Image: The graphic features the Modern Therapist's Survival Guide branding with the title
Google PodcastsApple PodcastsStitcherSpotifySubscribe on AndroidShare Listen in a New WindowDownloadSoundCloudLeave a ReviewSubscribe via RSSKeep me updated!Player Embed

Are We Really Better at Empathy Than the Bots? What the Research Says About AI in Therapy

If you spend any time in therapist Facebook groups, you’ve seen the take: AI is a soulless, venture-backed box that could never do the real work of therapy. Maybe you’ve said a version of it yourself. This episode sits somewhere more uncomfortable. To close out AI month, Curt and Katie set the hot takes aside and look at what the research actually says about where AI fits in mental health care right now.

Some of what they found stings a little. In short, in-the-moment exchanges, people sometimes rate AI’s responses as more empathetic than a human’s, including trained crisis responders. That is not a sign you’re about to be replaced. It’s an invitation to get honest about the parts of this work that are genuinely yours: the relationship, the clinical depth, and the willingness to sit in the mess with someone over time.

Transcript

Click here to scroll to the podcast transcript.

(Show notes provided in collaboration with Otter.ai and Claude AI.)

In this podcast episode: What the Research Says About AI Therapy Chatbots and Empathy

Curt and Katie dig into the newest research on AI in mental health and what it means for your practice. They look at the Dartmouth Therabot randomized controlled trial, where an expert-built, safety-monitored therapy chatbot outperformed a waitlist control, and a Communications Psychology study in which third-party raters judged AI-generated responses as more compassionate than those of expert human crisis responders. They talk through what these findings do and do not show, why a quick hit of empathy can feel good in the moment but skip the slower work of the therapeutic relationship (confrontation, repair, sitting in the mess), and the real safety risks of consumer-grade chatbots that can validate delusions or miss a crisis. The conversation lands on the question a lot of us are quietly asking, some version of will AI replace therapists, and on what only a human in the room can actually offer.

Key Takeaways for Therapists: AI Chatbots, Empathy, and Clinical Depth

“If you’re just telling your clients ‘don’t use AI for help,’ they’re going to stop listening to you. AI is getting better in so many aspects. The better you’re able to explain the difference between where it helps and where it’s dangerous, the more trustworthy you’re going to be in your clients’ lives. Be credible instead of fearful.”

— Curt Widhalm, LMFT

  • Therapy chatbots can work, but the bar is “expert-built and safety-monitored.” The Dartmouth Therabot trial showed a purpose-built, supervised chatbot beat a waitlist control. That is better than nothing, not better than a human therapist, and it is a very different thing from a client venting to a general-purpose tool like ChatGPT, Claude, or Gemini.
  • Clients are mostly using the unsafe option. Most people are not turning to vetted clinical chatbots. They are using consumer-grade assistants that can validate delusions, reinforce sycophancy, and miss crisis cues. Knowing which tools have escalation rules and safety protocols, and being able to point clients toward them, is part of the job now.
  • AI is rated as more empathetic than humans in the moment, and that should get our attention. In the Communications Psychology study, people preferred AI-generated responses for perceived empathy and responsiveness, even compared with expert crisis responders. AI does not get tired, rushed, or depleted, which is exactly when human responders tend to get short and directive.
  • Short-term empathy is not the whole of therapy. Quick reassurance helps in a crisis moment. The deeper value of therapy lives in the long-term relationship, the confrontation, the messiness, and the repair, the parts that sycophancy and “context pollution” actively undermine.
  • Borrow what works. If clients respond to warmth, validation, and well-calibrated, substantive responses in high-need moments, that is a prompt to deliver more empathy with more structure, especially when we are tired or pressed for time.
  • Stay at the top of your game. A burned-out, distracted, or irritated clinician can do more harm than a well-tuned crisis chatbot. Taking care of yourself is part of staying the safer, more effective option for your clients.
  • Lean into what cannot be commoditized. Clinical depth, presence, authenticity, complex cases, and in-person human connection are the work AI cannot fake. As Katie puts it, most of what we learned in grad school is now a few prompts away. Presence is not.

“Deliver what AI can’t fake. It can pretend empathy, it can sound very empathic, it can feel validating in the moment, but over time it turns to sycophancy, it turns to just customer service. Being authentic and real with your clients in the room, that’s our secret sauce right now, and hopefully forever.”

— Katie Vernoy, LMFT

Staying Credible When Your Clients Are Already Using AI

If you tell clients to simply stay away from AI, you will probably lose them, because plenty of them are already using it. The more useful move is to understand these tools well enough to talk about them honestly: what they are good for, and where they get genuinely dangerous. That is the difference between sounding fearful and sounding credible. You can treat AI the way you already treat homework, referrals, and other outside resources, as something that can support a treatment plan when it is used with discernment. You were never the only source of your clients’ healing, and acting as though AI changes that mostly widens the gap between you and the world your clients actually live in.

A Note on Safety and Regulation

Both hosts stress that safety is not only a legislative or client-education problem. A growing number of states are moving toward requiring a human in or on the loop for mental health AI, raising the bar for what can be called a “therapy chatbot,” or mandating periodic reminders that a bot is not a person. Those guardrails matter, because people have been harmed when left alone with consumer chatbots in crisis. Clinicians have a role here too: pointing clients toward safer, monitored tools, building human coverage and between-session support into higher-acuity practices, and advocating for regulation that makes sense.

The Parts That Stay Human

Curt and Katie keep coming back to a simple division of labor. Let the machine handle the boring, easily automated tasks, and pour your energy into the parts that need a human: sitting with ambiguity and mess, offering confrontation and repair inside a real relationship, working with complex and higher-acuity cases, and being present for clients who want a human-to-human experience. The research is not a retirement notice. It is a nudge to get clear about where your value really lives, and then to show up there fully.

“AI is genuinely great at parts of our jobs that are really easy to automate. But we’re still necessary. So go out and be aggressively human about those parts. Let the machine do the boring stuff. You go and really deliver the stuff that makes the human aspect of it well.”

— Curt Widhalm, LMFT

Resources on AI in Therapy, Therapy Chatbots, and Clinical Empathy

We’ve pulled together resources mentioned in this episode and put together some handy-dandy links. Please note that some of the links below may be affiliate links, so if you purchase after clicking below, we may get a little bit of cash in our pockets. We thank you in advance!

Relevant Episodes of MTSG Podcast

Meet the Hosts: Curt Widhalm & Katie Vernoy

Picture of Curt Widhalm, LMFT, co-host of the Modern Therapist's Survival Guide podcast; a nice young man with a glorious beard.Curt Widhalm, LMFT

Curt Widhalm is in private practice in the Los Angeles area. He is the cofounder of the Therapy Reimagined conference, an Adjunct Professor at Pepperdine University and CSUN, a former Subject Matter Expert for the California Board of Behavioral Sciences, former CFO of the California Association of Marriage and Family Therapists, and a loving husband and father. He is 1/2 great person, 1/2 provocateur, and 1/2 geek, in that order. He dabbles in the dark art of making “dad jokes” and usually has a half-empty cup of coffee somewhere nearby. Learn more at: http://www.curtwidhalm.com

Picture of Katie Vernoy, LMFT, co-host of the Modern Therapist's Survival Guide podcastKatie Vernoy, LMFT

Katie Vernoy is a Licensed Marriage and Family Therapist, coach, and consultant supporting leaders, visionaries, executives, and helping professionals to create sustainable careers. Katie, with Curt, has developed workshops and a conference, Therapy Reimagined, to support therapists navigating through the modern challenges of this profession. Katie is also a former President of the California Association of Marriage and Family Therapists. In her spare time, Katie is secretly siphoning off Curt’s youthful energy, so that she can take over the world. Learn more at: http://www.katievernoy.com

A Quick Note:

Our opinions are our own. We are only speaking for ourselves – except when we speak for each other, or over each other. We’re working on it.

Our guests are also only speaking for themselves and have their own opinions. We aren’t trying to take their voice, and no one speaks for us either. Mostly because they don’t want to, but hey.

Join the Modern Therapist Community:

Linktree

Patreon | Buy Me A Coffee

Podcast Homepage | Therapy Reimagined Homepage

Facebook | Facebook Group | Instagram | YouTube | LinkedIn | Substack

Consultation services with Curt Widhalm or Katie Vernoy:

The Fifty-Minute Hour

Connect with the Modern Therapist Community:

Our Facebook Group – The Modern Therapists Group

Modern Therapist’s Survival Guide Creative Credits:

Voice Over by DW McCann https://www.facebook.com/McCannDW

Music by Crystal Grooms Mangano https://groomsymusic.com

Transcript for this episode of the Modern Therapist’s Survival Guide podcast (Autogenerated):

Transcripts do not include advertisements just a reference to the advertising break (as such timing does not account for advertisements)

… 0:00
(Opening Advertisement)

Announcer 0:00
You’re listening to the Modern Therapist’s Survival Guide, where therapists live, breathe, and practice as human beings. To support you as a whole person and a therapist, here are your hosts, Curt Widhalm and Katie Vernoy.

Curt Widhalm 0:15
Welcome back, Modern Therapists. This is Modern Therapist’s Survival Guide. I’m Curt Widhalm, with Katie Vernoy, and this is the podcast for therapist about the things that go on in our profession, the ways that we approach our work. And congratulations, everyone, we’ve officially made it to the finish line of AI month. So take a deep breath, give yourself a hand, or at least type a prompt into ChatGPT to generate a sound of polite applause for yourself. So we have covered a massive amount of ground over the last few weeks. We looked at the updates to the ethics codes, we looked at Circuit Breaker Labs, where we talked about breaking the bots, we’ve looked at how AI is being brought into the room by clients and how platforms like Simple Practice are weaving it into their daily workflow. And today, as far as our grand finale of this month, we’re stepping back from more of a tactical how-to question. We’re moving past, how do I handle this kind of tool, and we’re asking maybe a more terrifying question: W hat is the science actually saying about where AI is at in our field right now? Because when it comes to technology, I’m seeing a lot of the therapist community fall into two kind of lazy traps. The first one is kind of the superiority complex, the one that we’ve been hearing about for really the last I don’t know eight or so years, about as long as the podcast has been out, that AI is just a soulless computer model, it’s never going to grasp the deep resonance of human suffering. We are artists. The second trap is really more that AI has absolutely no place in mental health. It’s a venture capitalist demon box. A nd we are here to, I guess, dive into this, and to say that, put it out from the very beginning, this is probably an uncomfortable episode that we are going to get into the research that says maybe we need to be nicer to the bots, and on some…

Katie Vernoy 2:29
Or else they are going to come after us?

Curt Widhalm 2:32
Some of it is also maybe that our professional egos need a little bit of a check, so we are diving into this. Katie, are you in one of these two lazy traps?

Katie Vernoy 2:47
I don’t think so. From the beginning of my business, I’ve heard about the folks who push back against progress, the blockbusters, the taxi drivers, the folks who say this is this is ruining the profession, this is ruining society, and they can’t do it the way that we can, and they don’t have any place in our profession. A nd those folks get left behind. A nd I think I’m cautious around safety. I’m cautious around whether it works. I also want to make sure that I still have a job for as long as I want one, so I feel like I sit somewhere in the middle looking at: what is it? H ow is it developing? W hat are the places where it fits? A nd what do I need to do to shore up my own clinical skills, my own business, in order to stay in the game?

Curt Widhalm 3:53
So, really, what you’re describing is something that we have talked about on the podcast from the beginning, which is that we have to embrace whatever the technology is that’s available to us in order to make things better, and if you’re one of the people who seems to echo the many various therapists that I see posting all over the internet that says AI is terrible and we’re just training our replacements, maybe this episode is more for you than anything else, because for all of the people out there who say, brace the research, when the research embraces technology or says that technology can get better and it can make us better, then maybe we need to look at the utility of it, but we’re also a field where you know historically some of the things that Katie first came to mind is remember how revolutionary it seems in the literature when therapy clients just got asked questions? Y ou know like you know the beginnings of rational emotive behavioral therapy. O r, y ou know, how revolutionary it seemed when Carl Rogers was like, hey, you know, what i f I just give you positive regard? S o we’re not necessarily a field that embraces anything kind of in a hey, these are new and wonderful things that are coming up. I saw a Reddit post recently that was kind of questioning what does everyone think of common factors research and my response was, you mean the research like this is stuff that’s that’s still out there, but you know the research that really is out here now and the stuff that we try to stay attuned to here on the podcast is the stuff that’s coming out, like right now, like the stuff that’s coming out this year, and we’ve pulled together a few articles that are really some of the first examinations into randomized control trials with therapy chat bots. W e’ve been tracking some of the chatbots and stuff that’s been out there for the last couple of years. Katie and I have been involved in legislation with some of the artificial intelligence stuff, especially as an extension of our work with the CAMFT Code of Ethics updates. So people have been turning to us, this stuff is fresh. Katie, go ahead and talk about this first article, this is one that I think you found.

Katie Vernoy 6:24
Before we jump in, I want to just make the comment that we’ve been making probably the whole month, which is although Curt and I are both on the CAMFT Ethics Committee at the time of recording, we are not speaking for them. It just informs our perspectives and our perspectives are our own. S o the first thing we’re looking at is really whether or not AI is working as a therapy robot or for mental health within the clinical sphere. This is not your note taker, this is client facing, and in this particular article, it’s Heinz et al in 2025 it’s Dartmouth’s Therabot, and it was a randomized control trial that basically said that this Therabot was better at therapy than the wait list control. S o it’s been kind of talked about as though this means that you know AI robots are better than therapists. This was better than nothing, so this is better than not sitting on a wait list, but it’s not better than a human therapist. They did talk about making sure that moving forward that there would be some sort of kind of head to head with human therapists, and another really important point here is that this was an expert-built safety monitored therapy chat bot, and so this is not Chat GPT, this is not Claude, this is not Gemini, this is an actual thing that was built in order to serve this purpose, and so when we’re looking at where we sit in this current moment, therapy AI companies are coming towards us. They’re, they’re building things, they’re trying to increase access, or trying to do all these things. They are better than nothing if they are expert-built, monitored for safety, but I don’t know that our clients are actually going there, they’re going to the chats, the Clau des, that kind of stuff, and those things are not safe, as we talked about with Arul and Shirali in from Circuit Breaker Labs. So I think there’s this element of we need to make sure that we pay attention to this, we want to make sure that we’re fixing what’s fixable, but so far there is still a place for human therapist, but these therapy chat bots are good. They are actually doing pretty good.

Curt Widhalm 9:14
And I do want to really highlight that this is not a new feature of technology moving into our space, our space being the delivery of mental health stuff, that this has been going on for really as long as the internet has been widely available, of being able to deliver services to more areas and to people who wouldn’t traditionally be seeking out therapy, so this is something in balancing the various aspects of our principles of looking at if we really are as professionals standing for therapy being available to more people at more reasonable prices, this is something that does meet some o f those social justice aspects that many of us strive for, and this is just the newest iteration of it, that we’re not all going to be able to deliver services to all areas, clients are going to face some of the barriers, this is a dropping of some of those systemic barriers. N ow I can already hear some of you arguing, but they’re not going to stop there. They’re going to come into our spaces too, and they’re going to take things over, like it’s some sort of, I don’t know if you play the video game Civilization, but you know, or Risk, if you’re a little bit in to board games, but you know they’re going to come for our practices too. A nd you know this is maybe the kind of space to look at where things are at now, and I think that this is where one of the articles that I found is by Ovsyannikova and Inzlicht, this is from 2025 in Communication Psychology, and they were looking at Chat GPT’s responses for those in the crisis responder networks, and this is one of the things that really stood out to me in listening to a lot of our audience, a lot of our larger therapist commun it y, it challenges the idea that people have been saying for especially the last five or so years, you know, the bots can never replace our human touch, our empathy, the way that we communicate, and this research actually shows that people who are on the receiving end of these services respond that the Chat GPT directed responses show better empathy than humans do. They had looked at what people’s responses were to the Chat GPT responses versus the ones from the humans and the people like the Chat GPT responses better. They’re nicer. They are even with the sycophancy, t he perceived empathy is really what people respond to. Now, th is is in short, compassionate responses. These are emergency text lines. Th is is what is being delivered here. It doesn’t necessarily hold up over time, but we are not as far ahead in the empathy game as we would all like to tell ourselves that we are. We’ve talked about sycophancy a s a problem, and this continues to be a problem for artificial intelligence, but we are still ahead when it comes to long-term relationships, and the touch on that, but there’s some space here where we might need to learn how to be nicer from the bots.

… 12:54
(Advertisement Break)

Katie Vernoy 12:54
It’s interesting, and I think this could get worse as folks use AI more and more because they’ll get used to that sycophancy, and so to me, I think there’s this, this element of being able to differentiate what is it they’re actually finding better about the empathy for AI, and what are the things that over time are going to show that they’re not getting the help that they need, there’s the quick empathy that that reassures someone when they’re in crisis, but part of the benefit of therapy is that confrontation, it’s that messiness, it’s the relationship, it’s the repair, and what I’ve seen in the use that I’ve had for AI in building out, you know, show notes for our podcast, for example, is that there there’s what a rule called a context pollution, which is over time it gets more and more based on only the most recent answer, and so it loses the plot, and so it’s constantly in this customer service, I want to please you mindset, and so if you ask a question and you shift and you say, but what about this, it refocuses and refocuses and pays attention to what’s most recent, although I think folks are going to like it better because it’s reassuring and it’s empathic, and that’s of course you’re right, and of course you have all of the context, and you are the, you are the person that should win the argument. You were the best person I’ve ever met. At a certain point, it can be very clear that this is really just deceptive empathy, there’s not anything real behind it, and that there’s also nothing that you’re really gaining. It’s helping to keep you stuck. It’s helping to keep you only within your own perspective. And my hope is that over time, that some of the AI that’s that’s built out specifically for therapy will solve for these problems, because people are going to use them. B ut when I’ve had some situations where I started, you know, kind of tweaking something that was being written up, it just completely lost it, and it started creating something that was so iterative past what I had originally asked that I had to start a new chat, I had to sort through it, like it was just, it was a mess, and I get concerned that these studies are very short term, as you talked about it. They don’t necessarily get to that place where therapists are assessing Chat GPT, for example, in this space, because that’s not – they’re not asking for that, they’re not asking for the clinical perspective on it, and so maybe this sounds too clinician-focused, but clients don’t necessarily know what they need from therapy, and so them saying I like ChatGPT better is kind of like saying I like to go to therapy and sit and chat with somebody and have them tell me that I’m right and give me some quick tools that’ll make me feel better, and so I think a lot more research needs to be done, and I think clinicians need to make sure that they aren’t being cheerleaders. We had that conversation with TJ about the lack of clinical depth, and I think that’s something that we’re going to continue to go into. That clinical depth is really important for us as humans to do, because I don’t know how quickly AI is going to actually be able to replace that.

Curt Widhalm 16:27
To add to what you’re saying, I think that there’s always going to be more room to do research, because as we pointed out, we’re taking on some of the first rounds of some of this research and looking at it, and these short-term research studies are going to lead to long-term research studies. I agree with you that a lot of people who are going to be experiencing these kinds of things as clients might not know how to switch to a new chat window, or know how to ask, or to prompt or argue back with their chat bots in the same way that clients don’t necessarily always feel empowered to argue back and push back against their therapist, so I think that there is still a learning curve that’s here, but I think that this is still something where there’s a lot of space for us to have to take and embrace, and this is not the only part of society that this is happening in. I continue to read articles about college professors who are having to reduce the amount of pages that they’re expecting their grad students and their undergrad students to read on a week by week basis, because society around us has shifted. T hat there are certain aspects of what I look at all of our historical research in our field and say that this stuff doesn’t necessarily continue to hold up in 2020 as when it was published in 1978 for whatever reason, because society has changed, the world has changed, our attention spans are shorter, we as a society do want some of these empathic things, as much as I am visiting into the artificial intelligence world, as much as I use artificial intelligence for some of the various brainstorming things that I do, whether it be for coming up with ideas for podcasts, whether it be things that I come up for how to word things better for a newsletter or something else, or just kind of creative stuff on my own. There is still kind of that pull of getting the response back, where it’s like, yeah, I do ask great questions. So…

Katie Vernoy 18:53
You do, though. You do ask good questions, Curt.

Curt Widhalm 18:56
Thank you. So I do think that this is a societal shift that even with the cautions and the familiarity that I have with artificial intelligence that we are moving in this direction and I think the crux of this episode is how do we take this with the advantages that we have as humans with our clients to use that long term relationship and the confrontation and the deep work of therapy, but also taking what seems to be working from artificial intelligence and being able to either incorporate or adopt that in the work that we are doing in person with our clients.

Katie Vernoy 19:40
I think it’s important for us to, to also make sure that we’re looking at safety, because we’ve talked about this before, but there has been a lot of that sycophancy, the maybe even the context pollution, or those types of things that will create a crisis, that will put into place a weird dynamic where someone starts talking about delusions, suicidality, mania, those types of things, and Chat or Claude, or whomever will start validating these delusions and encouraging potentially dangerous behavior. There’s also very easy ways to to break some of the rules by spelling something wrong or putting some context around it. I think Shirali talked about that in the Circuit Breaker Labs episode, and so understanding that these bots can be available 24/7 they might be able to respond in a crisis. I think that being able to identify and link folks to ones that are safe, monitored, have escalation rules, so that that a person, or you know, some sort of safety protocols are in place is going to be our responsibility, because folks have died by suicide when they’ve been left to their own devices with some of these, these chat bots, and regulation is coming in and starting to regulate what chat bots can do, and I think that does help to create a little bit of safety, both for our clients, but also potentially for us as clinicians, but I think there’s there’s really a responsibility that we have to make sure that for our own clients and the folks who we are striving to serve that we are part of the safety solution, and that we are advocating for laws that make sense, because in truth this is going forward, whether we want it or not. There’s a lot of money being put into creating more access, whether it’s on the backs of clinicians who are getting paid less, or on creating AI and other technology infrastructure that will make access basically free for insurance companies or for consumers, and so I think this is a place where we want to pay attention, and I think we want to make sure that we’re really aware, and we talked about this, I think, in last week’s episode around what our clients are actually using, but this is another space where a practice like yours, Curt, I think is a really good model, where if you’re working with high-intensity clients that you’ve set up coverage and coaching calls and those types of things that don’t live exactly where this lives, but provide a different human mechanism of safety that can allow for some clients to feel like they don’t need to turn to especially consumer-grade free chat bots to try to get support in the middle of the night, because there’s, there’s definitely, it can be very problematic, as we’ve talked about in other episodes.

… 23:02
(Advertisement Break)

Curt Widhalm 23:03
We have both now brought up about the regulatory piece of things that more and more states are moving towards enacting that there has to be some kind of human in the loop, human on the loop aspect of therapist, when it comes to some of the artificial intelligence roles, when it comes to mental health, some of them are going to be creating a higher grade of what are called therapy chat bots, if not outright banning them, unless it involves human oversight, others are going to include language around if there is mental health stuff coming up in a regular chat with Chat GPT or a Claude or something like that, that every so often it will need to pop up with a reminder of I’m not a human, I’m just a bot, and these are some of the first rounds of legislation that is going in to address some of this as safety stuff, and it is based on some of the very real concerns that all of us are having, and being able to move into some of the safety to catch up with where the technology is. I do want to go back to the point, though, that I had brought up, as far as what is it that we can use and adopt from this? B ecause the other piece of this in some of the safety is what is also going to be what we are taking from the research, as far as what is safe and how we communicate to clients, and this is, you know, some of the opportunity for us to take some of this research from people such as and Inzlicht, i s if there is ways that we can learn better empathy from some of the communication to be able to communicate with our clients b etter, their research is on people who need the safety in the moment. The coaching calls that you’re talking about, not that I think that people who are out there taking coaching calls are just answering the phone, going, ‘What? What do you need now? Like, get over it, rub some dirt on it, kid.’ But if there is places where we are able to take some of what is working from people in their time of crisis to be able to focus our energy in that when people seem to be responding at this high level of need, this is where we actually need to focus our empathy even more, and that’s that’s the part where safety is not totally just on the legislative end, it’s not totally just on the client education end, but safety is also in learning new ways for us to be able to say we have now seen hundreds of 1000s of minutes of clients, if not millions of minutes of clients operating with artificial intelligence. This is an insight into some of the interactions that people are actually having and actually rating that they’re deeming feels more safe than going to a therapist.

Katie Vernoy 26:20
What are the types of things that are in those reactions? Because you said we should learn from them. What are we? What are we? What are you talking about specifically?

Curt Widhalm 26:27
Some of that is first, we have to look at the human condition of the therapist. We have to look at where our own vulnerabilities are, as far as we get tired, we’re rushed in between sessions. Sometimes these coaching calls have to be incredibly short. AI doesn’t get tired, so we have an actual really good comparison here of our competitor, air quotes, around that never gets tired, it’s always there, doesn’t carry any emotional strain, it’s not doing any sort of task switching, but what we do when we as humans are in that very short kind of space, we tend to give shorter answers. We tend to be really directive and almost kind of pushy sometimes, but what people are rating their responses are at this point I actually just need more empathy for where I’m at, and those are the kinds of things where we have research that we’ve never ever been able to compare to before, of a cross comparison of people want nice empathy in these heightened crisis moments, they want the pushback in the longer term relationships, but these are the pieces of information where we really have to put on that empathy face when we’re reacting to clients at this higher level, and we have this research now smacking us in the face, saying therapist get better.

Katie Vernoy 27:54
It’s interesting, because I think there there is a real clinical utility in looking at the fine tuning that’s happened in trying to create these appropriate responses from an AI chat bot. I think the other thing just to pay attention to is that most of these chat bots are not sitting for licensing exams, and they’re not being held accountable by, you know, the like the Board of Behavioral Sciences, like you and I are, so I think there’s this, this element of taking what’s useful and also recognizing that we still have different responsibilities and accountability, and before we hit record, you talked through a couple of different responses that you know, one was human and one was from a from an AI bot, and I think that my concern is when AI gets very empathic, and maybe, as you had talked to me before, maybe it provides some call back to session, it might provide some, you know, plausible coping strategies or things to think about. I always get worried as a clinician when I’m creating those, those responses that I’m engaging in something that I can’t follow up on. Now AI can actually still continue to go, and you know, then we have those, those kind of long-term solution issues that we’ve talked about already, but, but I feel like there’s, there’s maybe even some new ways to think about communication based on what clients are actually wanting in those crisis moments. I don’t know if we have to go into it in more detail, but I think it’s really interesting when you’ve got whoever the clinicians are that are informing the creation of these AI chat bots, their stylistic choices are getting codified into how things are being at least designed from the beginning, and then ongoing feedback from clients is what’s fine t uning and honing those responses down to something that is more effective, which, as individual clinicians, we don’t have that. We, we certainly can have hundreds and 1000s of hours of clinical work, but not millions, right? We just can’t. We can’t have that much information, and so I think it is interesting, but I also, I want to hold that caution that I still think that there are things that we want to pay attention to, and I say even more dramatically, we need to make sure that as human therapist s that we’re staying at the top of our game, taking care of ourselves, so that we aren’t burned out, we don’t have that irritation or distraction that might cause us to be more harmful than an AI chat bot that is finally honed to respond in a crisis. Alternatively, we can also try to engage these and partner with these types of of AI platforms and monitor them, so that we can gain what’s best of both worlds.

… 31:09
(Advertisement Break)

Curt Widhalm 31:11
You’re bringing up a couple of things, so I want to bring up one more point about this article. One of the things that the respondents really enjoyed more was that the length of the responses from the chat bots tended to be better. They weren’t too short, they weren’t too long. Now they’re substantial in a lot of being able to explain what is happening to a client, why they’re reacting in the way that they are, but also giving some pretty direct advice as far as W hy don’t you try these skills? A nd I’m taking out of the article that by comparison the human-led responses didn’t go into that level of depth in their responses, so it became kind of more from the respondents, and I’m waiting for my human to tell me more of what to do, and these are the kinds of things that I think that we can adopt better in our delivery of services, is being able to take what works and deliver it to clients, but it also means that we have to look at Is what we’re doing actually working? And I think that that’s another very common theme of our podcast i s when’s the last time you looked at how and what your clients are actually responding to? It might be the relationship, it might be people that are just showing up to you because they’ve gotten over the shame-free aspects. You know, that’s another piece of why people are going to chat bots in the first place. Is this is not a human that I have to respond to. Don’t have to necessarily deal with my feelings with. They’re constantly available, they’re not burned out, they don’t go on vacation, whatever else it might be, but what does it say about us, and you know, the system of psychotherapy that the people that were worried about turning to chatbots feel safer going to a computer where their own data is being mined and potentially sold and put into the algorithm for everything else about them at the same time that we’re asking you as our listeners to deliver more empathy and deliver it with more care and deliver it with more structure. B ut we have to address this as there is a societal movement that we’re having to address. If you are feeling threatened by artificial intelligence, there’s the stick your head in the sand and hope it goes away movement. Business advice from Curt: I don’t advise that you do that. But we are really at this crux of more than ever people have more options to take their feelings and their problems than to just us.

Katie Vernoy 34:09
And I think the thing that I was thinking about when is when you were talking is that as soon as Google came on the scene there has been increasingly better ways that folks can get information about pretty much anything, and that includes at this point information about how to take care of their mental health, their physical health, plan their travel, all those things, and when we add AI on top, whether it’s the Google AI that pops up in your browser or Claude, Gemini kind of situations, w hat it really comes down to, we need to be something more than cheerleaders and coping strategy experts. We need to have presence. We have to be doing our own work. We need to be able to work with depth and authenticity and be able to sit with the messiness, because most of the rest of the stuff that we learned in grad school is now a few prompts away on folks’ favorite chatbots, and so to me it makes our profession, I think, a lot more interesting, and it maybe asks a bit more from us at times, because of needing presence to be what we offer. We can’t see a million clients half spaced out hoping for the best. We actually have to provide some depth.

Curt Widhalm 35:45
This takes us to the part of the episode, which is, what does this really mean for you as our audience? A nd we’ve got several ideas here, but the decision to move more into this month of AI, and really being able to talk about things more in depth, is we are in a place where we, as the service providers, have to be more literate about what is out there, how it’s showing up, how it’s showing up in our clients’ lives, so that way we can really better be able to explain what these tools do and don’t do. If you’re just telling your clients don’t use AI for help, they’re going to stop listening to you. AI is getting…

Katie Vernoy 36:32
Exactly.

Curt Widhalm 36:32
…better in so many aspects. It’s not perfect, it’s downright dangerous in some places, but the better that you’re able to explain the difference between those two things, the more trustworthy that you’re going to be in your clients’ lives. So be credible instead of fearful.

Katie Vernoy 36:52
And we have a whole episode on this, but I think being able to really engage in that conversation around how AI is being used in your client’s life, I think, is really important. I will reiterate: deliver what AI can’t fake. It can, it can pretend empathy, it can sound very empathic, it can feel validating in the moment, but over time it turns to sycophancy, it turns to just customer service and being authentic and real with your clients in the room, that’s our secret sauce right now, and hopefully forever, because that’s the best part of being a therapist, and just touching the surface of that is not going to be sufficient to keep your caseload full, I don’t think.

Curt Widhalm 37:40
And recognize that as there’s more tools available, that that can hopefully be a better part of your client’s treatment plans, knowing what the tools are that can better benefit them. We are not the sole providers of our clients’ happiness, we’re not the sole place of their healing, we’re not the sole place of them getting into a better life. We already refer to other kinds of professionals. We already give homework outside of sessions. This is stuff that knowing where the line is between what AI can genuinely help with: psychoeducation, finding better ways of researching things between sessions, also knowing where it’s dangerous, and helping to provide psychoeducation about that to our clients is a way that we can better adopt the world that we’re living in rather than the world that we wish was there, because it’s too hard to adapt.

Katie Vernoy 38:46
And to that end, I think there are places that we’ve talked about this in episodes previously, but you can incorporate your own AI to create stronger admin systems, make things easier for you, help you focus more on that clinical depth that we’ve been talking about, and also to potentially evaluate your clinical care. There’s there are so many different ways that AI is being developed to support clinicians, both to make their job easier, but also to make them more efficient and effective, and in clinical efficacy, especially when you have basically deliberate practice built in, because you’ve got a an AI bot helping you sort through what you’ve talked about, what you’ve not talked about, things you’ve missed, moments you’ve missed, that kind of stuff. There’s there’s some stuff coming out that I think will really challenge us, and hopefully in a good way.

Curt Widhalm 39:45
So to self reflect, if clients find AI more empathetic than you, what is your empathy actually for, and are you actually delivering it? T hat if w e were to take the general attitudes of what we read in Facebook online posts, we’re very grouchy bunch of people, and very critical…

Katie Vernoy 40:10
Falling asleep late, not answering the phone.

Curt Widhalm 40:14
So if we aren’t actually exercising our empathy with and for clients, we’re going to get replaced. I f AI can do things between sessions w hat’s our plan for clients who are needing help when we’re not available? If it’s just wait until our next session next week, you’re not operating in the world that we live in anymore.

Katie Vernoy 40:43
Yeah, and I think we’ve talked about this in other places, making sure that what we’re designing our practice to do really focuses on what is you, what can only you do, what are the things that AI has not done as well, and we’ve talked about clinical depth, but it’s also complex cases. It’s it’s folks who are not in love with technology who want an in-person human to human experience, and so some of that stuff is different than what I think some clinicians started providing during the pandemic, taking a hard look at how do you make sure that you’re showing up with the things that AI and the tech companies creating them cannot commoditize, the things that are like I said before the special sauce.

Curt Widhalm 41:36
And I think to close out AI month, one of the things that I think is the hardest piece of this is no matter where you’re at in your career, you can be a student listening to us. Thank you for joining us very early on in your career. You could be somebody who’s been well seasoned and listened to every single one of our episodes, and to the millions of people who fall into that category. Thank you. And the question really is, there’s not a lot of really structured guidance out there on how to solve this. This is something that is some deeper work for you to have to do about you and the way that you practice, and that’s something that can only really be solved by you. The data isn’t saying that you have to go and just retire immediately and go and just find a new job, find a new career, disappear someplace. It’s saying that AI is genuinely great at parts of our jobs that are really easy to automate, but it’s also something we’re still necessary, and go out and be aggressively human about those parts. Don’t be aggressive, torture clients, but really lean into the parts of this that you know, let the machine do the boring stuff, you go and really deliver the stuff that makes the human aspect of it well. All of the evidence of our field right now is that the big VC stuff is going to try to take every single aspect and make things cheaper. AI is going to make at least the paying the humans delivering the services cheaper, lean into the parts that you do well and be irreplaceable about it.

Katie Vernoy 43:28
I like it.

Curt Widhalm 43:29
You can find our show notes over at mtsgpodcast.com. Follow us on our social media. Join our Facebook group, the Modern Therapists Group to continue on with this and other conversations. And until next time I am Curt Widhalm with Katie Vernoy.

… 43:40
(Advertisement Break)

Announcer 43:44
Thank you for listening to the Modern Therapist’s Survival Guide. Learn more about who we are and what we do at mtsg podcast.com. You can also join us on Facebook and Twitter, and please don’t forget to subscribe, so you don’t miss any of our episodes.

 

/0 Comments/by
0 replies
SPEAK YOUR MIND

Leave a Reply

Your email address will not be published. Required fields are marked *