Latest / Brilliant Minds? / AI in therapy?

Transcript

AI in therapy? · Brilliant Minds?

• Olivia: Hello everyone, welcome back to a little bit of us, Brilliant Minds, Brains, Banter Being Human. I'm Olivia, and this is my loveliest friend. This podcast is for entertainment, informational, educational purposes only. Although the hosts are licensed therapists, this podcast is not a substitute for professional mental health services. Listening to this podcast does not create a therapist-client relationship. Andrea: Andrea! Olivia: Woo woo. today. Andrea: I don't know if this is a little bit of us. I feel like this is a lot a bit of us. If you are experiencing mental health concerns, please seek help from a qualified mental health professional in your area. If you are in crisis, call or text a suicide prevention line in your country immediately. In the U.S., you can call or text 988 or use online chat at 988lifeline.org. Olivia: It is, it's a lot of exposure. Yeah, for sure. We are going to be talking a little bit today about, although I kind of like surprised Andrea with the topic a little bit, is we're going to be talking a little bit about mental health and AI and kind of our perspectives on it and what we've experienced and all that good stuff. It's quite like an up and coming topic for sure. Andrea: Ha ha ha ha! Olivia: And our little disclaimer as always, we do do this for a living. However, right now are just trying to live. So do not come at us with your problems. ⁓ Andrea: Somebody let us run a business. So, this professionally. Olivia: Did somebody or did they just be like, okay, you can have a license. Andrea: You know, sometimes I wonder what actually happens at the Department of Licensing Olivia: I think everyone kind of wonders that. I'm pretty sure people that work there probably wonder that too. ⁓ Andrea: That's very true. They probably are switching to AI at this point. Olivia: Maybe they should, yeah. Maybe their processing times will get better. Ooh, DOH burn. Okay, so ⁓ my, I think the first question out of the gate, ⁓ is, do you think AI will replace your job? Andrea: My gut says no. And think because... See, I don't know. I don't know if I have like a legit answer to this because my gut just says no. Like it just doesn't, advice that I'm seeing people talk to AI about and the things that it's saying is just ridiculous. Like I don't think it can ever get to the point where it's actually us. Like look at us. Olivia: Can't be regenerated. Can't be copied. Can't be borrowed. Andrea: No! No! Olivia: Yeah, I would agree. I would agree. I think... I don't know. really shouldn't have to the caveat. This is my perspective. This is my bias, because this is literally the point of this flipping podcast is my bias and perspective. ⁓ I think that AI will ultimately as much as it can learn and try, it will never feel truly relational in the biological sense that we experience a relationship and that like. Andrea: Yeah, exactly. Correct. Mm-hmm. Yes. Olivia: biological connection. ⁓ And so I think it could get good at giving feedback or input or synthesizing information or providing psychoeducation, like aspects of our role. But the core of what I think our role is, just like relational human connection, it cannot copy, I don't think. Andrea: Mm-hmm. Mm-hmm. Yeah. Yeah. No, I think you touched on a really great thing is the biological piece that comes with relationships across the board, right? Like we actually do have brain research and neurology that proves that physical relationships, seeing an actual person impacts the way that we feel like it helps us regulate. helps us identify emotions, like so many things like Olivia: Yeah. Andrea: The first thing that comes to my mind in this topic is the idea of mirror neurons. In therapy school, we touched on mirror neurons so much, especially when working in person. And there's arguments out there that telehealth doesn't allow for that to happen, but I actually very much disagree with that. I think it still works with telehealth and over video. But basically the idea of mirror neurons is we watch someone... ⁓ Olivia: Mm. Andrea: our brain kind of fires in a similar way with regards to emotion or the way that their body is reacting or things like that. so AI can't emulate that. There's no way. That is a biological function that our brain has. Olivia: Yeah, had a supervisee, often call themselves a mammal, which I like, I just, I just found funny because like, normally people just get like human person, whatever. And so they'd be like, I'm just a fellow mammal with all the other mammals in the world. And like, I bring it up because it, to me, it relates to what you're talking about this idea that Like we are mammals. there is this like biological aspect that AI just like cannot produce in the same way, just like for any other like actual animal. Like AI can't, like parent or like mother, like, I don't know, like a chimpanzee or an elephant or whatever. ⁓ I'm going to save this for the rabbit trail. Andrea: Mm-hmm. Mm-hmm. Yeah. Yeah. Olivia: Okay, I need to save that. It was about giraffes giving birth. Okay, that needs to get saved. Okay. Logging that away. Logging that away. Logging that away. ⁓ Okay, so I think it would also be interesting to talk maybe a little bit about where we have seen AI in our field up to this point, maybe for people that don't know, or maybe for people that are like curious about how AI is getting used in mental health. Andrea: Writing giraffes. Olivia: And at least how I've most commonly seen it is ⁓ a lot of people have probably already heard about this or maybe even seen this when it comes to like going to the doctor, because I know they're rolling it out at my doctor's office, ⁓ ⁓ will listen to sessions ⁓ generate clinical notes for But was curious, do you know of any other way that Andrea: Yeah. Olivia: AI is getting used in our field. Because I was trying to think about this and I can't think of another way it's really getting used. Andrea: You know what, that's really interesting. If this topic hadn't been sprung on me last minute, I may have done some research. I'm just kidding. Olivia: I would never spring anything on you. Who was the heathen that did that? Wow. Andrea: Um, you know, I think that, I think that that is the primary way that our field uses it. I think there are two areas of AI and mental health that are very different that I think are important to address at some point too, which is like this idea of we can use it as a tool, doctors, we can use it in medical care, all of that. But then there is like the opposite of consumers using AI for mental health support. Um, which is a totally different pathway. And in this episode, you'll probably see our opinions differ very much on both. I don't think there's anything totally wrong with using it for documentation and notes and things like that necessarily. It's not something that I feel comfortable doing or safe doing personally, but. I think a lot of the reason for that for me comes from the fact that I do my documentation very quickly already and I don't feel that it's necessary. And I also have seen so many like data breaches and hacks over the like system over the last 10 years or whatever that I'm like, what ⁓ if of recorded sessions just happens to be part of a data breach? don't, can't, I feel like there is a risk in that. Olivia: Okay. ⁓ yeah. Andrea: for myself. Olivia: Absolutely. I know we've talked about this not recently. ⁓ I there's benefits and drawbacks of specifically having AI support for documentation. ⁓ think more drawbacks for me personally than I would say there are benefits. I think the benefit that I see ⁓ is of clinician, not most. Probably one of the most significant issues I see with clinicians in the field is around documentation terms of either consistency, efficiency, or timeliness. Now do not ask me why it's such a barrier for people. at the end of the day, you wonder the same about documentation. We're both really efficient. We're both really quick. We do our stuff. We dot our I's, we cross our T's, right? So it's really hard for me to understand or put myself in the shoes of somebody that really struggles with documentation. I can understand it from some perspectives, but like I have not lived that, so I don't get it in the same way. Andrea: Yeah. Olivia: but that is a really significant barrier for people. And if despite like multiple supports and multiple attempts and structure and accountability and all these things, if a clinician is still really struggling with documentation, would be a situation in which I would maybe say, actually, I think AI ⁓ support for this would be helpful because then it would be done. It's taking off some of that burden. But I definitely wouldn't use it as a plan A for a clinician that's struggling with documentation. ⁓ Andrea: Correct. Yeah. Olivia: Because there's like a strong chance that actually they're just seeing way too many clients and they're burning out and they don't have time for documentation. that's like, like, that's not when we should be using AI. Andrea: Hmm. Yeah, mean, is a whole other topic that we could talk about, you know, and might be really interesting for listeners to to kind of know the background of what we do documentation wise. But you and I both take a very like teaching approach when it comes to working with people on their documentation, right, and like, figuring out what works and what support systems are in there. But you and I are also very aware that like, that can just be a workplace accommodation if it becomes too difficult to do your job. Olivia: Mm. Andrea: or if it becomes debilitating or hard. That can be a thing. You're saying not a plan A. I think that we all need to practice and get the skill of that documentation and using clinical language. And then on the other hand too, I actually was talking to someone today who uses AI NoteTaker who I know does a really good job with documentation, because I was their supervisor. And I know that their documentation is great. And they brought up a really good point that like it Olivia: Yeah. Andrea: utilizes clinical language that they forget about all the time. And so it makes their notes like a little bit more insurance friendly, which is interesting because like that's how I was raised in the field is how to make them insurance friendly. So that's just part of my system. So it's interesting to hear that feedback that like AI is pulling different clinical language that they might not like think about. Olivia: Mm-hmm. Andrea: but I think it touches a lot of feelings that I have about supervision and about schooling and how that learning is happening for me. right. But I did. Olivia: Training and yeah, yeah, like professional development essentially. Yeah. Andrea: Exactly, exactly. And I think like, if you're efficient enough already, like I think it takes me maybe a minute to write a note. If you're efficient, I don't need to use it. Like it would take me just as long to review the note and whatever. And, then it's like lingering in the ether somewhere and I don't like that. But it's, it's a weird one. And like even ⁓ seen my doctor use it and I'm oddly okay with that, but I think that's just like a different setting. I don't know, it feels different when it comes to what I do. I don't know. Olivia: Yeah, it's interesting. the two main reasons why I would not use an AI support on notes. A lot of times it's called AI note taker. So if you hear me say that, that's what I'm referring to. One is along the lines as you, which is a privacy piece, which is I really cannot ensure that a recording of a client session is actually deleted. and confidentiality that privacy is incredibly important. I mean, not just in the field, obviously, that I think that kind of goes without saying, but like for me personally too, right? Like that privacy is really important. I really ensure that that is happening. So I'm not, that doesn't sit well with me. The other piece, which I know I have mentioned before with you ⁓ is see a lot of people turning to AI for advice and like, life coaching or whatever the case may be. And because it's AI and it learns and it grows, if it listens to me and records sessions, it's going to learn how to be a better quote unquote like therapist. Andrea: So this was part of the conversation I had with this person today. And I don't think we can guarantee that with all systems, but supposedly the AI system, let's talk about simple practice for example. The AI note taker within simple practice does claim that the only learning that it does is within your own. system. So the only learning it does is how you edit the note that it doesn't learn or hold anything into the cloud. And it's like, that's like the statement that they give ⁓ system. But like I, but the trust of that feels weird. Olivia: Yes, but like... Yes, but that's exactly what I'm saying because I'm sure in the fine print is also, ⁓ we only keep these like recorded sessions for this amount of time and then it gets deleted forever, you know, whatever. It's like... Andrea: Yeah. Olivia: but like you can't insure that. Andrea: Right, right. Like I would need to see the legitimate proof that it is gone and wiped and disappeared forever and ever and ever and nothing is gone forever in the world that we live in. Olivia: Right. Whereas in like a medical setting, this might sound kind of bad. It feels like I'm almost like wanting to dodge accountability, which is not it. There is so much malpractice that's happening and so much of patients being dismissed, things not getting recorded when they should. And also the idea that doctors have way too many patients that they're seeing and they can't keep up with documentation anyways. Andrea: Mm-hmm. Olivia: So to me, actually, I think it's a benefit to have AI because there might be a little bit more like accountability in terms of what's getting documented for both the patient and the doctor. So I say that not because, cause you know, my doctor uses AI and she asked me every time we get into an appointment and I'm like, that's fine. Um, cause I can tell that girl's behind the documentation. I was like, don't worry, girl, I'll do you solid. You can use AI. It's okay. Um, you know, like, I don't mind. Like, Andrea: Yeah. Mm-hmm. Olivia: Anything that's in my medical chart, I've probably already blabbed to like anyone that would listen anyway. So like it's fine. Right? Like there seems to be like, I'm not as concerned about my privacy when it comes to my medical record, but I am more concerned about it when it comes to my mental health record. And technically they are two different things. But yeah, so it's like, I say that not because I, I don't want to be held accountable as a therapist. If I'm like doing crappy things, like I absolutely be held accountable, but the medical system feels different to me because it's a lot. Andrea: Yeah. Same. Yeah. Right. Mm-hmm. Yeah. Olivia: more transactional also because there's like a lot of malpractice that happens. ⁓ Andrea: Yeah. Totally, and I think within our field, like we said at the beginning, it is so relational and there's so much nuance in a conversation that even if AI is learning from our sessions, if that is the case, if it's doing that, then it's probably not fully comprehending the context and the relational aspect and the biological aspect of that, and it's not gonna relay it correctly. Right? Like if somebody goes on to Gemini or chat GBT and they're like, help me figure out if I should break up with this person, which people do all the time. Like maybe you've made your chat GBT to be sassy and it says something and then you take it seriously when you don't need to take it serious. When it's like not intended to be serious. Like it's just, it's sketchy or it's like taking information that it's learning from therapists and spitting it back. And maybe it's like, Olivia: Yeah. and Andrea: something that's inappropriately said, right? Like, I don't know. It's not. very risky and there is no ethics or code of ethics or guidelines on AI right now. It makes me nervous. I'm efficient, I can do it myself. Olivia: yeah, ⁓ I mean like, let's be honest, this is one of those situations where we won't have laws around it until there's a legal case. Andrea: Right. Which like. Olivia: and then it sets a precedent, which like, I'm not going to get into like how crappy of a system that but what I will say, like as a clinician, if I was to use an AI note taker, and since it's so new, I'm kind of like signing myself up to be a part of the group of potential clinicians where there could be a legal case. ⁓ Andrea: Right, right. Yeah. Yeah, for sure. Olivia: about my client and, and it could become all about like, well, what was actually happening in these sessions and did the AI note taker this or, it, you know what I'm saying? Like I could, I could be the first legal example of how there's regulation in the field about AI. Does that make sense? Andrea: Yeah. Yeah, no, totally, totally makes sense. I think it also, you know, brings up an additional point to this note taker is like, if you're not reviewing what it's saying, then you're just like setting yourself up for higher risk with that too. Olivia: Yeah, but like how do you how do you prove that you've reviewed it? Andrea: Oh, I'm just saying like, if you're just using the note taker and signing, you're not reading it, you're not looking it over, this AI note taker could be writing the craziest bullshit and then you could get those notes subpoenaed and be like, wait, what the fuck? Like, what did I do? Right? And like, not saying you have to prove it, but like, you could just get yourself into a whole mess of trouble. And to me, like, why would I, like, why am I using an AI note taker if I would just have to review it? Like, that's so. Olivia: Yeah. Yeah. Andrea: That defeats the purpose to me. Olivia: Well, yeah, yeah, absolutely. Now that we've talked about this for a little bit, it reminded me, ⁓ from somebody in the field that a practice was using an AI auditor. for client files, which I thought was kind of interesting. And I think it's interesting for a variety of reasons. I think as an auditing tool, I think it could be really helpful. But then you still land in this conversation of does client need to consent to client file being audited by AI? Andrea: I mean, yeah, I would think so. Olivia: I mean, I would think so too, but generally, like a practice, like a group practice, they should be having some level of internal audit system that the client wouldn't inherently have to consent to. Andrea: Mm-hmm. Right, right. You know, that's an interesting thought. And now my brain is jumping to insurance. Like, do insurance companies use AI to audit people? Olivia: Like it. I bet they probably will. At least I think. I would think. It'd be more efficient and then they could be doing more of them. Probably. Andrea: Yeah, I'm- ⁓ I don't do more. Olivia: Well, I mean, I've lived through a few like insurance audits and then there's never been any issue. I've never run into any issue. So. Andrea: No, No, same. Yeah. Yeah, I think, and usually it's state insurance that's running audits. It's very rare that private insurance companies are running audits. Really? I never, not like a true one. I've had like somebody's file get requested for proof of diagnosis or something, but that's about it. Olivia: ⁓ I've lived- I've gone through actually a couple of private insurance audits. Yeah. Yeah, I've had it for two of my clients and then I also have had it for two people that I supervised. They're just like random insurance audits. like an insurance was like, okay, pull these like 10 random client files. Andrea: ⁓ But not in private. Mmm. So interesting. That's so interesting. No, I haven't had that in since being in group or private practice, just Medicaid audits, which are Medicaid audits are way more intense and like they come in, they will sit and thumb through every single file. I worked at a practice and we still had paper files when I started there, ⁓ was insane. Olivia: Yeah. ⁓ yeah, yeah they are. Andrea: they had an audit right when I started and the auditor like literally came to our office because we had paper files and was flipping through everything. Olivia: That's wild. Yeah, I did in my nonprofit job, we would have yearly county audits for all of our contracts, because almost all of our funding was contract-based or grant funded, but only a few grants. And so went through a ton of county audits and that was like, well, it was like, it was interesting because it was like, Andrea: Yeah. Yeah. Olivia: There needed to be paper copies of everything and like a paper client file that they could thumb through, but there also had to be an electronic version that had to match. That was the worst. They had to have both and that they had to like match. Anyways, ⁓ they would look through every single one. Cause like the programs were small enough where they could look through every single client file. like it was, it was a, it was intense. Like it wasn't maybe as intense as like the Medicaid cause there's a lot more stuff that they look through. ⁓ Andrea: Yeah. Yeah. Yeah. Olivia: but the intensity of having to have all the paper copy stuff and then the electronic stuff the end of match. That's a lot to check through. Andrea: Yeah. Mm-hmm. Yeah. I think AI could be super useful in those types of, excuse me, be useful in those types of settings. But again, I just keep coming back to this idea of like, I have no trust. Olivia: Yeah, well, it's curious because in an instance where an insurance company is going to use AI as an auditor tool, they probably wouldn't be asking their people consent. They would probably just be like, hey, just so you know, update policy update, you know. ⁓ Yeah, which is interesting. mean, again, like I think it could be a useful tool, but it does kind of like beg the question of, then client should be consenting to. Andrea: Right. Right, right. You know, and then a longer term question I have for that is like, at what point is AI gonna be in everything? Because it kind of is starting to get there. like, just in general, right? Like, ⁓ know I've seen like AI is used in this thing, but it doesn't ask you, it just says it is. And I'm like, okay, like. Olivia: directly. Yeah. Mmm. Andrea: Great. I mean, it's nothing private, but it's just you know? Olivia: Yeah. It's interesting. feel like there's both things happening where like AI is getting pushed and rolled out and used a lot. But then I also see a lot of people being like very against it and very anti AI. It's an interesting mix. I was trying to think about are there areas in the field where AI ⁓ be applicable and useful in our field? Not just like where they are currently. ⁓ Andrea: Mm-hmm. Yeah. Mm. Yeah. Olivia: like where it's currently getting used. And to me, I think probably the most useful applicable, if you could ensure privacy, think the most useful way that we could be having AI support the field is doing our billing. Andrea: Yeah. Olivia: Like, AI should be doing our billing. Andrea: Ugh, that'd be so nice. Olivia: Like AI should be doing medical billing. Andrea: Yeah. Definitely. Olivia: Like that would like to me, that's like a no brainer, like, duh. Like, why shouldn't it be doing that? Like that's like, that's a perfect job for AI is like billing. It's like the numbers essentially. Andrea: Just like. Yeah. Yeah. It also, yeah, and we could get into billing at some point down the line too, but like billing oddly, like us submitting our billing and our claims to insurance, that doesn't take that long. But like, it'd be nice if it just did it by itself, right? But the piece that's just like, you could probably use AI to pay us a little faster and like give us our remittance a little faster, because why the heck is that taking so long? Olivia: Well, yeah, that's what I'm saying. Like if, kind of like the whole system had AI support, like everyone around, think would kind of benefit because like the insurance companies could be processing claims faster. They could be like dealing with disputed claims faster. could, you know what I mean? Like, I think everyone would benefit if AI was supporting billing. Like, and I'm not trying to like, I don't know. Andrea: Right. Right. Mm-hmm. Totally. I agree. Olivia: completely fire an entire field of people that like do coding and billing because like you're still gonna need to have a human look over things, clear things, make sure like whatever. Andrea: Right. Right. Yeah. quality assurance, ⁓ Olivia: Got it, yes. But to me, that feels like a no-brainer. ⁓ Andrea: Mm-hmm. Yeah, I mean, I will admit there are some really great uses for AI. Like, you and I have used it in the past for things to help support build our business. And it has been really supportive and helpful. And it can be such a great tool for neurodivergent people, too. Like, You know, I know people that use it to help them create a grocery list and that is a very simple task that it's doing. like, I'll admit I've used it and because I can say like, I'm allergic to these things, I can't eat these things, I'm trying to adhere to this type of like nutritional support. Can you create me a grocery list for today? And I can choose to follow it or not, but it is like in those ways it's been so helpful. Olivia: Yeah, I feel like we've talked about this on the podcast before, there's something, assuming it's probably in order to virgin related. Haven't quite like figured out how it boils down, but ⁓ so much easier for my brain to edit something than it is to create something from scratch. And so that's where a lot of times I think AI is helpful. And like where we have used AI in, I know, just like. Andrea: Mm. Same. Olivia: setting up business forms and things like that is we know what we need, we know generally what it needs to include, but looking at a blank page and trying to write a disclosure form, for instance, which is like probably the most critical form that you need to have as a clinician to have clients sign to consent to therapy. It's impossible to know where to even start to write it and how to work, all these things. So AI is really great at that. So I definitely think it's a useful tool. And that was gonna be kind of one of my questions is where do you use AI in your life and is that helpful for you? And I think it's the outsourcing of creativity in that way. I don't know if creativity is best way. Andrea: Yeah. It's like a, I wanna say executive function, but it's not quite executive function. Olivia: Yeah, exactly. it's like, um, it's like, we're saying like, plan me. Andrea: initiation. Mm-hmm. Yeah. Olivia: Like it's outsourcing the like startup. ⁓ Andrea: of my clients the other day said that it helps them get their like analysis paralysis. So it's like ⁓ kind that idea of like ⁓ the of getting started on something like AI can really help create a structure around that and be like here's where you start or like give you the rough of what you need to say in your form or your email or whatever and then you can kind of go from there and edit it ⁓ and ⁓ Olivia: Hmm. Andrea: And it's helpful to kind of get past that analysis paralysis of like, where do I start? What do I do? How do I get this going? Olivia: Okay. Yeah, yeah. My brother uses AI a lot and I kind of mock him for it. because ⁓ don't use it really that much. We used it, absolutely. I don't think it's like, it's not like I'm actively trying to avoid it or anything like that. I just don't, I don't know. There's not many times where I go to use it or think to use it. ⁓ Andrea: Yeah. Right. Olivia: But my brother was just telling me, because he's like, you're always using it for something. And I'm like, what are you on? Like, what? And he was like, yeah, I realized I could set up to be a DM and set up campaign for me. Andrea: DM is Dungeon Master for people who don't know D &D. Olivia: Okay, yeah, that's fair. And so he, yeah, he like set up a like D &D campaign with AI and he's like, oh yeah, like I kind of just like use AI almost like a game on my phone because it will like give him like choices and like it'll play out stories and like all this stuff and he'll have to like roll. Andrea: That's so smart. Yeah. Mmm, that's clever. Huh, that's really clever. Yeah. Olivia: Yeah. Yeah. And I'm like, ⁓ that's cool. Still probably wouldn't use it for that, but like, cool. Yeah. Okay. Like, but I swear he's always figuring out a new way to like, he, like planned a house that he's like nowhere close to building through AI. And I'm like, like, one time he came over to my house. He said, what kind of flooring is this? And I was like, who are you? What ⁓ of flooring? Andrea: Right, right. Whoo. Olivia: What gene pool do you think we come from where I know what flooring this, like what kind of flooring is it? ⁓ I just like, I'm just learning about flooring. I'm like, okay, well, good for you. Andrea: Now, now how? Right? Now maybe this is a good time to transition into the other side of AI and mental health. which is the use of AI outside of a therapist, the use of AI as mental health support. Now this one. Olivia: see. Andrea: This one is dicey. mean, like we just touched on, I think there are some benefits to supporting people with executive functioning difficulties with ADHD and initiating executive functions or analysis paralysis or getting, making decisions, like helping make decisions. You know, I've had people use it for create a workout plan when they feeling stuck, you know, to help them navigate certain things. But I feel real icky. when we start to get into things like processing emotions, interactions, legal is another one. Olivia: that's spicy. Andrea: Yeah, yeah. Like we are starting to see issues come up with mental health and AI outside of the therapy space. Meaning we are starting to see people hurt people because of AI. We are starting to see people develop full blown, I'm putting quotes around romantic, air quotes around romantic. relationships and sexual intimate relationships with AI. And the Psychological Association is looking at eventually adding AI psychosis to our diagnostic criteria. Olivia: care. It's kind of inferred in the list that you mentioned, but it's also increasing suicide risk. Andrea: 100%. Olivia: So it's not just like safety of people hurting other people. It's also like just people safety. Andrea: Mm hmm. Yeah. Olivia: Which is, yeah, really scary. Andrea: interesting to me too because I feel like AI could have a feature in it where if somebody implied a lack of safety that it could contact authorities. Like it's tied to your IP address. Olivia: Or be like, you need to call, yeah, like, or you need to call 911 or like, yeah. Andrea: Yeah, give you crisis resources. Yeah, like... Olivia: You'd like to think that. Andrea: this ⁓ where we can't trust it, right? Like this brings that piece back where we can't ensure that any of that information is safe. like, period, across the board, I feel like. Even if the fine says it, or if the contract says it, or if you're signing an AI, like there is still a part of it that like, there's gotta be a hole in everything. Olivia: Yeah. ⁓ Yeah, and this is another case of... It's really hard for me to put myself in the shoes of someone who is thinking through something, struggling with something, feeling stuck about something related and they for advice is AI. Like, ⁓ never in life ⁓ would I going through something challenging. Whatever it is. I mean, I don't even know. Parenting my child. I... ⁓ Andrea: Yeah. Olivia: navigating a new diagnosis. I mean, like, I can't even like anything, anything like challenging, like a very like human thing to go through, right? I can't imagine being like, I really need advice, I feel stuck. I'm going to go to AI and ask them what they would do. Like, maybe as like a joke to see what it would come up with, but not like, for real. So for people to go to AI and be asking them for like life advice is really wild to me because that I just like Andrea: Mm-hmm. Mm-hmm. Olivia: My brain doesn't work that way and I wouldn't do that so I don't understand why. But it does seem to really alluring for people. Andrea: Right. Yeah, know, I think of two things. It ⁓ accessible. Which makes me cringe a little bit, makes me really sad. ⁓ yeah. And you know, it's accessible. People quite understand their insurance ⁓ benefits. don't ⁓ understand different types of therapy. They don't understand necessarily how many services are out there. Olivia: Yeah, I mean, that's a fair rebuttal. Like, that is fair. It is accessible. Andrea: Right, like I wish that could be different. Yeah, yeah. So there is that. And then the other one that I don't feel is as strong of a rebuttal is the fact that we as a society have really developed a strong interest in parasocial relationships. So those like online relationships. Olivia: or like what therapy is or like, yeah, all the things. Andrea: And I don't think that there's anything totally wrong with having online relationships. I have met people online and I have made friends, been in Discord chats before. That parasocial relationship isn't entirely bad. I've seen it work for people. But I think when we replace that parasocial relationship with a non-human... That's a problem. Olivia: Hmm. Yeah. I mean, now that we're talking about it, maybe it's... it starts as just blind curiosity. Andrea: Mm. Yeah. Olivia: I like, want, like, I wonder. I wonder what AI's analysis would be. Andrea: Mm-hmm. Yeah. And that's fair too. Olivia: and maybe just like starts with blind curiosity. Like I wouldn't even put something out there like that I was actually going through to see what it would say like as a joke or like because I think it would be funny. I wouldn't even do that. But I could see maybe somebody just being like curious, like, ⁓ like I wonder, because just like you said, the axis is right there. It's like the tips of our fingers. Then just like go do it. Andrea: Yeah. Yeah, I did the other day when I got my blood work back. I did have that curiosity. So I was like, I'm curious if I, I didn't put my actual blood work into Gemini just so we are clear. I just like said, these are my numbers with these things, like period. And I said, tell me what your analysis of this would be. And I was actually pleasantly surprised. It said right off the bat, I am not a doctor. Olivia: I am not a medical provider. Andrea: this doesn't mean anything. And I was like, oh, interesting. I've never seen that before. just like a, I was, yeah, I was actually really happy to see the disclaimer. And then it gave me some really interesting advice where it was like, here's the lowest, most likely possible things that could be going on. And then it was like, do you want me to keep going deeper? And I was like, oh, you're actually asking. I was like, no, it's probably just this low level thing. Olivia: like a disclaimer. Like, never mind, we're good. Andrea: But yeah, yeah. But I was like, that's interesting. So I can see the draw in the curiosity of that and the fast access. Again, we are a society that's building. convenience is what we strive for all the time in our society and we can't wait for answers. ⁓ But ⁓ a very real thing too. Olivia: Yeah, I mean, like, I think that example is interesting. Like, could they be using AI support through like MyChart for doing those things and giving like like potential initial interpretations of like your labs and that you'll get like, it's not official medical advice and that you'll get follow up from your provider or whatever. ⁓ Because generally I see like if labs are like truly concerning, they're not even going to put on your my chart first before a doctor is calling you. Right? So you already know if you're seeing the data results, you're okay. But still, like doctors see a ton of people ⁓ can take like days to get through lab results. It's like maybe that could be a useful way to be, I don't know. Andrea: Exactly. Yeah. Mm-hmm. Yeah, yeah, I mean, even when I did get the call from my doctor, was like, you know, we're gonna give you a referral to this specialist and that's not what AI told me. And I was like, ⁓ okay, like that's actually really like great. Like I can still kind of think about this and still move forward with this thing. But I cannot, I still in my brain cannot understand doing it for life, for like life and emotion. Olivia: Mm. Yeah, hey. Andrea: It's is so Emotion is so ⁓ like nuanced and you may feel is anger, I might not feel as anger. What you might feel is joy, I might not feel as joy and AI is not going to understand that. Olivia: Nuanced. Yeah. Yeah, think ultimately my stance on AI is just that there's so much we don't. There's so much we don't know about it. And there's so much we can't ensure or predict. And because it is so new, there's no laws or regulations around it. To me, it's too much of an unknown, too much of a risk. And when it comes to mental health specifically, obviously, we're not trying to poo poo AI. I think there are... Andrea: Yeah. Olivia: real life places where it can be helpful. Like we talked about where AI could be supportive in this specific field, for mental health specifically, like. Andrea: right? Mm-hmm. Olivia: I want to be as protective as possible around mental health because when it's not protected, we're landing in safety risk zone. And I don't like that. And nobody likes that. And so I am ⁓ way gun shy about bringing AI into the mental health field than I would be about maybe other fields where it could be supportive or helpful or make things faster or whatever the case may be. Andrea: Yeah. Yeah. you Olivia: Ultimately, like, I'm tenuous ⁓ AI. a AI note taker. If there were such things as like AI auditing tools, AI billing support, very much I think behind those things. AI note taker? I don't know if I like that. Andrea: Yeah. Yeah. Mm-hmm. Yeah, yeah, I think with reason, maybe. a line. mean, educationally, I think we all should learn how to do things by hand. Across the board, right? Even, I think we should all learn our executive We should learn how to do the dishes. If you choose to have somebody come do the dishes for you, that's great. Olivia: Yes. Yeah. Yes, yeah, that's important. Andrea: But like, think like learning how to do that, having that skill is just as important. Olivia: Mm. Yeah, I agree. Yeah, I totally agree. It's such an interesting thing. Like AI in general, such an interesting thing. So many people had different perspectives on it. And then I think the conversation can shift so much depending on the application of AI ⁓ health is definitely no exception to that. So. ⁓ Andrea: Yeah. Yeah. Right? Yeah, yeah, and it's depressing for me too to watch so much funding go into AI in areas that, you know, don't really help society with where we need the most support. And then cutting mental health support and giving that AI giving AI regulations. Huh. Olivia: Damn. Yeah, it feels like a fad, but unfortunately I don't really think AI can be a fad. You know what I mean? Andrea: Yeah, yeah, totally. I saw a funny video yesterday that was the question of like, do you think AI is going to take your job? And the person was automatically like, no, you've seen the movie. It went terribly. Olivia: you Andrea: And I was like, got him. Like, like Black Mirror episodes. None of those end well for anyone. I don't know if you've ever watched Black Mirror. Olivia: No, I know a little bit about it, but yeah, I haven't seen it. Andrea: Ugh. Yeah, it's, I mean, it's all social commentary, right? And a lot of them when they first started were about like AI features and things and none of it ends well, but it is so real. Like it is so real. Olivia: Yeah, it's kind of, yeah, it's also one of those things if you think about it for me, if I think about it too much, it kind of freaks me out. I'm like, no, gross, no, no, no, no, don't like it. Andrea: Mm-hmm. Yeah, yeah. we can, as an society, maintain boundaries and maintain regulations on and develop skills and understanding around this, I think it could be great. I love it. Let's use it. Create me a goddamn grocery list. Love you for that. But are we going to have that? Olivia: you Andrea: Don't have a lot of faith. Olivia: No. Yeah, yeah, I don't either. if you guys out there have any thoughts, questions, comments about just like AI, mental health, or like maybe even how you ⁓ used AI and what's been helpful for you. ⁓ always curious because like here my brother is running a like D &D campaign with AI, which I like literally would have never thought about doing. If guys are using AI in like useful, interesting I want to know. I'm like super curious about that. So. Andrea: Yeah. Yeah. Totally. Totally. And if anyone says that they've made an AI headshot, get out of here. Do not even leave a comment. Olivia: Definitely share if you guys are doing that. we didn't really touch that on that like AI ⁓ Yeah. Andrea: I could talk about AI in the art field for a whole nother 45 minutes. Olivia: I feel like I could sum it up with no. Andrea: Correct. Hire your local photographer. There are tons of us out there and we would love to take your headshot. ⁓ Olivia: Perfect. Please and thank you. Andrea: He said, thank you. I can tell that it's AI. Olivia: It is so real. Yeah, it's kind of freaky, honestly. Thanks y'all for coming on this journey with us as usual. Like I if you've been using AI and you found things that have been helpful for you, supportive for you, please share them. We want to know. We're curious. Do you guys have any thoughts or questions about things you want us to talk about within the field? Andrea: Yeah. Olivia: Always let us know. We're coming up with these on the fly. Like I surprised Andrea with this topic today. ⁓ So if you ever have ideas for us, definitely let us know. love to hear what you have to say. Other than that, you guys, we will catch you all in the next episode. Y'all take care. Bye. Andrea: Ha ha.