AI for Clinicians: Tools, Trust & Thoughtful Experimentation with David Lopis

Craig: Hello, everybody, and welcome to the Healthy Business Lab Podcast. We interview amazing allied health business owners doing cool things to learn their gold nuggets and how we can grow and evolve and experiment in business. And today, I’m super excited to introduce somebody who’s done exactly that and where we’re going to go with some of the experiments that David’s done. David Lopis from AI for Clinicians, welcome to the podcast today.

David: Thank you so much, Craig. Pleasure to be here.

Craig: Yeah, it’s great to have you on. And again, I love interviewing people that have run allied health practices but then evolved and taken different tangents in what they’re doing and in the business world. I’d love for you just to share a bit of a background about yourself, your starting point, and your current evolution in the allied health space.

David: Sure, absolutely. Thanks, Craig. I’m a psychologist myself and supervisor. I guess I’ve dabbled in working in private practice, worked as a school psychologist, and worked in community.

Over time, having worked as a supervisor, around late 2022 I really stumbled across some exciting AI tools that I thought were really interesting. Talking to a range of clinicians and asking them how they were using AI, I’d get all these really confused looks and this real uncertainty around what sorts of things you can do ethically and what the benefits were. It was all very unclear to the people that I was talking to.

I established a bit of an online community, started doing talks, started consulting, and have really pushed this whole space to better understand AI for clinicians—what it’s all about and what we can do with it.

David: Because I just think there’s so much possibility, and it’s just a really exciting space.

Craig: Yeah, cool. I’d love to unpack that a little bit further, particularly the evolution from running a practice as well. How have you found the background of leveraging your psychology skills in navigating that transition into the AI space? What have you used personally as your anchor points as you’ve made that evolution?

David: Look, I guess I have been using a whole lot of different kinds of tools, both more clinically as well as the more general tools—the ChatGPT, the Claude, the Manus, even.

I’ve just always been asking myself that kind of question: how can AI help here? But what’s also important as a clinician, and I guess as a psychologist, is that we’re very ethically minded.

It’s also about what we should be using and where we should be drawing a line. That’s frequently come up as a question. I’ve been very mindful along the way of being aware of what I’m comfortable with, what society and the profession would be okay with, but also staying across what new tools I need to know about and being able to promote what I think will be helpful.

From the perspective of psychologists and many clinicians worldwide, the buzz is around AI scribes, which I’m sure we’ll talk a lot more about, but also being aware of the many other kinds of tools that are available. The scribes themselves, I think, are where the excitement is because of the time savings—perhaps 30% to 80% of admin time. That’s certainly been my experience with playing around with a few different kinds of tools.

Craig: Yeah. I’d love to go back to your experience. You said back in 2022, when AI was still in its early days, you started to come across it. What were some of those initial exposures to the technology and AI space that got you intrigued and interested? What did you grab hold of at that point?

David: I played around with a tool called Jasper. That was my first one, even before ChatGPT. Did you ever see Jasper?

Craig: Yeah, yeah. No, I hadn’t. I’d heard about it, but I haven’t seen it.

David: It was mostly a marketing-type tool, from my understanding at the time. What blew me away initially was that I could be writing a sentence, click on a button, and have it think and go off from there based on what it thought I wanted.

That was my initial reaction: wow, there’s got to be something here. But then there was also the mindfulness around client information. You can’t do anything with that. So, how can we overcome this? My initial experience was, “Wow, but how is this going to be more than a gimmick? How is this actually going to be useful?” I couldn’t tell other clinicians that it had all these uses, but that they couldn’t really do anything with it. So I kept asking myself, where do we go from here?

It was an internal struggle that I’ve constantly had, and I still have, because you frequently see new tools coming out on LinkedIn and in the wider world. Very rarely are they answering questions around compliance and how we can actually use them with client information. I think that’s always got to be front of mind.

Craig: Going further on that point, how did that side of things evolve for you over the years as more models became available and there was more opportunity to access them? How did you reconcile privacy, permissions, and that approach? How did you get the comfort level you needed personally?

David: I started making inquiries when I saw AI tools in the clinical space that were specifically released with the intention of being used by health workers in Australia, and no doubt worldwide as well.

Because I could talk to a clinician with an AHPRA registration and ask them various questions, they could provide assurances around compliance, which was not something I could get before. Once that became a space where others were using tools that I could trust more, that’s when I began to think, okay, now this has a lot more legitimacy.

But in the earlier days I also hired lawyers to do some investigations and answer questions about the space. I touched base with my own professional advisers to ask some tough questions. I really think this is a process we have to build comfort with—as a profession and as a society—to figure out what we’re comfortable with, where we build trust, and where we’re taking things too far.

Craig: I love that. If I think back to the profession and what you would have done in your career, no doubt that curiosity and asking questions—whether you’re talking to a patient or client in a session, or whether we’re talking about AI and how it interrelates—is that interesting component of making sure you satisfy yourself, ask the questions, and then work out what the next steps are.

As you started to get more comfortable with what you were seeing from your own investigations, where did you go next? Was it the scribes? Was the time-saving aspect what really came to the front of mind, having worked in practice and run practices, and being able to spend more time with clients rather than doing admin?

David: That’s where the interest lies overwhelmingly, because I don’t think clinicians are being exposed to many other tools. At the same time, writing notes and reports can take many hours for clinicians over the course of the week, so it’s understandable that this space has had so much excitement.

From my own experience of having such an admin-heavy week, that’s not what clinicians get into the space to do. I don’t think many clinicians wake up in the morning thinking, “I really can’t wait to do my notes today.”

Craig: No, I’ve spoken to many, as I’m sure you have. This podcast has been symbolic of that. What do we get into practice for? It’s definitely not the writing, the numbers, and all those sorts of things. It’s a people profession.

David: It is, absolutely. But I also don’t want to downplay the business development side, because many clinicians don’t enjoy that either and can use AI to assist with it.

That can be taken to the extreme, where graphics generated by AI or the wording and phrasing come out all wrong. That also needs to be done properly. But it can take a real load off clinicians who want to spend less time on it and have extra assistance with AI.

Craig: I think we can go down two paths here. We’ll cover both: the administration and time-saving component, and then come back to the marketing and business development side.

With your experimentation with scribes and the admin side, how has that evolved over time in terms of the quality of the output, privacy, and adherence to regulations? How have you found that evolving?

David: Results do change. They have changed when I’ve tried tools earlier on and then there’s been a model change, an update, or new features. The speed of change has become enormous—outpacing anything else, I think, in the history of technology growth and evolution.

I’ve found it a struggle to stay on top of everything, but also to find the right kinds of tools, because I’m teaching people and sharing my findings in my communities. I don’t want anyone wasting time on tools that aren’t appropriate.

I do find myself going down rabbit holes and realizing that something isn’t helpful or worth using. Then, five minutes later, there’s another tool and a whole new set of uses to consider.

For example, I’ve been playing around with AI receptionist tools and setting one up for myself to see how it works if someone calls me and to automate that aspect. Earlier on, there were mixed experiences, but you realize it won’t be long before the tools become much more sophisticated and many of those issues are resolvable, either through your own settings or by talking to the companies.

Craig: I’d love to dig into your process when you do an experiment. Do you set it up as an experiment? Do you say, “This is the outcome I’m after, this is how long I’m going to give it”? How do you approach playing with these tools?

David: It’s a great question, and I probably do need to come up with a more systematic way of doing it.

It depends on whether I’m asking how I would use this personally, how I would use it as a clinician with my client, or how I would tell my client about using it. There are tools for clinicians, tools for clients, and tools that are part of the clinician working with the client. These are all going to be different.

Ultimately, it’s a matter of setting it up, seeing what its core uses are, and trying a few different common interactions to see what my experience is.

Craig: How long will you do that for with a particular tool? If you start something today and have a play, do you go down a deep rabbit hole and really work through it? Or do you give yourself a bit of time, see how it evolves, and revisit it later?

David: I’d say there’s an initial setup period of up to a few hours, but often a lot of these tools can be understood within the first hour. Then, over time, you fiddle with it, interact with it, see how it goes, and maybe try again in another month or two or three.

You also spread the word to others and reflect on their experiences. We have the ability through social media to leverage off each other, and that’s much more powerful than me using a tool just for my own purposes and vision.
Being able to say, “Here’s what I’ve seen, this is how I’ve set it up, and here’s how far I’ve taken it,” and then inviting others to share their experiences is really valuable.

Craig: I absolutely love that. Having developed software in the financial space, I know the value of getting other people’s opinions because we often find ourselves in our own version of how we use or do something.

I love that that’s part of your process. At the end, we’ll share details about your community, which no doubt gives you great feedback.

If I circle back to the marketing and business development side, allied health professionals don’t often get into practice or business to do marketing. They want to see their clients and look after them, but we need to make people aware of us in the first place.

How have your experiments with different platforms helped with that? You touched on some areas where it can be challenging, such as AI going off on its own tangent and sounding too technological. How have you found a good approach for people to leverage AI in marketing and business development?

David: I think there is huge opportunity and enormous excitement, with good reason.

The process is: set it up, get AI to do what you’re doing, and then refine, correct, edit, and oversee. That’s the process. I don’t want to give the impression that this is set-and-forget or a once-off, because then there will be garbage. There will definitely be garbage if you do that.

One process I love, which I think makes such a difference, is getting a tool to interview you and ask you questions so it can understand your business.
You could say something like, “I’m wanting to develop a master document. Ask me 20 questions.” If you want to be extra lazy, you can ask it to make the questions multiple choice with suggested and recommended answers.

You might respond, “1A, 2B, 3C,” or write out a sentence if the options aren’t very good. By the time you’ve gone through that process, the AI can build an understanding of your business. You can then correct and edit it, working towards a master document that describes how your business works in as much detail as you want.

You can put that into an AI tool in the future—any AI tool you like—and have it leverage that understanding. That’s really powerful. You can continue to refine the document and help the AI better understand how you work.
Craig: I love that. In my own practice and evolution of working with different models, I got a lot of value from having those working documents that the AI helped create. It’s like having an external brain sitting over there, but then asking, “Do I really think like that?” And then realizing, actually, yes, I do, but I need to change or recalibrate it.

Knowing that you’ve built a community around AI, what transformations have you seen across administration, business development, psychology, and allied health more broadly? Have you got a couple of examples that really stand out and have turned the dial?

David: For me, what is really exciting to hear are clinicians who say they’ve come out of retirement because now they enjoy their work. That, to me, is huge. They’re excited to use the tools and excited to focus on their client work.

How do you quantify that? I don’t know. But to me, that’s what this is all about: bringing inspiration and motivation back to the work.
We’re also seeing clinicians use tools in interesting and novel ways. For example, being able to better understand how to work with clients with unique needs and quirky interests, and being able to produce resources, strategies, and ideas that connect to their strengths and values, is now much more possible and easily achievable with AI.

Clinicians can even generate images. ChatGPT now has an amazing image generator. If you want to make a comic based on a child’s interest in cats to teach social skills, that’s now doable within a prompt.

That sort of thing is changing how many hours might have gone into preparing materials. Preparation is now much easier.

Craig: I love that. I want to jump in on two aspects, but particularly the uniqueness and the ability to craft something so personal to any interaction professionals are having with clients.

I got a few goosebumps thinking about how that plays out. We’re a neurodivergent family, so we understand. We’ve got kids working with occupational therapists, psychologists, and others.

The ability to tailor something so specifically to an individual—and the rewards and transformations from both sides—would be incredible. Practitioners could create something so quickly that really captures those moments. Is that what you’re finding in the feedback when people are doing this?

David: Absolutely. It’s a win for rapport between the clinician and the client. It builds bonds within the family when they can use such resources.
I think we’ll also see improved outcomes over time because clients will be able to better understand what you’re trying to teach. That’s a huge win in all the most important ways.

Craig: As you were talking, it made me think of one of my children who works with a psychologist. Even having the scribe there and looking at it together can build rapport. They can look at it and ask, “Did we say that?” Then they can edit things together, get engaged, and have a laugh about what’s happening.
I hadn’t thought of the rapport-building aspect of having them involved in the process. That’s super cool.

Even when you mentioned clinicians coming out of retirement, that’s huge in this space. The ability to get more practitioners back into something they love is important, especially when burnout has been so prevalent in the industry.

David: Absolutely. Even those doing big, long reports may be able to get a first draft in moments instead of spending hours. That’s going to change how much admin is required.

For NDIS purposes, there are big, long reports. In the forensic system, there are big, long reports too. There are ethical and legal issues involved, but conceptually, it changes how people see their work, burnout, and employment.
People working for clinics are going to ask, “Can I use AI?” Those who are told no may start looking elsewhere.

Craig: That’s an interesting point around AI adoption and how to approach it. If a practice was sitting on the fence and feeling nervous about the regulatory implications, what would your messaging be? What have you learned? How would you talk to them about why it might be worth their while?

David: In terms of the business case, every piece of admin—notes and reports, for example—can represent many minutes or potentially hours saved on each task.

If you value your time and the time of the clinic, you should put a financial value on that. A lot of places treat admin as invisible time that magically appears in the day, but it doesn’t. It’s part of the day, and just because it’s someone else’s time doesn’t mean it shouldn’t be valued.

You should also consider the emotional value.

From a regulatory perspective, look at what your regulator says. AHPRA has released information about the obligations for professionals using AI. They aren’t condemning it or saying not to use it. They’re outlining ways to use it properly and the major considerations around using it.

Professional associations and groups are similarly providing guidance. Very rarely are they saying not to use AI at all. Courts are also describing acceptable uses in various ways, although in many cases they’re saying experts shouldn’t use it at all, which is interesting.

Do your own homework. Things will get really interesting because clinicians whose work may go before the courts are already using AI. Clinics will need good systems in place to make sure work generated using AI is appropriately managed and documented in reports.

Craig: That’s appropriately managed and looking at the wider picture. I love that you’ve touched on the financial impact, emotional impact, and systems.
This speaks to how practices need to think about all these things, and how AI models can help with them. It’s funny that the answer again comes back to curiosity and asking questions: “I don’t know what I don’t know.” Having AI help moderate that process is super useful.

As we get towards the end of the recording, it’s a great opportunity to share information about the group and community you’re building around peer facilitation and learning. How can people connect with you, David?
David: The easiest way is through LinkedIn. Find me, David Lopis, on LinkedIn.

I also have a Facebook community called AI for Australian Psychology by Psychology Squared, and anyone is welcome to join. It’s not just for Australian psychologists. aiforclinicians.com is where you can join the broader AI for Clinicians subscription community. I also do talks and events and am happy to give sessions for clinics or groups about what the AI space is all about, great uses for clinicians, and the ethical and responsible-use issues to be aware of.
For example, there’s been a lot of debate about whether clinicians can say, “I’m not going to work with a client unless you give me permission to use AI.” That’s going to be a frequent question.

It’s important that clinicians form their own opinions, get their own advice, and understand the issues. There are going to be lots of issues, and AI today is not the same as it will be tomorrow or the next day.

People need a space to keep the conversation going. That’s what I’m trying to build through these communities: a way to stay up to date.

Craig: I love that. Doing similar things in the financial space, I know the value of shortcutting the process of looking at the multitude of tools, models, and other options.

Being able to start useful and interesting conversations brings awareness to AI and technology. Coming back to the idea of being intentional, practitioners, clinicians, and business owners need to be intentional about how they approach it.

David: Absolutely. The flip side is that a lot of people are going to an AI tool sales demo, thinking they now know everything about AI and are ready to go.
From a technological perspective, if you’re confident with technology, you might be ready. But understanding the ethical and professional issues is a lot more involved and continues to evolve.

By talking with peers who are using AI, we can understand great uses and continue building our knowledge base. It’s really to the betterment of our clients and ourselves that we learn these new tools.

There are many tools that will benefit clients, and plenty that will help clinicians too. There’s no shame in finding tools for both.

Craig: One hundred percent. We’re all about everybody getting what they want and need from systems, processes, technology, and the like.

I really appreciate your time and the gold nuggets you’ve dropped today. I’ve got a page full of notes that I’ll be going back to after the call.

We’ll add all the details you shared—the LinkedIn information and your group—and make sure we share them as part of the show notes.

Once again, David, I really appreciate you sharing your wisdom and knowledge and the gold nuggets here today.

That’s a wrap for today’s Healthy Business Lab Podcast. As always, like, subscribe, and follow the links to connect with our guests, because they’ve got so much information to share. They’re building plenty of great things and are all about networking.

David: Wonderful. Thank you so much. It was really a pleasure to be here.

Craig: Make sure you do all the things I said, and we’ll see you soon on the next episode.