We need to talk about the use of AI with coaching because something is happening. In fact, it just happened the day before I started recording this episode. Now, usually I spend a lot of time scripting and structuring and recording and editing these episodes so they follow a very helpful format. What just happened is so new and is probably going to cause a huge change in how people use AI that I wanted to do this episode. So if this one feels a little different than usual, that's why. First, some background on me and AI. My education is in computer science. I've been teaching computer science and IT for over 17 years now at the college level. So I have a very good understanding of how what we're calling AI actually works these days. And I also want to give a disclaimer. Yes, I do use AI. Later I'll explain when I use it and when I definitely do not, because that makes a big difference. I am not anti-AI. In fact, I like AI, but only if you use it in a smart way. So let's start with the big issue I'm seeing right now. People are just asking AI to spit out content for them. I see it constantly on Facebook posts, in coaching groups. It is absolutely rampant on LinkedIn, and I see it all the time on Instagram and threads, in a supposed storytelling post where people are sharing something that supposedly happened. And maybe they are telling the story, but when I hit some of these AI tells, I immediately start to question whether it's a real story or they just asked AI to write it. Within about 10 seconds, I can tell you if something was written with AI, because there are very distinctive structures and tells in the way AI writes. And the problem with all of this is that it makes everybody sound the same. Maybe you've heard this advice, and if you haven't, hear it now. This isn't my advice. This is from years and years of marketing across all walks of business. So the advice is you need to stand out. People buy from people, they don't buy from a company or a business. So when you use AI to generate your marketing content or your content in general, you're going to sound just like everyone else. Even if you add in your own expertise and your own pointers, you're still going to sound like AI. Are there ways to make AI sound less like AI? Yes, there are. And I've done a lot of that work with my own setup, but I will never ever put out content with AI without working on it myself. So let me tell you a couple of the AI tells. And once you start hearing and seeing these, you're probably not going to be able to unsee them. One of the big ones, and this is a thing called, it's not X but Y. For example, a coaching post might say something like, Building a group program isn't about getting more clients, it's about creating more freedom. Another tell is short staccato sentences, typically three in a row. On online posts, you'll see it three lines in a row, just a couple of words each with a period at the end, very staccato and back to back. You'll also see terms like, I needed to sit with that. And in storytelling posts, when people aren't speaking like themselves, they'll say things like, I got the phone call and my stomach dropped. Yes, we see that in literature and writing sometimes, but that's not how people typically talk and post. You'll also find lots of extra words stuffed into sentences that make them longer, and I'll say fluffier than a normal person would ever write or say when they're just conveying information. The more you read and start to catch this, the easier it gets to go, yes, that was probably AI. That being said, I am not anti-AI. If you can find a way to get AI to help you speed up your processes, I think that's useful. What you don't want to do is use AI to be out there and pretend to be you, because inevitably it's going to sound like AI or it's going to do things on your behalf that you normally wouldn't do. An example I saw recently, I was in a coaching group and somebody started posting about cruises. I'm thinking, that makes no sense here. Why are you posting about cruises in a coaching group? And a bunch of people caught on pretty quick. Yeah, that's probably somebody using AI to automatically post for them. Another one is using AI to have conversation. There's a technique in social media posts and groups where you ask open-ended questions. You're fishing for leads, fishing for people to share things with you, and then you engage them in a conversation, eventually try to move them to a DM, get them on a call, and so on. I'm not a big fan of that, but everybody has their own selling style, so you do you. The problem is they'd ask these questions. Somebody would give a short response, and then they'd ask a follow-up, like you're having a normal conversation, but it's not normal. If you ask someone what they do and they say, Oh, I'm in real estate, a normal person makes a little small talk like, Oh, how long have you been doing that? The AI version jumps straight to something like, Wow, you must feel so empowered when you help people find their dream home and get to move into it. What's the hardest problem as a real estate agent? And then no matter what the answer is, it immediately tries to do the same thing again and again, trying to eventually move them into a sales call. And that's just not how normal conversation works. And I've seen people get caught by this. In one thread, somebody was posting these questions. You could clearly tell they were AI. And another person said, I feel like I'm having a conversation with Claude. The poster said, Oh yeah, you caught me. The questions were mine, but I was using Claude to make them cleaner. And I get that. Maybe the person speaks English as a second language and was trying to use AI to smooth the language. I could understand that. Hey, I myself have often used AI to sort of smooth things over. Sometimes when I'm trying to write an email response that maybe I'm a little too emotionally invested in, I will ask AI how it's coming across and I'll use it to, you know, sort of tone it down, but I will then go back and edit it more. You can't just take it and put it out there as is. Because it's still not you. It still sounds like AI, and people are going to catch on, like in that example. Okay, I'm not a marketing person. My place in the coaching world is helping coaches design their group programs, the courses, everything inside of a group program to grow their business. Marketing is part of what I have to do, but I'm not here to sell you marketing. What I do want to stress is this. If you sound like everyone else, if you sound like AI, you're not going to stand out and people are going to stop engaging with you. Why would they want to work with you if you're using AI to get all your content? How are they going to know you're not just using AI to answer their questions? And if that were the case, why don't they just ask AI themselves? They could pay $20 a month for Claude or Chat GPT subscriptions and get the same answers you're giving them. And that brings me to the bigger thing. The whole reason this has become such an issue. In the past 24 hours of me recording this, Anthropic, the company that runs Claude, announced they are going to start watermarking everything because of European Union laws. They're going to start with their new models and eventually work it into their older models along with offering tools so that people can check whether content was generated with AI. Which means even if you spent a lot of time getting your bot of choice to sound like you in a month or two, especially if you're using Claude, people are going to be able to copy that, drop it into an app, and see that it was AI. Anthropic said when one of their models generates text, it weaves an invisible watermark right into the words. It doesn't change how the text reads, but it travels with the text when you copy and paste it, and it can survive some editing. Now, to be accurate about it, Anthropic is clear that the mark tells you the text was run through Claude. It doesn't prove a human never touched it because people use these tools to proof, read, and tidy up their own writing all the time, but that still gets marked. For our purposes, that's the part that matters. If you had Claude clean up your post, that mark can still be sitting there, and there are going to be tools that let anyone check for AI use. So it's going to be kind of like a scarlet letter for those of you old enough to remember reading that book in high school. It flags that Claude was in the mix, even if you did some of the writing yourself. And it means you're not going to be able to quietly lean on AI without people figuring out that you're using AI. And you may be thinking, oh, that's fine. I don't use Claude, I use ChatGPT or Gemini or whatever other model. The truth is, if Anthropic is bowing to the EU on this, the other AI companies will most likely follow suit. So it's probably only a matter of time until all of it is watermarked the same way. Now in the business world, this isn't such a huge deal. Businesses are pushing AI so hard right now, and that's a whole other topic. But when it comes to helping people one-to-one or even in a group program, you don't want to use AI to build your course or your program. If the chatbot is just going to spit out the same stuff whether you told it to or your client did, then why are they paying you? You need to be the expert. You need to be the person sharing that knowledge and helping your clients get that transformation. So one of the big things I'll tell you is this don't use AI to generate your program or course content. That said, I think it's perfectly okay to do a brain dump and ask AI to help you organize your thoughts. Put it into an outline for you. But don't just drop in a little bit of information and say, okay, go turn this into module one. That's the same mentality as people who used to info dump into a video, call it a module, and then wonder why their course didn't help anyone. There's a science and a process to building educational resources that help people learn new skills and get unstuck in their lives. Whatever kind of coaching you're doing, there are processes we follow to help those people. And as coaches, we need to be the coach doing the coaching. We do not need to be a coach who asks AI how to help a person and then just regurgitates it. Now, having said all that, let me tell you how I use AI because I do think there are perfectly reasonable uses for it in a business. The first thing I won't do is use AI to build, examine, or critique any content a coach gives me. If I'm helping a coach build, let's say, a live course, I will not dump their info into AI and ask, does this work? I will sit down and use my 17 years of experience as a college educator to go through what they have, find what's working, find where they can make it stronger, and figure out how to build in the processes that help the client get through the material and have the same kind of transformation they get with one-on-one coaching. What I will use AI for, like I've said, is doing a brain dump and asking it to organize it. I do that all the time. In fact, I often do it for these podcast episodes. I'll sit down and say, here's what I want to talk about. I'll ramble off everything I can think of, and then I'll ask AI to put that into a skeleton outline for me. Then I'll go through, write the script, edit everything until I have a script that I'm happy with and I think works. Sometimes I will ask AI to go through and make sure I haven't said anything stupid or have weird grammar issues, but other than that, the script that comes out is from me. I'll also use AI for research. I have Claude and I use the cowork tool with a Chrome connector. So besides searching through Claude's own systems, it can actually go out in the internet and interact with things on websites. I'll often have it go look things up and check things out for me. And even though I'm pretty organized with my computer files, every now and then I can't find something, I'll ask Claude to go through my Google Drive or Dropbox and find it based on what I think is in the file. And it does that pretty well. I'll also ask it to go through my email and see if there's anything I missed that needs a follow-up. And then I do the follow-up myself. So in a sense, I use AI like a virtual assistant. I ask it to help me with tasks that speed up the process, but I don't ask it to do the final versions. And to be clear, when Chat GPT first hit the scene several years ago, I was doing that. I would just say, hey, do this thing, do that thing, and try to use it for marketing or whatever and so on. But I quickly realized that wasn't the way I wanted to go. At first, I thought it was a shortcut, but then I learned it sounded like AI, and if I sound like everybody else, no one is going to listen to me. That's when I came up with my own AI policy. And my AI policy is basically this. I'll use it to help me with tasks that are repetitive, or that can be sped up, or that can be done while I'm doing more important work, like communicating with clients, holding client meetings, and writing scripts. But I will no longer use it to create the content itself. It can help me brainstorm, organize an outline, but after that, I do the work to generate the content and put it together. I won't use it to do the work I do in helping people with their programs and courses at all. I will use it for note-taking during meetings, though. After I finish a meeting with a client, the AI does the note-taking and I can post those notes and a transcript in our shared client working folder. If either of us have questions about what was said or decided, we can go back and look. So that's my argument to you. Come up with your own AI policy. Think about what you're comfortable using it for. And if you hate AI and don't use it all, hey, more power to you. I am not trying to push AI on anybody. I know people keep saying, oh, it's coming, it's coming, you've got to get used to it. And that may be true to a point, but I don't think that not using AI means you're going to lose your business. I know people will say, but if everybody's using AI and marketing, how do I keep up? Well, until the large language models get better at sounding unique, everybody using AI for marketing is going to sound like AI, and people are going to start to see through that. If you're out there even just once or twice a week, sending your newsletter or doing whatever marketing you do, and you're doing it in your own voice, it's going to be more persuasive than pumping out tons of AI content. That's my opinion. I strongly urge you to think about how you're using AI on your public-facing side. In your business, for the tasks that just burn up your time, the things a virtual assistant could do, absolutely go for it. But the big one I'd really recommend against is using AI to build your own courses and your own programs. That has to come from you. You need to be the person sharing that information with your clients. And if you're not sure how to do that, I've got plenty of episodes on it. How to get started with group programs, what to put in your courses, all of that. There's plenty of that in my previous episodes to help you get started. If you want the next step, you can go and listen to my episode on the fastest way to grow and save time as a coach. It's right alongside what we're talking about today because using AI well is really about freeing up your time for the work that only you can do. I'm Dr. Curtis Satterfield. You've been listening to Program Design for Coaches, and remember, you've got what it takes to build your group program. I'll talk to you in the next one.