Loading summary
Transcript70 lines
- [00:03]
A
Welcome to the AI Explored podcast, helping you put AI to work. And now, here's your host, Michael Stelzner. Hello.
- [00:14]
B
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 put AI to work. Today we'll explore how to productize your expertise with AI. My special guest is an AI strategist who helps entrepreneurs leverage their expertise and AI to grow their businesses. Her podcast is the entrepreneur school in the AI era. She's co founder of Waiv by Gravia Studio. Kelly Sinclair. Welcome to the show. How are you doing today?
- [00:51]
A
So great. So excited to chat with you, Michael.
- [00:53]
B
Super exciting to chat with you as well. I'd love to hear a little bit about your journey. How did you get into AI?
- [00:58]
A
So I feel like this is a hilarious turn of events because I am not the typical started with ChatGPT on day one kind of guest that you often have on this show. For me, I have a background in communications and public relations, and I actually, the first time I tried ChatGPT, I was like, wow, this is fast, but it's kind of garbage, right? So as far as the output, I was not very impressed. But once I learned how to actually train it with. With brand context, give it what it needs to know about me in order to produce good results, like, you know, who your audience is, how you like to speak, your style, those kinds of things, then I saw the opportunity that I could actually use generative AI to help my clients with implementation. So I was working as a brand and marketing strategist, helping entrepreneurs to get visible, figure out how to get in front of the right audiences and reach their ideal clients, implement. And we would build these strategies and then they would just not do anything with them. So I saw the opportunity to bring Genai into the game and actually build tools that would help them take action. And so this was a game changer in terms of getting results for my clients, and it was a way that I could help them to just improve and deliver on the strategies that we were creating together.
- [02:21]
B
I mean, I know since ChatGPT came out, we're coming up on four years now. So just out of curiosity, when did you start integrating these AI services into your client offerings?
- [02:29]
A
Early 2024, probably. So, yeah, more of a late adopter myself. I mean, it's still early. I mean, I ended up getting to be like a top 1% user of ChatGPT once I started diving into it because I was like, okay, this has so much potential, especially when we think about how our clients can actually access and, and implement and do these things. So. But then I started running up against some of these hurdles that I know we're gonna talk about later on delivery and how to actually get these tools to clients, which sent me down this other rabbit trail of having conversations with a local software developer who, long story short, is now my business partner. And I am now an accidental tech co founder and we work with entrepreneurs on productizing their expertise with AI.
- [03:14]
B
I love it, I love it, I love it. Okay, cool. So such a great journey and so many people I think can relate to this. Cause I mean, we' to look back in the future and say we're still really early. I mean, even people listening to our voices right now, folks, it is really, really early. Even though the whole world is talking about AI, it is very early days. I mean, it's kind of like social media in year three and a half, you know what I mean? Like, the whole world is going to change in a pretty dramatic way and it's not too late. And that's why I'm really excited to get you on the show. All right. When it comes to actually this concept that we're talking about, which is to somehow take your insights and knowledge that you have as an expert and, and leverage AI in some kind of capacity, what are some of the misconceptions or false beliefs people have when it comes to leveraging AI in this kind of way?
- [03:57]
A
So, Michael, I think that one of the biggest misconceptions is that if I take my expertise and put it into AI and make a product that somehow devalues me as the expert. And I think that we can really unpack this way of looking at it, because I really believe it's actually the opposite. I am a firm believer in human first AI adoption and that we should be incorporating AI into what we're doing so that we get to extend ourselves and our experience as experts and really leverage that to allow us to do more of what only the humans can do.
- [04:39]
B
And I'm with you on this for sure. I've evolved my thinking on this as well. In the very early days, I'm like, I guard my intellectual capital, my ip, my whatever the heck you call it. My thoughts like, guard it, guard it, guard it. But then once I began to understand how AI can be a massive enabler, that allows my many decades of experience, in my particular case, being an entrepreneur and also Being someone who talks a lot about marketing, to leverage that in a pretty powerful way, it just, it's an unlock. So when we actually are able to properly do what we're about to do today, right? When, if people pay real close attention to what we're going to talk about today, what, what's the upside? What are the benefits that are waiting for them?
- [05:20]
A
So two things I really want to talk about here. Maybe first we should contextualize this in the concept of like how things are changing because AI has come into the scene. So I think for anybody who is a consultant or coach or service provider in any kind of way, if you're in the knowledge business, as they say, and you teach other people how to do what you do, or you use your brain to help people, other people, then the, the game is shifting, the expectations are changing. And what we have now is the opportunity of what I like to call expert backed AI. And I think it's important to differentiate that from generic AI. So if somebody says, oh, I can just go use AI to do whatever it is you teach me how to do, the answer is sure, but you don't have the ability to like validate if the results are any good. You don't know if that's actually producing something beneficial for the user versus when you as the expert who has, let's be honest, years if not decades of experience of wins, of losses, of failures, of knowing exactly how things work, what your frameworks are, that is like hard won expertise and that can be embedded into AI tools that actually allow your clients to use that thinking. So digital courses are like teaching people how to think, but AI tools are allowing people to use your thinking and that's where implementation comes in. And so this expert backed AI is really the opportunity and how things I'm seeing are going.
- [07:01]
B
I know there's a second thing that you want to say here, but I want to double down on this a little bit. As you were speaking, I was thinking to myself, okay, anyone can pick up a camera and take a picture, but it's people that have been extensively trained on how to frame a picture, how to get the lighting, how to get the composition, how to get the subject to do the things that you want the subject to do. So you could say that the camera is effectively like AI, it's a tool, right? But when you bring all that unique background and expertise to the table, you're able to take this very powerful tool and do something with it that others do not know how to do with it. Just wanted to throw that over the fence, because I thought that was a really good analogy because I could go up against the best photographers in the world, quote unquote. Right. But I'm not going to know what they know. Right. I'm just, I just have access to the same tools that they have. But what I have is experience and knowledge and insights that are allowing me to use the tool in a powerful away. That's why I really like that analogy that you're talking about. This expert backed insights, right. Taking what you, what you've developed in your own brain and from your own learned experiences and applying it to AI allows you to do things no one else can do.
- [08:03]
A
Oh, I love that analogy. I'm going to pocket that and use that one because it makes total sense to me. Absolutely.
- [08:09]
B
Okay, so the question was how? What's the upside? What are the benefits here? Right. And you set this groundwork on this expert backed AI concept. What are some of the benefits? I know we're going to get into the how we do this in just a minute, but I just want to plant the carrot in front of the horse, if you will, for us being the horse. Like what is waiting for us on the other side when we get this
- [08:27]
A
right big analogy guy, Michael. I love it.
- [08:30]
B
They're just coming out today.
- [08:31]
A
I don't know what else can we come up with while we're on this podcast? So I want to say that like, so we're talking about human first AI adoption. We're talking about expert backed AI. And what does that allow you as the expert to do? Well, what our clients are finding is it's actually allowing you to go deeper with your clients to actually leverage and be more in the expertise that you have. You're having more nuanced strategy conversation because your clients aren't coming to you with a how do I get started? They're coming to you much further along where you're able to really apply that outcome and give them better results. And I think the other thing here too is to consider the fact that, you know, with digital courses as an example, I did some research into this and Thinkific had put out a study in 2025 where they talk about completion rate of digital courses. Right. And it's abysmal, to be honest. 10 to 20% of people will actually finish a course. Like hands up if you have a digital course graveyard. That is me too.
- [09:35]
B
Or hands up if you bought a course and never finished it. That's.
- [09:37]
A
Yeah, exactly. Yeah. Never finished it, maybe didn't even start it just felt Good, because you bought it. You're like, I have this information now. That's great. So the completion rate thing is, that's not what you're here for. As an expert, you want to help people. You want them to get results. You want them to have the opportunity to apply what you know so that they can, you know, you can make an impact in the world. That's really what a lot of the people that are, I'm sure, listening to this show and that I circulate with, want to do. And that completion rate, when you add AI tools into it, it skyrockets to 70 to 80%. So huge improvement in terms of how your clients are actually getting results. And that whole ability of them to use your thinking is a big part of that, too.
- [10:22]
B
Okay, so a couple things I want to clarify just real quick here. We're talking about how basically to get AI to effectively monetize your expertise. Is that really what we're talking about? Because I want to make sure everybody kind of knows the direction where we're going with this. Like, if that's the case, are we talking about using AI to effectively create tools? Tools that we're going to sell to our clients? Because I think that's the part maybe that we're missing here, that if that's part of it.
- [10:46]
A
Yeah, absolutely. And that's why I kind of said the word product ties, too. And so almost like giving you the idea of, okay, so if I have a digital course or if I was to teach my expertise through a digital course, that's what that pathway would look like. But now that we have AI on the scene, we have to think about how can we actually add AI into our offers. And so there's a hybrid approach where you're building AI tools that your clients could potentially subscribe to or use. And then potentially, you're also adding layers of coaching or office hours or things like that to bring the human component to the table. Because I think really both are important. And again, we're not just fully outsourcing to AI.
- [11:31]
B
Okay. So I just want to explore this a little bit before we get into this too deep so people can wrap their heads around this. We all have expertise that we have developed over many, many, especially those of us that are gray hairs that have been around for a. And that expertise might be something that we sell in the form of a course or in the form of a membership, or in the form of consulting, strategy sessions, all that kind of stuff. And what I'm hearing you say is that if you can Properly somehow use AI to create some sort of a added value service or software or whatever. Right. Something that will allow you to provide even more value to the people that you're serving and, and also free your time for the higher order things. This is where you can actually scale a business. Is that kind of what I'm hearing you say?
- [12:22]
A
Yeah, absolutely. Like I call this bot squads and we're going to talk about like how to actually build these and what different opportunities there are with the technology that exists right now. But I mean, I want to like boldly say bot squads are the digital course circa 2026. Like this is where we're going. This is what people are expecting. Your clients are expecting to be able to do things with the ease of AI. So this is the opportunity for your monetization to blend this together and create tools that they can use. And we'll get into exactly how to do that.
- [12:56]
B
Folks that don't know this, I mean, I talked to at least over a hundred different people on the two different shows that I do. And there's a big trend going on right now where people are moving away from courses because that knowledge is becoming commoditized. Right. And AI sucking up all that information. But they're moving towards this new model which is to have some sort of a software product that provides super instant access to insights that were uniquely created by folks like Kelly or me or others. That is a reason for them to keep on a subscription model. Right. So there's a movement towards this concept of a monthly subscription or a membership, for lack of better words, that doesn't just include access to knowledge and access to live people, but it also includes access to tools that were specifically designed to solve whatever the problem is that you ultimately solve. And what's cool about this concept is this is something that's sticky. What I mean is that a lot of people now are going to keep memberships because they don't want to lose access to these tools. Would you agree with that?
- [13:59]
A
Oh, absolutely. And you nailed it with the AI has commoditized knowledge. So it's not the knowledge itself that's valuable anymore. It is actually your expertise, your lens and the way that you tackle whatever it is that you do. And that's the thing to figure out how to package that up.
- [14:16]
B
Perfect. Okay, so now we're going to transition into like how to do this. Let's start with the basics. Where do we start?
- [14:22]
A
So the first step is identifying the opportunities within your frameworks and your processes. So I will do a little Bit of a teaser that I'm going to share at the end. A tool that I built to actually help you do this, called the AI Tool Launch Playbook. But I'm going to give you some questions that you can consider that you can throw into your favorite AI LLM to have a conversation to try and extract this. Apparently you just need to also talk to Michael because he's a big framework guy. You can pull this together. I was never really an expert at this, but identifying the pieces of the puzzle is the first step. So question number one is around repetition. So where are your clients asking the same questions over and over that you can create a tool that would help them with that piece of what it is that you do? The second question is implementation gap. So this is like what I identified before with my visibility strateg. I saw them not taking action on the visibility, and then I was able to build a tool that would help them with momentum and actually understanding what they were doing and how that was working and what the ROI was on that. So where do your clients regularly get stuck that you could provide something that would help them get over that? Hurdle number three, is there a skip zone? So what is the thing that your client's like, I don't want to do that. Like, I'm just not gonna. And you know that it's a really important part of your process that they have to take those actions in order to get the results that you're helping them to achieve. So what does that skip zone look like? And what would really support them? And even just sometimes it's about, like, getting to a first draft of something where they can't get over that blank page syndrome that they're having. Right. And they need to take action there. And then the last one is around a confidence gap. And so this is like mindset. Where do they get stuck? Where are they, you know, getting all in their head that you could provide some guidance and support that would make them feel more confident to move forward and take action on the framework that you have?
- [16:28]
B
Okay, so let's explore these a little bit. We talked about four things here. First of all, these are all things that anyone who's listening could explore as possible starting places to potentially develop something with AI. You said repetition. What are the things that people keep asking you over and over and over again, right? And then you said implementation gap. Whenever you work with someone or someone works with you, what's the area where they. Maybe it's a momentum issue. I always call it a flywheel effect, where sometimes Getting started is the hardest thing. Right. So some sort of thing that could help them get off the mark to get that momentum, as you refer to it, the skip zone. I really like this. There are always going to be things that some people just don't want to do because maybe they don't. They don't know how to do it. Right. And it's like, but you know how to do it. And when you teach it, you presume they know how to do it, but they don't know how to do it. So maybe you could develop something that helps those that don't know how to do it to get over the next hump. And then this whole confidence gap, this mindset concept. Right. Which is this belief system that we have about ourselves that may or may not be active accurate. So these are all kind of places to kind of use the magnifying glass, Is that what I'm hearing you say? And look into our experiences and say, which one of these things resonates with me the most. Any tips on how to know where to start here? I mean, because these are four different things, we don't want to tackle them all, presumably in the beginning. Right?
- [17:41]
A
Yeah. You could create AI tools for any and or all of these pieces, honestly. But I think what you said there makes me want to remind everyone that when you're the expert in science, something you think everybody else understands how to do it. And that they also like, oh, this is in my brain. I forget what it's like to not know how to do this or to not understand where I'm at. And that is the opportunity to reflect on for yourself and realize that, oh, no, this isn't easy for other people. And something that's really interesting about some of the clients that we have right now is what they're doing from a pattern perspective is I feel like they are creating tools that help their clients perceive these things to be a heavy lift. They think that it's hard to do the thing the expert is teaching them, and that is what's keeping them from doing it altogether. And now we're able to present, well, there's AI involved in this. So actually it reduces that lift. It is not as much of an objection to get started anymore because this has been brought into the process.
- [18:49]
B
So tell us about how you first did this, because I know that there's
- [18:52]
A
a story there what my first AI tool was.
- [18:56]
B
Yeah. Just so people can wrap their head around an example of how this might have worked for you.
- [19:00]
A
Sure. And I have a couple really great client examples That I can explain as well. But for me, the first one I created, her name was Valerie, the visibility auditor. And so what I noticed is that I could tell my clients, you know, the best thing for you to do is get on podcasts or go to networking events or do collaborations with other people. And they would still be like, well, I gotta post on social media every day. And so I trained this bot with a framework that would help to identify what are you doing. Like, let me know what you did this week. Let's go through that together and kind of grade it on roi. So if you spent all week, you know, prepping for this one big huge podcast, hello, it's me, then that is a good use of your time. Good ROI can come from this versus I spent all week trying to put together one Instagram reel and I just posted it, had no engagement, that kind of thing. So my tool was actually evaluating that, helping them to get momentum and redirect them into how to continue to apply the strategies that we built together.
- [20:13]
B
Okay, perfect. I love it. So, so at this stage, we're auditing, for lack of better words, our processes to try to find something that we could have AI do for us and maybe even better and faster than we can do for ourselves. Is there anything else we need to talk about here before we move on to the next step?
- [20:30]
A
Well, I think we can talk about the framework of how to actually structure the tool. And for me, that's. I call it the IPO framework. So that is input, process, output. So what's really interesting here, if you think about it like at a super high level, is that you can create a user agnostic tool that can still produce user specific customized outputs. So that's when you break it into those three steps that you can understand that it's the input that changes the output and then the process stays the same for every single one. So that's, that's where the user agnostic piece comes in.
- [21:14]
B
Just explain, when you mean input and output and process, are you talking about like the person using some sort of tool, they're inputting information and then there's a process happening and then the output that comes out of it. Is that what you mean by that?
- [21:26]
A
Yeah, absolutely. So the input is what does the user have to bring to this conversation, to this tool, in order to get the output that you are trying to help them get? So I'll give you a specific example in a second, but maybe we should go through like what would be included in the process and what kind of output you might want it to have. So for the process you want to identify the goal, what is its job and again, keeping that specific. So like just like want to give a moment to chatgpt and custom GPTs, because a lot of people are going to. Custom GPTs are dead. Well, I don't believe that the concept of a GPT is dead. Like having a specific goal for a tool helps it really stay in its lane and do a really good job. So you have that clarity around what its goal is. Then you need to give it instructions on what you want it to do and you need to give it resources. So what is all the training that it needs to do its job properly? This might look like transcripts from coaching calls. Your all of your content inside of your courses or your modules, whatever templates or frameworks or worksheets or examples that you have that would help get the output. Again, what is it that we're trying to achieve with this tool? What is needed in order to get there? So input, process, output.
- [22:48]
B
Love it. You know what most marketers are doing manually? Ad creative writing the copy, coordinating with the designer for the images, waiting on video edits, managing revisions every campaign, every cycle. I wanted to fix that for our members. So on July 16th, I'm bringing Caleb Cruz to teach inside the AI Business Society. He's going to show members how to build personalized AI pipelines using Claude code that runs the entire creative process on its own. Copy, images, video revisions, all customized to your brand. And you don't need to be a coder to do it. I'll be in the room for the entire session. I'm in every session and I can tell you members are going to walk away with a clear buildable system that they can put to work that very same week. If you can't make it live, the full recording and resource kit will be waiting for you. Transcript, slides, key takeaways, the whole shebang. This is what we do every month inside the AI Business Society. Come and see for yourself and don't miss this training. Join@SocialMediaExaminer.com AI Again SocialMediaExaminer.com AI okay, so at this point we have thought about like some of the structure that you just talked about and we've hopefully identified some sort of area where I want to develop something. What's the next step?
- [24:11]
A
So maybe it'll help just to give some examples and thinking about this actually. So I want to talk about two that are clients of mine and they're doing very different things. So I like to show how it can be so different, but the same process, like framework is applied. So one person, Nicole, she is a PR coach and a journalist, and her clients want to get media, but they find it very hard to do that. They think, okay, I've got to do all this research. I've got to write pitches, all this sort of thing. So she built a bot squad that actually does the intake process. So it asks a bunch of questions of the user, and then it creates a customized media messaging document. So that essentially becomes the input that is needed for the next bots which do podcast research. So that will apply the user's information in order to select appropriate podcasts, not just like the top podcasts in the world, ones that are a fit for the type of business, the type of things that the user actually talks about. And then it will go to the next spot which does podcast pitch developments. We'll actually write that pitch in the voice of the user again, because that messaging guide exists, and it can create that customized output no matter what kind of user comes to the table.
- [25:40]
B
Very cool. And in this particular case, this is not something that Nicole created for herself, or is it something she created for herself, or is it something she created for her clients so they could just autonomously go do it without her? I'm just curious.
- [25:53]
A
Both, actually. She uses it for herself to get media because a lot of people who are experts, they're like, here's what I do, and here's how you can do it too. Right? So that fits there. And she actually told me she had reached out to a media, like a TV outlet. So she also has, like, a TV segment, one that does this too. And the producer was like, this is the best pitch I've ever received. Which is incredible, because she's like, that's the part where your ego gets hit a little bit. As the expert, you're like, oh, I'm the one who is supposed to be really good at this. And the bots are doing such a good job, and that's great. And the way that she does it with her clients is she actually gives them access to this tool and they can use it regularly, and they get some touch points with her as well to, you know, review things maybe before sending pitches if they want to, just to keep the human in the loop, which I think is really important.
- [26:44]
B
I love it. You said you had a second example.
- [26:46]
A
Yeah. So I also have an example of Michelle, and she is a messaging strategist. She says one of the biggest objections she has in her own sales process is that people Are like. Like, oh, messaging. Like, that's gonna take months to get that right. I better make sure I don't have any other projects on the go. I need to be able to, like, clear the decks and have all this time. And she's like, well, now we actually have the ability to use AI to help with this. So the first step in her process is to do voice of customer research. So she has her clients do the research. They bring it to her bot squad, which she calls Moxie. And Moxie will analyze that and pull out of the key pieces and turn it into the types of marketing messaging that her clients can now use and know that is actually going to be effective, which I think is really cool. Because she's like, I have a doctorate in communications. My clients don't want to learn how to do research analysis. They don't want to, and they shouldn't have to. But I can give them a tool that will do that in the way that I'm trained on it.
- [27:52]
B
Love it. Okay, now there's some people listening that are like, okay, I want to do that just so I don't have to do that part of my job anymore. These examples that Kelly just shared here, you could design these and just use them internally to free up your time. But what I like about what Kelly's recommending here is actually we productize these into an actual thing that we can make money on, which I think is really cool. So it does beg the next question, which is how do we actually build these kind of things? Because it sounds awesome and it sounds hard. So let's talk a little bit about that, because I'm sure people are really excited. Like, a lot of people, like, have trained up, you know, cloud projects or custom GPTs on their insights, but they have no idea how to actually turn that into something that they can sell. So let's talk a little bit about that part of the process, because I'm very fascinated about that.
- [28:39]
A
Yeah. And this is where I'm like, oh, if only there was, like, something really great, which is something that I'm working on for sure. But I will say, like, the whole thing about AI is that it's always evolving. There's always new tools. There's always new capabilities within the tools that we have. So let's talk about, like, the maybe three core ones that you might actually, actually use to execute this, or at the very least, to create, like, a proof of concept for yourself using the framework, the IPO framework that we just talked about. So the first one would be a Custom GPT. So building a custom GPT using that framework, with the thinking about what the user has to bring in, thinking about the process that GPT should run them through and then thinking about what the output looks like. So building a custom GPT is very easy to do with chat GPT because it will walk you through exactly how to do it. Just have a conversation with it, apply the frameworks from this conversation and you can get a bot out of that. The thing I don't love about that is that you get one bot that does one job really well and that means that there's no connectivity if you have a multi step process and you want to be able to bring AI into a lot of different elements of your process. And I'm seeing like some of our clients, they're like have 17 different bots and what they're ending up doing is making a PDF with a bunch of links that are saying go to this one for this and then take the output and go to this one for that and do the next thing and follow it along. So that's one of the issues with chat GPT, the other one being security, when you're actually sharing it. So right now you can send a link, you can make it public, you can allow people to use your GPT, but you don't have the ability to like gate that, turn off access if you're actually wanting to monetize. This can kind of more like on a subscription or something like that.
- [30:30]
B
Yeah. And before we move on to the next thing, I do want to say that like we've interviewed a lot of people on the show, this is episode 115 I believe. And there have been people who have, you know, talked about how, I think they've just taken a link, right, and they shared it in their members area, right. And they just kind of say please don't share it. Right. That's kind of the hope. But the problem is if somebody leaves the membership, they're still going to be able to potentially use this custom GPT. So I, I see the problem. This is not going to be something that's going to get someone to necessarily want to stick with you month after month after month. And Also with custom GPTs, I've heard lots of stories about, sometimes they break, you know what I mean? Because like the models shift and they change and all of a sudden like your custom GPT doesn't work the way you thought it worked. And I also know that OpenAI was thinking about building a great marketplace for these kind of things, but it Seems as if they, they've put their attention elsewhere. So it used to be like the only option in the early days, but it seems like now things are changing. Would that be fair in your professional opinion?
- [31:33]
A
Yeah, absolutely. I think you nailed kind of all of the same frustrations that I was having when I first started as well. And what I see some of my clients doing, which is like, you know, they're password protecting their GPTs, so this is one kind of hack that you can do, but then you have to like, remember to change the password and share the password again. And so it just becomes like a management challenge as well.
- [31:55]
B
Okay, so what are the other options?
- [31:57]
A
So what we love about Claude skills. So move over custom GPTs. Claude skills are here and everyone's thinking, this is the new hot thing. And I love it. I love it because it operates differently in terms of it creates what nerdy AI version of me now says, which is multi agent orchestration. And that means that, that a lot of things can be happening at the same time. That connectivity issue that you have with custom GPTs being siloed individuals changes with a clod skill. That can be multiple steps happening and accessing information from different places within somebody's Claude account. So you can apply this to a cloud skill, you can sell that Claude skill. And so lots of people are doing that too. They're just, here's a product, here's a Claude skill. You now take that and you plug it into your own Claude account and you can now use my expertise throughout your own Claude.
- [32:54]
B
Okay, that's fascinating. Let's explore this a little bit. The benefit of a skill is, and I've learned this mostly because of all the guests I've had on the show, is it's portable, which is really cool. So you could take a skill. And Claude was the first company to come out with Skills, and then everybody else now seems to support Skills, Chat, GPT, Gemini and many others. Everyone is embracing skills. So they're portable, meaning you can take them, like if you don't use cloud lot in your Gemini house or whatever, you can take them over. And they're fundamentally very similar, which I think is awesome. So I guess the one advantage to going with skills is that you're going to update and change the skills, which might be a reason why somebody might want to maintain some sort of a subscription model, is because as your quote unquote skill evolves, right, you might have new versions of a skill. Kind of like, you know how in the olden days you would pay one price For a piece of software. Do you remember those days like you buy Photoshop and then it would last you for years until the new version came out. So I think we're moving towards an era where there's going to be versions of Skills and if you want to get all the newest versions of the skills, you're going to have to subscribe. Right. And there's a lot of people going towards this route with software. Right. So it's a good model where like, hey, you pay an annual fee and you're going to get all the updates that I make to the skill. So this is almost rethinking like in the beginning when we start creating skills, it might be very simple, but we could evolve them and make them more sophisticated over time. And therefore if we keep upgrading skills, there might be a reason to keep people on a membership. I'm just free flowing with you a little bit, but is this where you see it going?
- [34:21]
A
Potentially, yeah, for sure. And also the capabilities within side the quad interface and connectivity, I would say this is like next level where skills can actually pull on other tools. So if you're a business coach who's teaching people how to create digital products or something like that with AI, you might also be like you want the tool to pull in in your airtable database and do something with your email service provider and do something else with your, you know, video creator, all of those things can happen. And as AI evolves, it gives you the opportunity to continue to upgrade that and, and evolve it with it. I will say though, the one thing I don't love about Skills is again, it's like handing over a zip folder of your expertise. So this becomes the conversation around your IP and how that really is prot and wanting to keep that lockdown or deciding that maybe it doesn't matter.
- [35:25]
B
Yeah, it's really interesting. I love it. Okay, cool. So we've Talked about custom GPTs, we talked about cloud skills. What's the next level? I know we're leveling it up each time, right?
- [35:36]
A
Leveling it up. Well, I mean, for anybody who has started, I'm laughing because like this is me, I guess. But anybody who has started like really going deep into AI, I can see how it actually maybe isn't as difficult as we thought to vibe code something to create our own platform that could replace other things that we could, you know, build a kind of digital product interface that's specific to the tools that are built on our own IP and frameworks. And that is the next level of thing. But if you want to start doing that, which may not not seem like a big lift. The problem becomes what kind of business are you in? Or do you want to be a software business? Because, hi, that's what happened to me. Because that's what we're doing. I know, it's happened to you too. With note go, Michael. So is that what you're doing or do you want to be the expert? So do you want to teach people how to use AI? Do you want to have AI help them learn what you do? Or do you want to just like become entirely a software company? Because there's a lot of things under the hood that are a lot more complicated than you may think to get started with that.
- [36:49]
B
Well, and this is important, right? Because folks, the kind of things we're coming back to with this is access control, right? And ideally, if we're going to use AI tools and we want them to just be available to our clients or our students or our members or whatever you call it, we have to figure out how to, number one, one, build these things and number two, how to control access. And that is like the next layer because for anybody who ever gets in the software, you start to realize, okay, there's a database involved, right? And there's access management and all these kind of things that you've learned and that I've learned as well. So let's just start with the very basic levels. How could someone, quote, unquote, very easily, vibe code, if you will, some sort of a, a tool that does some of these things. What are some of the tools that basically you might recommend?
- [37:44]
A
Well, some that I've played, I've played around with. Lovable, for sure. Before I actually ended up vibe coding myself a family command center to help manage my life during my daughter's softball seasons and the end of school and how do we eat and where do we go and because the reason I chose that process was because it had to integrate with other tools. Like, I wanted it to have access to my calendar and be able to look at Google Maps to determine how long it would take me to get wherever in the world I had to drive for a softball game that night. So that's one tool that if you basically are just actually just like maybe a high level thing. I think to be good at AI, you have to be good at two things primarily, and that is communication and organization. So if you can communicate and articulate what it is you're trying to do, you could talk to any model, any product, any tool, and get it to essentially do what you want because you have that clarity. And if you have the organization skills, you're able to see like in steps like this. And I never used to be kind of a structured thinker or linear thinker, I don't think. But since I started using generative AI, that has become so important to be able to help guide a tool to essentially like give it a job description, tell it exactly what you want it to do and what you do want to it to come out with on the other end.
- [39:10]
B
So just out of curiosity, on Lovable, I've never used it, but does lovable effectively handle the whole darn thing for you? Does it help create the thing and host the thing and access management, all that kind of stuff in one swoop?
- [39:21]
A
Yeah. And it does like security checks for you, things that you don't think that you need to do, but do need to do. I feel like one of the biggest things that I haven't, I'm like considering whether I should make this a tool that I sell. All but one of the biggest things. If you want multiple people to be able to access your thing, you need to have what's called multi tenancy availability. So you need to be able to ensure that one user is going to log in and all their data stays in their lane versus it's not like getting mixed up with everybody else who's maybe using your tool. And that's why I have a software developer who does these kinds of things and builds our platform thinking about these things. And I just know that it's something that I would have never thought about before.
- [40:07]
B
Are there any other things we need to be thinking about? Because at this point we talked about custom GPTs and we talked about like the pros and cons of that. We talked about Claude skills. Cloud skills are amazing when you have complete dominion over it. But the idea of giving away your cloud skill is scary for people. And then we've talked about the next level, which is to develop your own product. And I know you have a product that does this for people and we're going to get into that. But before we get to the to that, is there anything people need to be thinking about when they're actually designing something that is intended to have multiple users and have access? Just any, any kind of things people need to be thinking about other than what you just mentioned.
- [40:42]
A
I mean, I think the other thing would just be testing. Like that's something we also don't really think about in AI is what you call non deterministic. So it's different you're going to get a different output from a custom GPT from even just, even if you just put the exact same prompt into your quad twice, you're going to get a different output. So what you think about when multiple users are coming is they're going to approach this differently. They're going to have a different input into the process and the output is going to be different no matter what I put in the middle to filter that. So we have to do as much like guard railing, I guess, in that middle piece of the P so that the I and the O can be as close as possible and not like wonky different results. So testing is annoying and important.
- [41:41]
B
One other tool that we didn't mention, but I've had plenty of guests on the show talk about is Claude Code. And Claude code is more than just a code of coding tool, but it is the coding tool. There's also OpenAI Codex. Both those tools are very good and very easy to use. You just kind of talk to it and it will create whatever the heck you want. But the problem with both these tools is it's just going to give you the code you're going to need to take it somewhere. And that's where it gets complicated, right? And that's where, for example, in our case, you know, Claude code is like our tool of choice to develop noco, you know, and it's ridiculously powerful and almost every major company in the world that has coders are using cloud code. It's ridiculous, but at the same time complicated, right, because you're dealing with something that feels super, super technical. But like Kelly said, there are people out there you can partner with, right, who are technical to take this to the next level. But there also are tools that people have built that help with this and I believe you've built such a tool, so maybe you could just share a little bit about that.
- [42:38]
A
Yeah, sure. Thanks for the opportunity, Michael. All of these things were frustrating to me and so I was like, well, I started having that these coffee dates with Andrew, who is my now business partner and he's a software developer. And I was like, I wish that custom GPTs could connect to each other so that you could actually go through a whole process of multiple steps without having to copy and paste and copy and paste and copy and paste all over the place. And I wish that you could share them without being worried that someone's just going to forward that link. And all of your like hard earned energy and effort efforts are just like being shared around like a reel on instagram he said, well, let's fix it. And I was like, oh really? And here we are on this journey and we now built a platform called Wave W A I V and that actually provides the infrastructure for creators to actually implement this. So build your AI tools, your, we call them bot squads because I think that's cute. And multiple steps can happen in the same interface. You can have multiple squads, you send your clients one link, you can see the access that they have, you can deactivate them if they are no longer part of your program. And all of those things are now kind of being solved with this interface that we created. And the other thing that's cool about it is that you can use any LLM. So we thought it was important to have that portability, to have that flexibility when the LLMs change, when new ones launch, every friggin three days at this rate, new ones, like things are coming in there. You, you use the right one for the job and it doesn't have to be like all, you know, an Opus 4.8 that's like a PhD doing what's maybe an intern's job that you could use like a Claude Haiku for, for example. And don't worry I'm saying all these things, but our platform helps you to make these decisions and you don't need to understand all of the details behind the meeting there.
- [44:36]
B
Awesome. Tell everybody where they can go, check it out and also explain to everybody if they want to connect with you on the socials, where they could go as well.
- [44:45]
A
Yeah. So if you want to start playing with building your own bot squad, I built a tool called the AI Tool Launch Playbook. And so our company is called Grabia Studio and we make a platform called Wave. And I made a special link for your listeners to go and grab that for free. So it's graviastudio.com SME and you can find me on Instagram elimakes Wave. And I just look forward to nerding out more about this. Please send me a dm. I just like want to hear all of the ideas that have come up for you and explore those and help you build your bot squad.
- [45:21]
B
It was Kelly makes Wave, not Waves, right? Or was it Waves?
- [45:24]
A
Well, it's Wave. W A I B oh W A I.
- [45:27]
B
Okay, that's important. Kelly makes Waiv, which is the name of your software. Kelly Sinclair, thank you so much for sharing all your wisdom and hopefully inspiring people to go ahead and create their own cool things with AI.
- [45:39]
A
Thanks so much for having me, Michael.
- [45:40]
B
Hey, if you missed anything. We took all the notes for you over@socialmediaexaminer.com a115. Be sure to follow this show on your favorite podcasting app. And if you've been a listener for a while, we would love a review on whatever platform you you're listening on. And do check out our other show, the Social Media Marketing Podcast. This brings us to the end of the AI Explored Podcast. I'm 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.
- [46:09]
A
The AI Explored Podcast is a production of Social Media Examiner.
- [46:16]
B
Before you go, if you want to stop grinding through the ad creative process manually, the AI Business Society can help. Caleb Cruz is teaching members how to automate the entire process on July 16th. Full recordings and all the resources are included with your membership. Join@socialmediaexaminer.com AI again social mediaexaminer.com AI.