5 Levels of AI Marketing Mastery: How to Level-Up Your Use of AI

Happy 2026. Hey, how's this for a New Year's resolution? How about you Master AI marketing? The AI Business Society gives you live training with experts who are getting real results. No hype, no fluff, just proven workflows that work. And you've got access to an incredible library of training from many of the incredible guests that you've heard on this show. So secure your AI transformation by visiting socialmediaexaminer.com AI.

Welcome to the AI Explored podcast, helping you put AI to work. And now, here's your host, Michael Stelzner.

Hello, hello, hello. Thank you so much for joining me for the AI Explored podcast brought to you by Social Media Examiner. I'm your host, Michael Stelzner, and this is the podcast for marketers, creators and business owners who want to know how to use AI. Today, I'm going to be joined by Dan Sanchez, and we're going to explore five levels that every marketer goes through as they are up leveling their use of artificial intelligence. And here's the deal. Whether you are a marketer or not, I guarantee you're going to find a lot of value in today's episode. In addition, there are a lot of cool little nuggets that I discovered that I did not even know in today's episode. So you for sure want to listen all the way to the end. If you're new to this show, be sure to follow us on your favorite podcast app so you don't miss any of our future content. Let's transition over to this week's interview with Dan Sanchez.

Helping you simplify your AI journey. Here is this week's expert guide.

Today, I am very excited to be joined by Dan Sanchez. If you don't know who Dan is, you got to know Dan. He is an AI strategist who helps marketers prepare for the future. He's host of the AI Driven Marketer podcast and his course is AI Fundamentals. Dan, welcome to the show. How you doing today?

Thanks for having me on. I'm excited to be here.

I'm stoked you're here. Today, Dan and I are going to explore a step by step process where you can kind of take your AI proficiency to the next level. That's probably the best way that I can describe what we're going to talk about today. Now, Dan, before we actually get into the goods, I want to hear a little bit of your story. How did you get in to AI Start wherever you want to start.

You know, I've heard a number of your episodes already, and I Think everybody kind of has that common theme of, like, November 2020, I guess it was two, you know, when ChatGPT drops and you're like, OMG, like, what is going on? So I have a similar story. But I had done all the marketing stuff. I'd spent time in every channel, most recently podcasting, like this one. I love podcasting. And I'd been running a very popular marketing podcast called B2B Growth and Doing all the topics, reading all the books, talking to hundreds of guests, and then AI hits the scene. And I'm like, oh, my gosh, this is the newest thing that I've seen. And I haven't had this feeling about something since, like, social media started picking up. And that was a while ago. And you were a big part of that, right? And then that's how your whole start came out of, like, the whole social media marketing world. But I didn't know what to do with it. It was so new. It wasn't a new channel. It wasn't this new way to approach marketing yet. And everybody was just been like, wow, this is the new thing. So I think I. I did what most people did that first year and just kind of experimented with it. And that's what everybody said to do. That's what the advice was like, oh, just play with chat GPT. So I did, and I learned a little bit. And I worked for a tech company a little later on, and they had a CEO that was very big into AI. And I learned some things from him about super prompts and chain prompting and a couple of different, like, rudimentary, like, how to actually not just prompt, but how to engineer your prompts to get better results out of AI. But it wasn't until December of last year I was like, this is it. I'm convinced AI is the thing. This is the next big wave. It's going to take over the world. At least it's going to take over marketing.

Well, just to be clear for people listening in the future, December of last year is December of which year? Just because they might.

December of 2023.

Okay, got it. So keep going with the story.

And I was like, this is it. I'm going to do what I like to do. Which, if you can see my video screen, like, I have a lot of books behind me. I'm an avid reader. I read everything and take all the courses and have the degrees and like, like, I love to learn. So I'm like, but there's no books on this topic. And anything that was written on it, it's outdated. You know, I go to search for podcasts and I can hardly find a handful of podcasts. Your podcast wasn't around back then, you know, So I started listening to the few that were out there. There's hardly any YouTubers that are talking about except for, like, channels that are just focused on broad AI, not AI for marketing. So I'm like, well, then I guess if it has to be, it's up to me. And, you know, I start rolling up my sleeves and I start trying to create custom GPTs at the time that was noon. I start experimenting aggressively with it. I start my own podcast, the AI Driven Marketer, and start just talking to anybody I can hear doing anything remotely interesting with AI. I'm like, hey, be a guest on the show. Hey, be a guest on the show. Tell me exactly what you did. Tell me everything you know. So I started going into aggressive interviews and experimenting with what they learning. And then I finally created a framework for myself that December around how to know if I'm actually making progress on that journey towards mastery. Because if AI is going to be the next big thing, and I think it's going to be even bigger than social media, then surely I need some kind of framework to know I'm making progress on it. And that's what I want to actually talk about on the show today.

Very cool. So love the story. So you started the podcast around What, December of 2023, is that correct? Or when. When did you start that show?

Yeah, December 2023. December.

Tell everybody kind of what you cover on that show.

I cover practical AI. I also get into strategic AI a little bit, like how to use AI to augment or enhance your marketing strategy. But a lot of it's very practical, very tactical. That's the thing that I thought was kind of missing back then. I was getting a lot of. I caught AI in the sky, like, oh, the sky's falling. Oh, that theoretically in the future, it's going to take over. It's going to be Utopia. Oh, look at this cool AI thing that doesn't mean anything. You know, like we, we can now automate cars, blinkers. You're like, I don't really care about that. It's all these weird little use cases I wanted to dive into, actually creating practical episodes that were useful to marketers now, whether on a strategic level or on a tactical. Here's how to create better social content today with it, with what's available now, or even maybe just soon in the future, like right around the corner.

Is it a solo show where you're just giving thoughts and advice. And I don't mean that in a negative way, I mean that in a positive way. Or is it more interview show still?

It's both. I do. It's probably half and half. It's interviews and solos. My solos are always me sharing some breakthrough, like a revelation that I've had about how something works, or it's very tactical, like, hey, look how I built this thing. I learned how to automate sending out newsletters from my podcast. Here's step by step how I did it and how you could do it too.

Love it. And folks, if you want another show to listen to, definitely add his AI driven marketer to your listening lineup. All right? There's plenty of people listening to this that are early in the journey because we're all early, let's be honest, as of this recording, we're not even at the two year mark. I don't think we're about to come up to it. If I'm doing my math right. Yeah, we're coming up on two years since the ChatGPT moment. So those that are early in the discovery phase of, you know, learning AI, specifically from a lens of marketing, what do you want to say to the marketers who are just getting started? Why is this so important for them?

A former podcast I used to be a host of. We did this hundred marketing leaders leader survey. And I don't mean like they filled out a form. I mean like we literally interviewed a hundred people asking them about 10 questions a piece. And there was a lot of interesting things that came out of that, that survey, that qualitative survey. But the overwhelming one that I left, I walked away from was out of these hundred leaders, when we asked them, like, what's the major problem you're running into? It was unreal. Like almost all these marketing leaders, and I'm talking like CMO vp, maybe high level director, marketing leaders said there's too much to do, there's not enough time to do it, not enough talent to go around, and there's not enough budget. Like, the expectations of marketing are sky high. And what we have to actually reach it is just so little in comparison. And you might have a massive budget, but the expectations are even more massive. You might have a small budget, but it's still a high, high bar to clear. And this is happening across marketing all over the place. And that was the big pain point. And the reason why I think all marketers should be learning at mastering AI, because AI can help you scale. Like you've never been able to do before. You could do more and you can even do better with AI if you learn how to leverage it properly.

I love this. And it's such a game changer for those of us that are employing it. And it's kind of like new discoveries every day. Don't you find that, Dan?

Oh, yeah, it's like Christmas every other week at least, but every other every week almost.

Yeah. It is quite literally like a sandbox box, like a kid in a sandbox with sand and tools. And you just can create almost anything. Just like, you know, metaphorically. When I would go to the beach in San Diego and hor del coronado, there'd these be these crazy sandcastles people would build out of sand. And I'm like, how is that even possible? And sand is kind of a good metaphor. Like you can build a lot of things with sand, you can build glass with sand, you can create all sorts of things. And AI is like sand almost. Right. It's such a powerful tool that if you know how to do it well, you can accomplish absolutely amazing things. And that's what I love about it. And that's why I'm super excited to get you on the show. So, all right, we're going to talk about what you've already hinted about, which is that there's various steps or stages or levels that anyone who's employing AI goes through, but especially marketers. So let's start with kind of the first step where you see most people beginning.

Where most people begin is using AI as what I call personal assistant. And it's on the one to one level. You, you open up ChatGPT or Claude or Google Gemini. You ask it for something, it gives it back to you. And it's a lot like how we would use a personal assistant if you ever worked with a personal assistant. That's how it goes. You assign a task, they go and do it, they report back. And that's how a lot of people use it. It's very helpful. You can do it for one off task. You can use it to supplement your work. It generally makes you faster and more efficient, just like a personal assistant would. Even a highly trained or somet we at least I've heard a number of people in the AI community talk about like your genius intern. Like it's really smart, but you kind of have to give it some guidelines because it will go off and do some dumb things sometimes. Right. Some things that I like to do at this level that I think are very practical are using it to do kind of like a little bit of that administrative work that I know I don't like doing. Like, I don't really like drafting proposals, but I understand the power of them to put an idea down on paper or a Google Doc so that people can actually understand it. And it's in the same word so everybody can understand what the idea is. It's way easier for AI to make it. Coming up are project charters. I'm amazed how many complex projects marketers go off and do without having a basic understanding of like, hey, what's the objective? Why are we trying to do this? What are we actually trying to accomplish? When are we trying to do it by? Who's in charge of this project? Like the basics, meeting agendas. Nobody takes the time to write them. All really important things, Chat GPT can just do them for you real fast. One of my favorite things is to open up the Mac app. Just like turn on the record feature and me just kind of like word vomit into it and be like, okay, now turn that into a meeting agenda. Now turn that into a proposal. Done.

Now just clarify, when you say Mac app, what do you mean by that? Because people aren't going to. You mean ChatGPT on the Mac?

Chat GPT has a Mac app. If you're a plus user, you can download onto your your computer. Same for the iPhone app. IPhone app works the same way, but there's a Mac specific app. But if there is an iPhone app where you can open up your phone and just speak directly into it, some people like writing into it. But I find that if I have a lot to say, more than a sentence or two, I just kind of like to turn on the microphone record and then just like, essentially like you're dictating to a personal assistant being like, hey, here's all the things to put in there and consider. Write me a first draft of a proposal. And then it goes and knocks it out. It's amazing how efficient.

Okay, right there, right there. A lot of people are like, are you kidding me? I didn't even realize there was a dedicated app on the Mac. I'm just loading it up in my browser tab. And what's really cool about that, and people have heard this on other shows, the voice interface is kind of a big deal, right? So mostly I've been talking about taking your mobile app and walking around with it. But if you're sitting at a desk, the idea that you can talk. Does the voice feature work? Does it talk back to you on the Mac or is it just Input.

I generally just use it as input for now. Obviously, OpenAI has talked about coming out with their new voice model where it's much more engaging and you can interrupt it. The more important part, because right now, like, when it talks, you kind of have to wait. And it's long, it's winded, right? It doesn't go fast, but in the future, it'll be able to talk and you can interrupt it and then actually have a really good dialogue with it. Go. Go to YouTube and just Google OpenAI voice assistant. You could see all the demos. It looks amazing. Like, if you've ever seen the movie her, that's what everybody compares it to. Even Sam Altman has compared it to that movie, of that AI voice assistant in the movie.

And I use it on my phone, and I'm not using the latest version. I'm just using the one that they have. And I'm sure maybe by the time this show releases, they might have already released it, but it's got a slight pause, but it's pretty darn good. Now back to this stage one, right? This personal assistant stage. There are some people that are like, there's got to be a stage below the personal assistant, because I'm not even using it for personal assistance, right? Like, some of them are just using it to, like, do silly little things like, here's a sentence, improve it, or. Or using it as an alternative to Google Search. So there's almost like even a stage zero. But what I love is that your step one is actually more sophisticated than a lot of people are even at right now. So do you have any tips on using it as a personal assistant before we move on to the next step?

Yeah, I mean, I just gave out three ideas right there. Proposals, charters, agendas. But often ones with. For marketers in particular, a big one is repurposing. And I think this is the most powerful way to use AI is to not ask it to come up with content for you. This is where a lot of people find, oh, AI can't write. No, it can write, but just like a genius intern, it doesn't know what to write about unless you give it context. So repurposing content is one of the greatest things AI can do for you. That can help you speed up now, Especially now that ChatGPT has dropped the Zero1 preview, its most recent model. It's way better at this than it's ever been. But you could say, hey, take this podcast transcript, turn it into a blog, here's our writing style, and then, bam, it's written a blog. Don't like it, need it to change the tone or have less headers or more headers or something? Just ask it and it'll do. Do it again. Like, you can keep doing it. Go from blog to social post. You can even take a social post from an influencer, right? Maybe something that's trending in your target audience and everybody's talking about it. Shoot, you can ask it. Hey, take this post, take all the comments from it. Turn that into a script for a podcast episode where we can give hot takes on this. And here's some of our points of view we'd like to add. And then it'll come up with a script for it. Like, its ability to repurpose content is very strong, especially if you're already injecting existing content into it rather than depending on it to come up with the content for you. Love it.

Okay, so the first step is to use your AI, ChatGPT, Cloud, whatever. You're using Gemini as a personal assistant. Now, let's say we've been doing that for a while. What's the next level, if you will?

So, and I got to preface this with this step by step model, something I made up for myself in not knowing exactly where I was going. Right. Okay. I'm like heading out onto. I'm like Lewis and Clark on a journey to find the Pacific Ocean. I know it's on the other side. I just don't know where I'm going. Yeah, and I'm kind of charting it out here. So the farther we get out, the more ambiguous it gets because there's even some parts that I don't know. But before we jump into the more ambiguous parts, this next part is one of my favorites because it's the one where I think the people on the cutting edge are now really starting to dig into the most. So once you get good at doing these OneToOne functions with AI, the next part you want to get to is using AI as process automator. Now, I come out of a marketing ops background, like marketing automation. I was heavy in Infusionsoft, HubSpot. Right now I use high level as a CRM and marketing automation platform. If you're not, then just realize that any type of thing you can have like trigger on the Internet, whether someone fills out a form or responds to a text message or clicks a link and email can be a trigger for something.

Right.

And all these triggers can lead to things. And in the marketing automation space, we've been using that to say someone fill Out a form, send them this email sequence. Well, now you have to think all that stuff can have AI injected into it. There's probably things that you're doing repetitively all the time that could be automated. Once you get good at using AI as a personal assistant, you can think like, well, what things am I asking AI to do over and over again? Maybe we should automate some parts of that so we don't have to keep grabbing that mega prompt that you keep copying and pasting and then going and grabbing the context out of this. And maybe this little part from this word doc of this strategy thing that you need to give it for context because you now you've gotten good at prompting it and you just want to automate the process to save you some time. This is where this process automation part becomes important. And you can use tools like Zapier or High level or even make.com has become a big, big thing that a lot of AI automators are using to inject AI into their processes. But let me tell you about a practical example that I've, I've. I've had. I like you, Michael, am a podcaster and I'm interviewing people half the time and doing the prep work for an interview. It's a lot, you know, it takes some time. You got to do some research on the guest, you got to frame up the interview, you got to come up with questions, you have to write the outline for it, maybe your intro and your outro, the title, you know. So there's a bunch of different elements you have to work through. But I was using AI to do it, and then I discovered that you can actually automate it with something called a custom GPT, which sounds complicated if you're not in the AI space. You've had past guests kind of explain it. So I won't repeat what a custom GPT is too much, because I think some past episodes you've had already have done a good job of that. But they're essentially just custom instructions for a chat GPT and all the major platforms have them. Claude has artifacts, projects also. Yeah, projects. That's what it is. I keep missing that one.

That's okay.

Google has Gems, I believe it's called. Right? That's their, their custom G. They're just custom instructions that follow whatever you put in there. So I made something called My Showrunner that takes it essentially prompts me. I say, hey, I'm going to start a new episode. And it says, cool, Are you doing a solo or a guest? I'm like, Guest, great. You want to copy and paste their LinkedIn profile. And I just go and copy and paste, put their LinkedIn profile in there. It's like, great, let's talk about what Susan knows. And it gives me a little summary of Susan if that's my guest. And it's. Then it asks me again, I gave instructions like, hey, once I give it the LinkedIn profile, summarize it and then ask me this question next. So every time it asks me a question, because I've told it to ask me the question so I don't have to think as much. It's driving the conversation now. It goes and asks me like, great, what would you like to talk to Susan about? What angle would you like to take? I'm like, oh, Susan and I Talked briefly on LinkedIn about this topic. It's like, great. Well, based on the angle of your show, which it already knows based on Susan's history and based on that angle, here's five different ways you can approach that. Which one do you want? 1, 2, 3, 4, or 5? You're like, huh? You look at it and you're like, three, right? And it's like, great. Based on that angle, here's 10 titles that would fit that. I love title two, you're like, but tweak it a little bit. You're like, great. Notes noted. And then it goes and actually runs through the process of taking the guest and my show angle and that title and that angle, and then comes up with the questions and the whole outline for the whole episode. It's just like, bam, it knocks it all out. And I've given an example of what an outline looks like. So it looks almost perfect every time. Since it's an internal doc, it doesn't have to be a hundred percent, it just needs to be 90%. Right. It even writes questions, an email for me to then send to the guest. Once all that's done, it waits until after the episode's over. Because once the episode's over, what do I have a transcript of the episode, right? So I take the transcript, feed it in there. It has the whole context, that whole conversation. It knows the angle, it knows the guest, it knows the title, it knows the outline. And now it has the transcript to go along with it. And it will start creating the assets that I need to finish the episode. The, the podcast description, some social content and other text content that I want to make from it. But that's all automated. Sure, it's still have to manually work through the process, but I don't have to think what used to be a 60 minute long process is now done in 5 minutes, 10 minutes on the back end.

Okay, so just, just so we're clear, first of all, I love this example. And yes, we've definitely covered a lot of different angles on custom GPTs. And for those that don't know, you just need a paid chat GPT account to do this. But in your case, you're just working in your custom GPT. You're not necessarily integrating with other software products. That's a more sophisticated thing. But you are essentially letting it guide you as like a more sophisticated assistant and. Right. That's been trained and it's a process that is automated because it's been trained on what to ask you. Right. And then you can come back to it later, give it the stuff and it knows what to do with that. I love that. That is brilliant. Now, is your custom GPT my showrunner, private or public?

It's public. I've given away the full instructions for how to run this thing. You can copy and paste it into your own GPT. You can even take a custom GPT that's based on those instructions. You just start using it right out of the box. If you go to myshowrunner.com it's gated by email, but once you fill it in, I've given you multiple ways in order to take it. But I've nothing's hidden. All the instructions are there and I hope you take the instructions and customize it for your own show. If you're a podcaster, the reason why I like promise you will get so much more out of it if you just read through it, you will find that it's really not complicated. Like I know how to code websites in HTML and css. It is way more complicated than writing step by step instructions. You'll see it, read it, and you'll find out that it's not that scary. And then you could start using the same format to make your own custom GPTs, which is where I find the power of custom GPTs are. It's not usually taking somebody else's, it's making your own customized to you and your situation, your marketing department, your target audience, all that.

Quick question. How many podcast episodes have you recorded? How many webinars have you hosted? How many client calls or team meetings have happened that are now just sitting there in your files? Here's the big unlock. Those aren't just old recordings. They're full of insights about your audience, your brand voice and your customers actual objections. You just need to know how to extract them. That's exactly what AI educator Erica Stanley is teaching AI Business Society members on January 8th. By the end of this training, you'll have a repeatable workflow that turns every past recording into a content and data powerhouse. One that keeps paying dividends long after the mic's been turned off. This training is exclusive to AI Business Society members, but here's the good news. You can join us right now by going to socialmediaexaminer.com AI and start transforming your old content into into new marketing gold. SocialMediaExaminer.com AI to join right now. What I love about what you're doing is you're allowing someone to reverse engineer or model what you're doing. And somebody who isn't even a podcaster but has a similar procedures for other kinds of tasks like creating YouTube videos or any other kind of content could potentially learn from something like this. Okay, so first step in your multi step process is to use AI as an assistant. The second part of the process is to begin doing some sort of process automation. And for people that want to listen to other we've got a lot of other shows that talk about how to use make.com and do more sophisticated things, but we're just scratching the surface here. But now the third step I'd love you to share what's the next level, if you will, in your process.

So the next step in the process is using AI like a coach or consultant, not just helping you do better and do better on autopilot, but actually using it to help you think better. And this is challenging and it's a little bit more ambiguous, which is why it's another step up. Like the first one's easy, it's one to one and now we're automating the one to one, but now we're actually having to explore a more ambiguous area. But AI is getting better and better at this. So I'm excited, especially with that, that new its ability to reason now is making this way better.

Yeah, and just to be clear when you say it, stability to reason, we should state that we are talking about ChatGPT 01 and 01 mini I think it's called or 01 that's right is the new model that was recently released and it thinks right and it does reasoning and I've not really messed with that. So we can talk about that a little bit later if you have thoughts on that. Well actually let's just ask right now, is it worth using in this Coach consultant role 100.

I mean, I'm actually using it as the default one more and more often because it's just smarter. Unfortunately, you can't bake it into custom GPTs.

Okay.

You. You can't talk to it.

Yeah.

You can't upload documents to it. So it has the O1 preview model, has limitations to it, but its ability to do just about everything you were asking an AI to do is better because it actually thinks through what it is trying to do before it does it. Which, like humans, like, if we think through what we're going to do before we do it, it usually turns out better. So that's the power of the new model. And it just came out last week. Exactly a week ago. Actually. By the time this podcast comes out, it'll probably been like, in a couple of weeks, but hopefully by then there'll be more case studies, a little bit more information out there about it, whether.

We use this particular version or one of the prior versions or a future version. Talk to us about using this as a consultant or coach, because I don't think a lot of people have thought that through.

So you can use it as a consultant or coach. And I use it all the time. I ask it questions, even just tell it my problem that I'm running into. For example, you might be running into a current marketing problem where maybe your boss has asked you to do something difficult or to merge these two things to accomplish a result. You're like, hey, I need more sales and I need it more from social media for the next couple of months and this is the way we're going forward. You're like, but it needs to be email marketing. They're like, no, it needs to come from social, whatever it is. Like, someone's giving you a difficult problem. You don't know exactly how to problem, how to solve it. Tell it to AI just like you would a consultant who specializes in that thing and explain it and give it your synopsis what you think the possible solutions are, and then ask it something. Ask it what considerations you might be missing. Ask it what angles you could possibly take, maybe even ask it for opportunities you might not be thinking about. And I find that AI, when you start asking it questions like this is fantastic even now. And it will get better at helping you think beyond what you've been able to think about before. It'll help you think in new ways, just like pinging a co worker. And I, I work remote most of the time, so, like, I, I miss the days where I could just turn around and be like, hey, Mike, what do you think about this? You know, and just being able to ping them. AI can do that and it can do it remarkably well, but it can go farther than that, of course. Automate the feedback. And honestly, this is something I thought about when I was thinking about examples and I'm like, man, actually I need to set this up because this would work, this would work right away. But let's say you're a podcaster and you want to get better at podcasting. Maybe you want to have ideas that are stickier and maybe maybe you've read some book like Made to Stick. Very popular book, oldie but goodie on how to make ideas sticky. Or whatever book you're reading you want to get better at for podcasting. Well, imagine you ask AI to be like, hey, based on this book, give me feedback every time I upload an episode on how I could have made my idea stickier based on the principles from Made to Stick. So take any book you want to learn and drill into your life and be like, hey, based on this thing that I'm doing, tell me how I could have gotten better based on the principles from this book that I'm loving right now. And you could do that. You can always upload your blog post, your speech, if you're a speaker, your podcast. How could I have made my idea stickier? Shoot, I'm probably gonna do it now because I really love this book and I wanna get feedback. I'm like, this is an automation that I need AI coaching me in every time now, but using it for consistent feedback because that's usually what we're missing now. Eventually it'll get better and keep watching your progress over time. It can't do that now, but that's probably going to come in the future. And then there's one more thing I want to talk about because it's, it's a, I think a little known thing about this, this other AI program called Grok.

Okay, you're talking about Buy X, right?

Yeah, Grok is from X. You know, it's Elon Musk's AI and it's gotten a lot of heat from the media because there's no filters on. There's no, like, there's hardly any security filters on the photos. So if you've seen like Trump hugging cats and dogs recently or weird Mickey Mouse photos, that's probably coming from X's. That's what it's been notorious for. But what people are missing about this platform is it's the only AI model tied directly to a social platform and it has access to real time information from that social platform. So as a marketer or even anybody in business, you can now go to X. You have to have a paid account for this. It's like X Plus or something. It's $8 a month. It's pretty accessible. You can go and ask X and it's performing at a high level and you can ask it, hey, what is my target audience thinking about X? What's trending with my target audience? What are they complaining about? And you can now have real time market research. And it actually goes. And it's probably using some old school tech of just using the algorithm to find them, categorize them, and then feeding the AI to give a summary. But I find it's working remarkably well. And I ask it and it'll, it'll bring up the tweets as examples. It actually brings the receipts. You're like, hey, this is trending. Here's the examples of 5, 6, 7 tweets of your target audience complaining about a certain thing.

I love that. And also Meta has meta AI which is integrated across Instagram and Facebook. I don't think it's real time. I haven't messed with it enough. I'd have to ask some of my more Facebook and Instagram savvy friends. But specifically on X. And you know, if Reddit also does this, it'll be very powerful. There's a lot of people that talk about things on both Reddit and on X that are, you can get the pulse, especially in the AI world. Like X is one of the big platforms that there's a lot of news going on and I would imagine you could use it to kind of discern information. How are you using it? Just out of curiosity, are you just quite literally using it to find news or what are you, what are you doing with it?

I'm using it to find out what our target audiences are talking about. When I'm consulting with companies, it's one of the first places I'm going to be like, oh, I wonder what this target audience means by this? Or what are they talking about with this? So I'm usually using to find trends, I'm using it to find problems that I'm bringing back to clients to find solutions for. And then I'm even asking it like, what could I possibly do? What, how could I match these trends if this business is selling this to this customer, how do I map these trends to the company? And this is where it starts to become more like a consultant or Coach because it has the access to the information. And then you can ask it how to bridge the gap with a blog post or a podcast or with new offers that you might be able to develop for the audience.

So just staying on GROK for a second, everybody. You have to have the little blue check mark you pay for with Twitter slash X, which, as you said, is not very expensive to get access to it are using GROK to do some preliminary work, and then you're bringing it into, like, chatgpt to do deeper analysis, or are you doing it all in grok? And I have not used GROK enough to know if they even have all the sophisticated things that ChatGPT has or how that works at all, or if it even remembers prior conversations.

I'm keeping it all on GROK for now because I find that its model is pretty much at probably just under ChatGPT4 levels, okay? And it's pretty good. And I'm usually just doing there, doing market research, getting some analysis, finding some ideas, and then copying and pasting insights from there into a Google. Google Doc for now. Man, I'd love it if that could happen automatically, but it's not there yet. I haven't gotten to the point where I'm like, building a custom TPT that I'm dumping it into, and then it just runs through the stuff to do it. But that would be the next level. It's a good idea.

One of the things that we were chatting about when we were prepping for this was reviews and how you can scrape reviews. You want to talk to me a little bit about that and how that would work?

Absolutely. That's probably one of the best ideas I gotten from one of my own interviews recently. But I found out that one of the greatest things you can do to get better feedback or to really put AI to work is like, just scrape all the reviews. If you sell on Amazon, you can scrape the reviews there. You can scrape the reviews on Google or Facebook or wherever customers are speaking out about your product. You can go and just copy and paste them. Just, you know, click, drag, copy, paste, and then give it a prompt. It's like, hey, tell me what insights I would have if I were analyzing this as a marketer, or analyze this as a marketer and give me some insights into how I can improve my marketing. Give me the good, the bad, and the ugly, you know, and its ability to actually give you great insights out of it without you having to read through step by step and then quantitatively putting it into an Excel sheet and Shoot. You might even actually be able. I can't believe I've thought about this. I'm usually just asking for insights, but you could actually have it do a quantitative analysis and actually put it in a chart and then have it do the insights part would probably even better.

That's exactly where I was going to. I'm like for anybody who's collecting testimonials and you have lots of testimonials that come in through whatever medium and maybe you've got them in a spreadsheet right now, but you don't know how to categorize them. I would imagine you could drop that CSV file or whatever into your preferred AI and you could ask it to put them into categories. Right. So that you have access to different. Like for example, social media marketing world. We've got people that talk about the speakers, that talk about the networking and that talk about whatever else. Right. So we could have it automatically put them in buckets for us so we can find it more quickly.

That's right. You will run into limits though because of the context window. And this is one of the big limitations of AI right now is it can only take in so much data and, and play with it and understand it at once. Is what it's what I call it short term memory versus it's long term memory that all the data it was trained on but is loosey goosey. Kind of like our own long term memory is loosey goosey. The short term is all the information you give it in a chunk, which for OpenAI is about, you know, a hundred thousand words or so. You know, like one long book. When you have the survey data that you're probably dealing with, it can be over a hundred thousand. So what do you do? What I find a lot of people are doing is doing it in batches. So maybe it creates a short summary of, of an individual's survey and it goes and creates a summary of it and then it takes and summarizes all the summaries in order to get around that context window. Lots of companies are getting around it that way and it's really interesting way to do analysis and to get insights from a large data source like surveys from, from a large conference like yours.

I have not messed with Google Gemini, but my understanding is it's got a massive context window. So that might be worth people experimenting with as well.

Absolutely.

When you're asking your AI to be a coach or consultant, do you tell it to act as if it's a coach who specializes in blank or do you just go right into your discussion and ask for feedback. Do you have any wisdom on how to kind of tell it to act as a coach or consultant?

You know, I've been doing that less and less because I find it intuitively knows how to coach me through the thing. There is times for it. Like, even if you've learned how to break down a prompt, one of the early things you learn in prompt engineering is defining a role for. For it to take, and that can change it. But if you have a very specific way you want the coach to approach you, like, do you want it to act like it's a CMO versus a CRO or someone more in a sales role? Like, it just depends what kind of coach do you want it to be? But I find, like, if I'm asking a marketing question and I want it to coach me through how to think through this marketing thing, generally it's going to coach me correctly as like a marketing coach. On my custom GPTs, I'm still throwing a role in there just in case. But I find where the roles make the most difference is if I wanted to come at a specific angle and that angle would include a role like, oh, like, I wanted to ask me questions more in this vein, not that vein. Well, what role would that normally be? So it's. I'm almost doing it more in a troubleshooting process to make it better at that point now.

Now, one of the things that I have discovered, which I'm sure you already know, but a lot of our guests may not know, is that when you're showing off chat GPT on your phone to a friend and you ask it to act as if it's something like a chef or whatever, and then you see it updated memory, or even when you ask it to do something that's a little off, sometimes it shows updated memory. And that's scary because it's essentially telling the chatgpt from now on, this is how you should behave. And it can get kind of convoluted and confused, and you got to go in sometimes and look at your settings and clear some of that stuff out. Have you found that to be true as well?

Oh, yeah. I regularly have to go through there and clear it. It's getting more. It's exciting, but it has that eerie feeling to it. Like, I built this custom GPT. I homeschool my kids, and I was the bottleneck because I wanted them to write down a summary of what they were learning every day. And I got tired of grading it. So I created a Custom GPT. I literally just take a photo of their handwritten journal on the topics and built a custom GPT with the grading rubric on it. And I just take a picture. It can not only read their chicken scratch. Writing is my. My. My youngest is in fifth grade, and he's writing his writings. It can not only read it, but analyze the content and give feedback that's appropriate for a fifth grader. It actually writes the feedback directly to him. I just copy and paste it and give it to him. What was scary, though, is I usually tell the custom GPT who it is. Is it my daughter? And I give or I give it the name of my son, and I'd forgotten to give it the name. It already graded my son's papers. I forgot. But I then took a snapshot of my daughter's, and I didn't ever. I got to give it the name, but it already knew, huh? It gave the feedback directly to her by name. And I was like, how did you know that this was my daughter's?

It's got to be the handwriting that's.

It was like, well, the handwriting was more so. This is the handwriting of a seventh grader versus a fifth grader. And I know you have two kids. This was a fresh GPT incident. So it's not like I had done a bunch of grading for her in this same conversation. It just kind of knew, and it kind of creeped me out for a moment. I was like, like, what is it? Remembering. It, like, remembers. Yeah.

Okay, so we have two more steps. We just finished this consultant coach step, and we've got this next step, which is the forecaster step. And you mentioned earlier that you're working some of these things out. So some of these things aren't quite there yet. But why don't you explain what it is when you're using AI as a forecaster and, you know, what you believe it will be able to do?

Theoretically, we know that AI will be able to predict the future much more accurately than humans is it'll be able to take in more data and run the numbers and do the math faster than a lot of us, right? I mean, we all know, like, when we. Probably one of the first, earliest thoughts as business people when we heard about AI, we're like, oh, Wall Street's gonna love this, right? Like. Like there's implications for there for it, but that's gonna be true for marketers, and it can be true in small ways now, but you'll be actually be able to use it to actually predict your Sales for next year. You'll actually be able to go into AI and imagine just giving it your passport performance on revenue, market data of how your market's currently doing, economic data, and then your marketing plan. Have it do some analysis on your marketing plan and have it predict like, oh, this is probably a good, better, best forecast for your current marketing plan. Right? And it'll be able to do that. We're not there yet, but we're also not too far off. Again, the big thing that we're missing is the context window. As the context windows grow and it can take more data in at once, it'll be able to do this kind of stuff. And some companies are doing it already. For example, a friend of mine is the CMO over at a company called Known. Well, and they're doing this more for a customer service place than marketing. But it's really helpful to understand how they're doing it and how it'll probably come over to marketing eventually. But they're using it and hooking up AI to your Slack and social media and your company email so that it can monitor how a service based company is doing with its clients. If you ever worked in a service based company, you know, like sometimes you think a client's doing well and then all of a sudden like red lights are going off everywhere. They're like they're churning. Why? Oh, and then you do a deep dive on why they churned later and you're like, oh, if only we would have seen like, oh, there was that little conversation on Slack but nobody knew about that. They posted about it on social and somehow we didn't catch that. And gosh dang it, by the time the email came in hot, it was too late. Well, AI can be keeping tabs on that way ahead of you and actually organizing it and doing the crunch work of just keeping track of social posts. Is this related to our company? Is this related to our company asking those questions? Very good at that. And then chunking it to get around that context window problem, to be able to surface issues before a customer is starting to send that nasty email, giving the red flags that it's going to churn soon. Right? So that's how AI can help customer service. But those kinds of things are coming from marketing as well. Being able to actively, not just you prompting GROK and asking for customer feedback, but actively listening for you and being like, hey, there's this new thing trending on TikTok that would be perfect for you to jump on now. Do it now before the trend is over which trends go fast on TikTok. Right. But how much would you pay for a tool that would show you just right when it's starting to trend and it analyzing it to see if your company would actually be a good fit for this one, Especially if you're a B2C company, that would be golden to get in early on a trend and your product actually would fit perfectly in the trend. That would be something that's coming. That ability to forecast something before it's passed is a huge thing that will come with AI eventually.

And then what's your fifth level, if you will?

The fifth level is the most ambiguous. But I, I'm convinced that it's, it's here and it's the ability for AI to create original value because I had to put something, it's almost like this catch all bucket. If it doesn't fit into one of these four other steps, you know, it's not just one to one tasks. It's not automating existing work, it's not forecasting it. It's not helping me think better as a coach or consultant. What is that? Well, it's, it's creating original value and that you can actually accomplish some of this in small degrees. Right? Now that course that you mentioned earlier, this AI fundamentals course I created, I actually do to a small degree. Every course is just a five. There's five videos that walk you through some of the elements that I've talked about like in prompting and all that kind of stuff here. But each lesson is actually customized for you, so it's hyper personalized. No, email is the same because on the form I ask what's your job title? What industry are you in? As in like what are you selling and who are you selling to? And based on those three pieces of information, every transcript of the lesson is sent to AI and the lesson is customized just for you. So that the lesson on prompting, for example, is just going to be different for a CMO in a beauty industry selling to tweens, cosmetics to tweens, than it is going to be for an email marketing specialist selling SaaS to a CTO. Right? It's just going to craft different lessons for them. It's hyper personalization, but it can do it at scale because I wrote the instructions and the prompt once. And now every time someone runs through that course, it's hyper personalizing it. I think we'll see a lot more of that into an even greater degree. I make kids books like how to become a marketer when you grow up. But like, like it's a fun little thing if you're in content marketing to come up with kids books for your target audience, even if they're adults, because they probably have kids. But now I worked with a kids book author to make this ChatGPT, especially that, that 01 preview that I was just talking about. I gave it a prompt to come up with this exact book and it was dang close, really, like coming up with the content for this. But imagine being able to do it at scale now and you can come up with the book for every client with their names and maybe their kids names in it. Like that's the kind of stuff that you can now do. That's original value that I'm always coming back to. I'm like, how could I curating original value here that's not just enhancing something else. What is now possible with AI, And I think that's the question we have to keep pushing ourselves with, is not just what can we do better and faster with AI but what new things are now possible that were not possible before. And as someone who, I don't know, maybe that curiosity me, maybe I'm just the sci fi lover that likes to dream of like, what if that's the part that like really gets me excited about this whole topic and then making it practical to actually implement today.

Dan, this has been a really fascinating discussion. Rich with little nuggets throughout and also rich with really good blueprint framework. Whatever phrase you want to use of how we can, if we're not doing X or Y, for lack of better words, we can kind of work our way down the path or up the mound, whatever metaphor you want to go with in order to kind of take it to the next next level. With AI and our work, if people want to discover more about you and your business and all the stuff you got going on, where do you want to send them? And also what's your preferred social platform if they want to connect with you online?

Absolutely. You can find all the links to my newsletter and the podcast@aidrivenmarketer.com and my preferred platform is LinkedIn, so you can go to LinkedIn.com in digital marketingdan and that's my profile. I never turn connection requests away, so connect. Would love to hear if you found me through this episode.

Dan Sanchez, thanks for coming on and enlightening us with all your wisdom.

Thank you so much for having me.

Hey, if you missed anything, we took all the notes for you over@social mediaexaminer.com a23 and be sure again to follow this podcast on whatever app you're using. And if you love the show, would you share it with a friend? If you want, you can tag me. I'm Elsner on Facebook, Stelzner on LinkedIn, and ikestellsner on X. And do check out our other shows, the Social Media Marketing Podcast and the Social Media Marketing Talk Show. This brings us to the end of the AI Explored Podcast. I am your host, Michael Stelzner. I'll be back with you next week. I hope you make the best out of your day and may AI help you become more successful.

The AI Explored podcast is a production of social media examiner.

AI Business World, early bird pricing ends on January 16th. You can save $300 on your tickets. Register at AI Business World Live today.