access and care that they need to really get the help.

[00:34:36] Let's think about that for a sec. What Gina and I just spent 30 minutes talking about was that AI can do a very convincing impression of giving you personalized engagement, but it actually really isn't. It's, it's a certain amount of pattern identification. Maybe, maybe. But even that is really limited.

[00:35:03] So when we go out here and, and make that the, the main marketing shtick that, you know, look like this is, you know, like you don't need a human to understand a human. You need a, it's pattern recognition. It's not actually giving you individualized help, but it's saying that it will. In fact, we have the CEO specifically saying that it will. I'm like, yeah. Um. And all of this stuff that AI is trained on.

[00:35:31] Like we said, it's like a hot take machine. You know, it's, it's pulling from the cesspit that is the internet. So there's all kinds of crap on the internet. Some of it's great and some of it is garbage. And I know that, yeah, I, I couldn't make an AI myself. I don't have that kind of training to do that.

[00:35:56] But so I know that, you know, certain types of content is, uh, weighted over others, but it's still raises a lot of alarm bells that, um, we're basing these machines on this space. That is, it's very flawed, very flawed space. And then we're going to ask it for, yeah, this, this kind of very, uh, personal advice.

[00:36:23] The, the, the CEO goes on to say that, um, the paths specially trained AI model has scored a 90, a 95 on the mental health safety AI benchmark, something called Vera MH. I'm not familiar with it. This compares to a top score of 65 for the consumer bots. So what they're talking about here is a benchmark, a litmus test somehow that, that evaluates the safety of mental health oriented AI products and services.

[00:36:55] Um, I would be super curious about what all that entails because the things that make AI therapy less safe, I don't know how you can inherently address that in the model itself, you know, with the exception of, you know, with the, there are, with every AI model, there are some guardrails around what it will and won't talk with you about. Mm-hmm.

[00:37:18] Um, however, the, the fundamental issue with the, the lack of safety is much like every self-help product and service. It's the lack of oversight, right? It's the lack of regulation. Yeah. And again, I'm not Mr. Government regulation solves everything. I'm really not, but there's a reason why you regulate the healthcare professions.

[00:37:38] Like there's a reason why I, as a psychologist, just can't go out in public and say whatever I want, you know, under the, uh, with the credibility of my degree and my license, right? Like I have to restrict myself to what is appropriate for psychologists to say. And I'm not sure how you regulate that realistically, um, with, with a chat bot, nor am I particularly sure, like, look, we're not talking about licensed therapists.

[00:38:06] When a licensed therapist says something that causes harm, there's recourse. Like if I say something that causes harm, like they can, you know, eventually they can bring me up on ethics violations. They can take away my license. Ultimately, what are you going to take away from the AI? Yeah. Um, and it is a quote unquote, it's, it's meant to challenge you. It's not just meant to agree with you. He says, this is the CEO saying, in fact, he says the apps model is post trained from open source models. So it doesn't use the major consumer models of all.

[00:38:36] Mm-hmm. It's not simply a wrapper over them. Mm-hmm. The path, which lets users get this, choose from 11 virtual AI therapists and customize their preferences for directness and other details is currently free as it gains users. Eventually the startup plans to charge $40 a month. Mm-hmm. Um, we got problems all over, which is not to say that I don't get it.

[00:39:04] Like, of course I get the, you know, like I, I get, and I even on some level admire what this company says they're trying to do. Mm-hmm. My question is going back to the title of this article, the path founded by Tony Robbins and blah, blah, blah, hopes to offer safer AI therapy. Gene.

[00:39:27] Look, I'm a, I'm a tortured Tony Robbins fanboy, but even I can say what, what qualifications does Tony Robbins have to talk about safe anything? Yeah. Safe therapy, let alone therapy. He has no training in therapy. He's really like leaning into this mental health guy branding. He is.

[00:39:49] He's, and we talked about this on a previous episode, how he was named to this advisory council, uh, for, about mental health. Um, yeah, it's, uh, which honestly there's probably some connection between these two things happening. I also find it interesting him becoming like, um, an investor, like, uh, you know, getting in on the ground of this kind of thing, like with a startup.

[00:40:18] It seems like there's just this sense that that's where the money, but then also kind of the, the fame comes from these days is being in on the ground of something like this. Um, so yeah, he's maybe not doing the firework at big, you know, events with thousands of people, but maybe like it's cooler these days to be, you know, part of some flashy startup to get your name in there as a founder.

[00:40:48] Um, yeah. Who's totally unqualified to, to be doing that, but. Well, it's, it's so interesting that, you know, people equate Tony Robbins with, uh, personal transformation. And then they equate personal transformation with mental health.

[00:41:11] And therefore, you know, Tony Robbins kind of through, through a string of associations gets, gets paired with mental health. And, and look, I'm not one to talk. I, why did I become a psychologist? At least in part, because I read Tony Robbins books. So, and I, I get that, that connection.

[00:41:29] However, I think it really behooves the mental health community to get very clear on the fact that, you know, man, mental health is not entirely about personal transformation. Like, I think of mental health as kind of the ground floor of personal transformation.

[00:41:49] So, a criticism that I've long had of the trauma treatment industrial complex, of which I am reluctantly part of, um, is that, so the, the model for treating complex trauma disorders is, there's a period of it. So, there's stage one, that's all about safety and stabilization. And there's, there's, there's stage two, that's all about processing trauma memories and trauma feelings. And then there's this third stage, it's called life reintegration.

[00:42:18] And I was trained by some of the best, some of the most famous trauma, uh, therapists in the world. And let me tell you, they're very into stage one. They really like that safety and stabilization stage. Stage two, trauma processing. Nobody really knows what the hell that means. Like, I have my version of it, but I had to come up with that on my own. And stage three, no one talks about life reintegration.

[00:42:46] Like, like what makes a life worth living? Like what, what makes, why are we bothering to do all this trauma work in the first place? And so, I've long had this criticism of the trauma treatment. I was like, man, we talk a lot about, um, you know, these terrible things from the past. We don't really talk a lot about how to really create a life worth living. Like, man, if I could just wipe your system of, of all your trauma and, and, and your trauma responses and, and all that pain. And you had to start over right now.

[00:43:16] Mm-hmm. Are you equipped with goal setting strategies? Are you equipped with financial literacy? Are you equipped with, you know, you know, the, the, the, the resources and the tools to like go forward and, and to create a life worth living? And this is where I think a lot of people turn to folks like Tony Robbins or, or, or other gurus. Cause that's what they promise. They don't actually, well, James Arthur Ray went through a period where he couldn't shut up about trauma, but. Um, but they, they very explicitly promise this.

[00:43:45] They'll often say, this is the thing. If you go to talk to a therapist, they're just going to keep talking about the past. I'm going to give you tools for your future, the future that you want to build. Like that is an explicit statement that is often made in the self-help industry. You bet. You bet. It's a, it's a, it's a hook. It's a shtick.

[00:44:06] And like you're saying there, there is some validity to that criticism, um, because sometimes, yeah, you can get bogged down in the, in the past and you need to maybe need a little more to push you ahead. But. So all of which is to say, I can't believe that what we've been talking about tonight, um, either, you know, the, the, the further integration of AI into cell.

[00:44:36] Self-help spaces and the industry. And by the way, like as a, I often say on this podcast that I'm a self-help person as a self-help person, as somebody who looks at that, like a new book or a new theory or whatever. And my first instinct is like, Hmm, maybe I'm interested. Right. Mm-hmm. I can tell you that the idea of, of AI self-help products and services is cat. Nope. I can tell you.

[00:45:04] And that has nothing to do with anybody being stupid. That, that has to do with us being hopeful. Um, and, and look, it has long been one of the frustrations for folks who use self-help products and services that, you know, man, like this author has some good ideas. I really wish I could get a more interactive experience. Look, that's why this, the, the entire LGAT industry took off. Right. Mm-hmm. Because folks were reading these self-help books. Like there was a time, believe it or not, there was a time when self-help was mostly just books.

[00:45:33] Mm-hmm. And what happened was, you know, gurus figured out that if they offered some sort of access to them and, and, and LGATs or seminars or, or retreats or like, like whatever it is, that people were hungry for that kind of thing. Mm-hmm. Like, yeah, man, I want the person who wrote this book to give me some, some individualized feedback.

[00:45:54] Now, just at the point in the industry's history where that kind of thing has come in for some scrutiny, thanks largely to us, honestly, this other vein is, is, is starting to develop. It's like, well, maybe you don't need to come to a seminar. Mm-hmm. Because Tony Robbins, as, as we find out speaking of Tony Robbins, like, you know, he was first on the ball with the AI. You can get AI Tony Robbins. You remember this? Yeah. Mm-hmm.

[00:46:25] Anyway, I, I think that because it is so seductive and because it, it scratches a very specific itch. And I get it, man. Like, I get it. Like, if, if, if, if they figure this out. By the way, I'm also mindful of the fact that we are at the beginning of this AI thing. Mm-hmm. Mm-hmm. And who's to say in five, 10, 20, what, however many years, who's to say that, you know, those, those quirks of grok that made interacting with a kind of, gave, gave some friction for me.

[00:46:55] Like, who's to say they won't iron that out? Mm-hmm. Yeah. Yeah. We need to be mindful and then this needs to be a part of Sikh's mission. This is one of the reasons why, why we chose to talk about it on the podcast tonight. Because instead of calling a board meeting and just saying to my board, hey, I think we should talk about this. I figured I'd just put it out on the podcast. Mm-hmm. This is, this is how we, we communicate now. Mm-hmm. Yeah. Anyway. Yeah.

[00:47:20] Well, you know, my hope is that, especially with these, these types of tools that are, that are definitely very, very clearly getting into the mental health space, that the professional mental health community will not just bristle against it, but will recognize that these things aren't going to go away. Um, so there, there needs to be some involvement from the professionals to say, hey, you know what?

[00:47:50] We're regulated. Let's, yeah, okay, these tools, like, they're interesting. People want them. There's a demand for this. But it needs to have some guardrails just like any of the rest of us do. Um, because, yeah, it's getting very muddy. I mean, I think we'll see, start seeing the same things just in, in the actual, like the rest of the healthcare industry as well as, as AI. There's more and more AI integration.

[00:48:21] And, and just saying, yeah, like we can't just let the technology go and see what happens. Let's, let's, let's do it thoughtfully because we're actually talking about, you know, this isn't just, uh, asking the AI what wine will go well with the dinner that I'm making. This is asking, uh, for a lot, a lot more serious things. So yeah. Yeah. Yeah.

[00:48:50] So if there was an AI version of Jenny Brown, what would it say most often? Hmm. That's an interesting question. I don't know. What do you think? I was trying to think. Yeah. Like, like if we, cause I told you, I think I told you on the episode that we did previously, I told you about AI Dr. Doyle, right?

[00:49:17] And how AI Dr. Doyle was, it had been fed all of, or a lot of whatever my tweets and stuff. And so if you talk to Dr. AI Dr. Doyle about, about trauma, you got an okay conversation. Stuff that I would probably say more or less. But then you started talking to AI Dr. Doyle about his favorite beetle and he just went off the rails. My favorite animal or whatever. It said Ringo, didn't it? It did.

[00:49:47] It said Ringo. So I often think about this. I often think about like, you know, man, if we had AI versions of like, if I told an AI to role play Jenny Brown, what kinds of things? I don't know what kinds of things it would say. I can, I can hear the, the, the voice though. By the way, my future stepkids love the way that your mom talks. Have I told you this? I told you this. But I hear this a lot from lots of people.

[00:50:15] I, cause I never, I don't, I hardly register it. I do more now that I'm away. But, uh, yeah, my mom, you know, she's a good girl from Queens. So. My soon to be stepkids who had, who were raised here in Southeast Texas and have never really been outside the state. Tuned in to some event we were doing Seek Week or something. And the 12 year old, especially like, like honed in on your mom's, like the, the way that she speaks.

[00:50:45] So it was, you guys don't know Jenny Brown, Jean's mom. Yeah. Like you can tell she's a Northeastern. Person. And, and, uh, and, and she's like, wow. She's like, she talks like Lois from family guy. She does. She does. Oh my God. That's funny. Well, isn't family guy there, they're on long Island, right? They're in Rhode Island. Oh, Rhode Island. That's right. Got the islands.

[00:51:14] This is what happens when we're off skirt late night. I mix up my islands. Um, yeah. Well, yeah. Sometimes, sometimes that, uh, New England mixes a little into the. True story. Yeah. All right, gang. Thank you for tuning in to the seek safely podcast. Um, if you liked our conversation tonight, if you like our podcast, please feel free to hit up your, your podcast.

[00:51:43] Uh, the catcher of choice and give us a glowing five-star review. Um, if you didn't like our conversation here, forget I said anything and we shall return to talk about more gripping issues in the, in, in the self-help space. They're gripping. They're going to grip us and we're going to grip them right back. We're going to grab them by the personal empowerment. There you go. It's late guys. It's late. All right.

[00:52:13] Thank you. Thank you, Dr. Glenn. Uh, thank you everyone for listening and we'll see you next time. Ciao, ciao. Bye-bye. Thanks for listening to this episode. We hope that you have found it enlightening and we'd be so, so grateful if you'd share it with the seekers in your life. We all know at least one, right? Until our next episode, you can find us on Twitter, Instagram, and Facebook at SeekSafely.

[00:52:41] Connect with Dr. Glenn Patrick Doyle at Dr. Doyle Says and me, Jean, at Jean C. Brown on Twitter. Feel free to send us an email, info at SeekSafely.org. To support SeekSafely, you can make a secure donation on our website, SeekSafely.org slash donate. The SeekSafely podcast is produced by Citizens of Sound.