10. Writing with AI in Your Private Practice: Keep Your Copy Genuine and Engaging

In this episode of Run Your Private Practice with AI, Kym Tolson shares practical tips for integrating AI into your copywriting to make sure it sounds authentic rather than diluting it and leaving it sounding like a robot wrote it. She emphasizes avoiding common AI clichés like “unlock your potential” and “transform your life,” which can signal AI involvement in content creation. Kym suggests being precise with prompts to reflect your unique voice, whether it’s humorous, serious, or inspirational. By using detailed and creative language in prompts, you can make AI-generated content sound less robotic and more personalized.

Kym also discusses the importance of using AI with a critical eye, particularly in avoiding the overuse of emojis and generic greetings that can undermine professional credibility. She advises on how to direct AI to produce succinct, relevant content that truly resonates with your audience. The episode concludes with a reminder about the SEO implications of AI-generated content and encourages listeners to join the Clinical AI Club for further learning and support.

Key Points:

  • Avoid clichés and common phrases that signal AI usage.
  • Tailor AI prompts to ensure content reflects your unique voice.
  • Use AI responsibly to maintain professionalism and enhance engagement.

References:

Transcript: 

00.00

Hey everyone, welcome back to another episode of run your private practice with artificial intelligence. I’m Kim Tolson, founder of the clinical AI club. And today I want to discuss how to blend AI into your copywriting. So you can ensure that it complements your writing rather than overtaking your unique authentic voice. So let’s break down some surefire ways to know that your content is AI generated and not human generated or specifically generated from yourself. We want to focus on how to use AI to you know, enhance your your world, your emails, your all the marketing you have to do, we want to enhance that we don’t want to just make it more complicated for you and then really turn people off with different AI giveaways that you might be using and not even realize our own AI telltale signs that we might be using already. So first I want to talk about like all the cliches that AI tends to use, like specific phrases like unlock your potential. And then the word revolutionize and the phrase transform your life, those sorts of things are all really common AI cliches. So if you see that in writing, it’s almost a surefire giveaway that a person used AI to either write that copy or that subject line or any of the content you are reading, you’ll even see it in Instagram captions, is another one I see and hear all the time is game changing to describe things AI uses that for almost everything. It will say something is a game changing solution. You know, if you’re reading that somewhere, it’s probably written by a AI. Okay, so here’s a funny story from when I first started using AI, my boyfriend, Ken watches all of my reels. He’s my biggest fan. I know. It’s super sweet. But anyway, now he says to me all the time. But is it a game changer, Kim, it’s pretty funny, but also not funny, because that’s how much I use that term at the beginning of my AI journey. Apparently, in my early AI days, I thought Game Changer sounded great. Now, not so much. So if you’re using any of those words that are just really more commonly associated now with AI, I would suggest changing them. And we’re going to talk a little bit about how to do that. And so the really the solution for that is precision in the prompts that you’re using. So you can tailor your AI interactions to really reflect your own unique personality. So if you’re super funny, and you always communicate with humor, you can instruct your AI to inject a little wit or a little sarcasm or a little humor into your copy depending on your personality. For instance, when you’re creating prompts, instead of using generic words like inspire my readers, you could say something really more specific, like give me a pep talk that’s so inspiring that it would make a sloth want to run a marathon something like that very specific. And that nudges the AI towards using more creative language and ensures that your content is going to stand out and not just sound like a robot wrote it. And when we’re focusing on trying not to sound like cliches, it’s really about trying to infuse your content with relatability. So some ways you can do that is by understanding a major core concept. If you can start to realize that generally readers of emails and even social media consumers, they generally crave connections and stories that relate to their own lives. So when you’re using artificial intelligence, you can ask it to weave in anecdotes or ask for examples that really resonate with your audience’s everyday experiences. And instead of letting AI predict what your user needs based on very little information, you can start to train it and saying things like, Oh, this is Susie, she did this and that and had this outcome. And then you can guide the AI to have more insights basically, from real interactions that you’ve had with people real life stories around whatever you’re trying to talk about. For example, if you decide you want to write a blog post that addresses the struggle of balancing parenting with self care, and you want to use humor, and you want to use empathy, this sort of guidance will prompt This will prompt AI to craft content that feels more personal and authentic and sounds like it’s coming from your voice and not from a robot. Basically, it’ll help you default from those predictable patterns that you know, always comes out those really wordy sentences that are just like so generic. So hopefully using some of those tips is going to help you get through that. Okay, so let’s move on the next dead giveaway that somebody’s using AI to write their content is the emoji overload that can happen when you use chat GPT emojis of course can brighten a message but when it’s overdone, it really can dilute your professional credibility. So the strategy here is a strategic deployment of emojis like one or two maybe if it’s appropriate for your type of client or the type of message that you’re trying to get out or even the type of emotion you’re trying to convey. But AI tends to put 1000 Emojis into copy especially if you’re telling it to write copy or to be funny or something like that. It’ll use emojis to try to explain the concept even more. So if you see an email or some social media copy out there or even a website with tons of emojis you can pretty much guarantee that AI wrote that because it’ll come up with tons of emojis and sometimes even a different emoji for each sentence. If you’re telling it you want it to be funny or that sort of thing. So if you are getting copy like that and you’re putting emojis all over the place, I would just suggest so that you rethink that strategy that you make sure you’re using just a couple of emojis that really convey your message. It needs to feel appropriate for what you’re writing. But just one or two emojis will do usually you don’t want it to be a ton of emojis and like a thought chain of a progression of Oh, somebody gets out of bed and they have their morning juice, and then they go for a walk, and then they’re doing this and then this, some people do emojis and a chain like that. And it’s just like way overwhelming for the audience. Usually, instead, you could simply just use like an arrow pointing one way to explain like, you get from here to there, instead of using a ton of emojis trying to explain a concept. So all that to say, if you’re using AI, and it’s putting out a ton of emojis, you can give it directives around how to use emojis, you can say, look, I only want no more than two emojis in this entire thing that we’re writing right now. So you can be very directive with it just because you want to make sure your copy isn’t coming across as AI written because people just glaze right over it. These days, they see those big words that I was talking about. Those like cliche catchphrases, game changing, revolutionize all that. And then they see a ton of emojis, people are not going to read that content, they’re not going to pay attention to that they’re going to skip right over it. Okay, so let’s talk about the next thing that is a dead giveaway for using AI to write your copy or your emails or whatever you’re using it for even your own social media posts or your emails to friends and family. The thing that AI loves to do, especially if you tell it to write is it loves to have these really generic cheesy greetings things like Hey, there, you empowered therapist, something like that. It seems like a message like that as a catch all for the audiences you’re trying to reach. It’s like, yeah, my audience, they are empowered therapists. That’s great. But it’s just so cheesy. People are just gonna see that and skip right over it. So I have a business coach Brittany long, she gave me this advice a long time ago that she said, just skip the greetings altogether. You don’t need to say hey, so and so welcome to my email it actually in today’s times, it takes up time. It’s like something It’s like another barrier to getting to your content. And AI loves to put these like fluffy language and greetings like that. And everything it writes for you is so you can be real directed with AI, you could say I don’t need a greeting. I don’t need you to add extra words into these sentences. I want you to be succinct into the point. And a lot of times that’ll help it avoid that like just reading that sounds like a robot wrote it right at that greeting, hey, empowered therapists, because, you know, it’s just so cheesy. It’s like, okay, that is not something you would write personally to somebody that you’re trying to, like nurture or give helpful advice to, or something like

07:28

that, you’re not going to greet them like that, not in real life, not like a human. And sometimes I’ll say that to chat GPT,

07:33

I’ll say you sound like a robot, I need you to sound like a human. Don’t say game changing. Humans don’t say that. They would say this instead. And I give an example of what we would say. And then it starts to get the message like, Okay, I shouldn’t be talking like that. Because again, if you’re, if you’re putting this out to your audience, it’s just, you know, it’s not helpful. And it’s not going to get the result that you’re looking at here that people don’t like it, they’re going to glaze right over and probably just send you to their junk mail. So the next sort of dead giveaway with AI is those power words. And I already touched on it a little bit with the cliches but sometimes with the fluff. Sometimes AI will make sentences more fluffy. But it uses these power words like transform, elevate, revolutionize, awaken, unleash, these are like dead giveaways that AI is writing that because like I said before, humans don’t talk like that. I’m not going to say, you know, unleash your power to a person. If I’m talking to you in person, I’m not going to say, hey, so and so don’t you want to unleash your power? I’m not going to talk like that. I’m not going to say that. So when AI starts spitting this stuff out, sometimes it can sound really good when you’re reading it. But if you really think about it, it’s like, oh, I never would have put it that way. And you will think gosh, this, I mean, this does not sound like a human rights. This sounds like AI. So if you’re putting out content that’s got a lot of these big power words, I would suggest that you start giving directives to the AI and saying, again, this is not how a human rights and give really specific examples of how a human does right for example, instead of transform, I’d like you to say you can get to the place where you want to be or something like that something more like in just basic human terms, because if you’ve got a whole like training document, and it’s using that kind of language, people are going to tune out, they’re not going to connect to that. And that has been my experience. And you know, staying to the topic of this episode, it’s really going to turn people off. And it’s just the dead giveaway that you’re using AI. So if you’ve got, you know, these power words, and these like cliches, and then a bunch of emojis and a cheesy greeting, that is like a dead giveaway that this is an AI written topic. So I hope that helps give you some clarity around just what your writing might look like to others. And now that you’ve heard me talk about it, you’re probably going to notice it even more. I use AI so much that now I can just tell I can spot it a mile away when somebody is using AI to write their like captions on Instagram or something like that. And I’m guilty of it too. I’m pressed for time all the time. I don’t have a lot of time to sit down and think about copy. I’m not a good writer naturally. So a lot of times it’s just the easy way out You know, I it’s like, okay, this output is great, that’s a long caption, it gives good tips, people are gonna like that copy, paste it, plop it down with some emojis into my Instagram. And those don’t get a lot of engagement really, it’s easier for me. But it’s, it’s worth it probably to put more time into correcting it to make it sound like you than to just plop it down or send an email out with something like that, that just seems super generic. So pay attention to that. But also, as I was talking about in the beginning, getting much better with your prompts getting super directive around, you know, the prompt the structure that I’ve talked about in a couple of episodes here. But the prompt structure basically, is you want it to be specific, you want it to be directive. First, you want to do an introduction to what you’re looking for, like the context of the question you’re trying to put in. And then you want to make a specific request. And then you want to talk about any constraints or conditions like I want you to not include any emojis in this, you put some sort of constraint in there. And then the closing of the prompt is a call to action. For example, I want you to summarize all this in a bullet point format for me. So just to review, you want to do the introduction, the specific requests, the constraints, if there are any, and your closing call to action, how you want it to summarize the content that you’re asking for. Here’s an example, I’m a therapist in private practice, I’m overwhelmed by paperwork, I’m looking for AI solutions to streamline my progress notes. And to ensure confidentiality and compliance. Please give me a list of ideas on how to do this and the bullet points. So that’s, that’s a summary of how you want to make the prompt. So when you’re trying to write your copy, you want to say I’m a therapist in private practice, I need to write an email to let my clients know that I’m now specializing in EMDR, and I need it to be in a friendly professional tone of voice, then please don’t use any emojis. And finally, I’d like you to ask them to sign up for this new group program that I’m offering or something like that. So if you construct it that way, it’s going to be a little more likely to sound like you and not a robot. And another thing you can do to make your AI sound more human is add writing samples that you’ve done before. So if you’ve written blog posts before, or emails that really converted well to people you were sending emails to or anything like that, even your journals, if you’ve got journals lying around where you really liked the way you wrote about something, you can use all of those in different ways to train your AI, you can use it with chat GPT, if you have the paid version, you can actually take a screenshot on your computer of things. And now and you can just upload it to chat GPT. And you can tell it to read the picture, you can put your writing samples in like a Google Doc, you could download the document and then upload that as an attachment and tell it to read the document and learn your tone of voice. You can also like like I was saying, like your journal or something like that, you can just take a picture of it. And you could just upload the picture to chat GPT now the paid version does that and you could tell it to read that and learn from it and write like you and more advanced features is you can build a GPT. But if you know you’re going to be writing a lot, and you really want to train it around your voice, you can totally do that by giving it writing samples and you know, blog posts. And you can even feed it samples from authors that you really enjoy or people you love their tone. Like I love Laura Belgrave, she has the best emails, in my opinion anyway. So sometimes I’ll copy one of her emails that I’ve saved. And I’ll just give it to Chad GPT. And I’ll say I want this to be written in the voice of Laura Bell gray. And here’s a copy. Or here’s an example of her writing or my business code, she does great emails, sometimes I’ll tell it to write like Brittany long, I’ll tell it to go to her website, when with systems.com and analyze the website and the tone and you know, write some website copy for me or you know, if I’m writing a clinical topic, I’ll say, Gosh, I really want to explain this from the perspective like Brene, Brown would or something like that, you can say go to this website, or you can go to a website, copy something Brene Brown said or some quotes or whatever, put it into chat GPT and say this is these are examples of this. But if you put it in your own GPT you never have to program that again. So that’s the really cool thing about GPT is where with Chad GPT or all these other API’s out there, you’re constantly having to like retrain the chat threads. So building your own GPS gets rid of that for you, basically. And it can help you train it to write like you and not sound like a robot. And next week in the AI club, I’m actually doing a two hour workshop with the AI club members that have the paid version of chat GPT and want to build their own GP Ts. So I’m super excited about that. I’m sure some people will want to just learn how to write more like themselves and create their own voice within their GPT or they’ve got special ways they do clinical techniques and that sort of thing, really anything like that you can create a GPT and it’s basically a clone of you. I did a private consultation the other week with somebody that he’s trying to get his whole team basically to be on board with AI and he’s got all these progress notes and a particular format that he wants to write in. So we worked on creating a GPT where it will analyze chart information and analyze like components of a treatment plans and that sort of thing and make sure that it’s in the right format that he wants it to be in so he trained this GPT he created So he’s just gonna give it to all his clinicians and tell them to use that as a tool. So there’s lots of ways to do that. So anyway, really excited about GTs and how you can train it. And I hope this was helpful to just recognize the pitfalls that are like the giveaways that you’re using AI to help you retrain it and to help you kind of give it that glance over before you publish something to make sure it doesn’t sound like AI. And also, it’s important these days, because a lot of the SEO out there like Google and all that is starting to really pick up on these patterns and messages like this. So if it’s super AI FIDE but like I was saying lots of power words, and lots of cheesy cliches and weird greetings and tons of emojis, the SEO bots out there are going to start realizing that this is not authentic writing that somebody’s putting out there to the world. It’s aI generated, so it’s not going to register with SEO, if you’ve got a lot of that stuff going on. So anyway, I hope that helps. If you have any suggestions for an episode or something you want to hear about reach out to me, I’m happy to record an episode to give you more information about any of these topics. If you want to really dive into using AI in your private practice, join the clinical AI club at WWW dot clinical ai club.com. I also have a free Facebook group on Facebook. It’s called run your private practice with AI. There’s not tons of engagement in there, but it’s a good place to go and ask questions. You don’t want to join the AI club or if you need help with something real specific. I can answer that over there for you. Okay, so thanks for listening and I’ll see you next time on the run your private practice with AI podcast.