SHOW TRANSCRIPT

This transcript is automatically generated by Descript.  Any errors or omissions are unintentional.

Mike Allton: [00:00:00] the cool thing about AI today that is unlike every other technology we’ve ever had in the past is if you don’t know what to do, you can just ask the ai, right? So you don’t have to know how to create a brand voice document.

You don’t know have to know how to create a target persona. You can just open a new chat in whatever LLM you’re using today and say, Hey, LLMI am so and so working for this brand, or me, me personally, and I want to create a brand voice document. Ask me questions, walk me through this with the end goal of having a complete set of instructions.

For myself or for the LM in the future. Mm-hmm. And in a nutshell, what it’s gonna do is it’s gonna ask you for, you know, what is your brand? You know, who are you working with? And most importantly, give us some examples. That’s the best way to train the ai.

Jeff Sieh: Hello folks. Welcome to Social Media News Live. I’m Jeff c and you’re not. And this is the show that keeps you up to date on what’s happening in the world of social media and so much more.

Have you ever wondered how AI could actually make your workday [00:01:00] easier? Not just in theory, but like in real practical ways. Maybe you’re trying to figure out how to use AI tools without getting overwhelmed or stuck in tech jargon. Or maybe you’re curious how successful marketers are customizing AI to fit their own unique business needs, and well, if that sounds familiar or something that you wanna do?

Well, you’re gonna love this episode because we’re sitting down with my friend Mike Alton, a seasoned marketer and the mind behind the AI hat who’s breaking down exactly how to make AI work for you. From building your own AI assistance to crafting better prompts. Mike’s bringing the strategies and examples you can start using right away.

So sit back, clear schedule, clear your mind, and get ready for this week’s episode of Social Media News Live. Mike, it is so good to see you again, my friend.

Mike Allton: Thanks so much for having me. This is always a pleasure. I think I’ve been on, I don’t know, three, four times now. Yeah, always a blast.

Jeff Sieh: if you don’t know Mike, like I do, he is an AI consultant and trainer.

He’s host of the AI Hat podcast, an international keynote speaker and author at the social [00:02:00] media hat where he is tirelessly helping businesses and marketers get ahead of the AI revolution and apply advanced technologies to their roles. He’s also the chief storyteller at Org Pulse, and he has spent over a decade in digital marketing bringing an unparalleled level of experience and excitement to the fore, whether he’s delivering a presentation or leading a workshop.

In fact. My being on Mike’s podcast got me underneath Stelzner’s Eye once again, and that’s why I was able to go to Social Media Market World and present about AI images there. So Mike, I have all you to think because of your amazing podcast, so thank you very much. I’m used to that. Yeah, that’s right. All the famous people are on your podcast.

That’s what it’s so, so let’s just jump right into it because, um, yeah. really quick though, somebody who is sponsoring our show that I need to make sure we mention is our friends over at Ecamm Ecamm dot com slash jeff use code Jeff 15 to save 15% on your purchase of, Ecamm. But, um, AI business growth, with Mike [00:03:00] here.

One of the things that I really wanted to talk about is, you know, some of these questions that, that people have about. Customizing ai. ’cause that was kind of the whole title of this show is how we can customize AI and not just have it be this cool little thing that’ll make a wrap or, you know, some fun images, but actually help our business grow.

So what does customizing AI mean for a beginner? Like, you know, we hear about tailoring AI for your own needs, but a lot of people don’t know where to start. So can you break down the idea of customizing an AI tool like chat GBT, for example, to fit your business or brand needs?

Mike Allton: Yeah, I think it’s helpful to recognize just kind of what you implied.

We are talking about chat GT or Gemini or club. We’re not talking about investing in an expensive custom solution. We’re not talking about investing in thousands of dollars of paid consultancies having to come in and spend months building something just for you. And, and we’ve seen, you [00:04:00] know, large corporations do that, but that’s not what we’re talking about today.

We’re talking about going into CHE PT or going into Gemini and making sure that with each interaction, with that particular tool, it already knows whatever you need it to know in order to do an even better job. Imagine treating Che GBT or Gemini, like an intern where you have an onboarding phase. Mm-hmm.

With an intern, it’s gonna be, you know, measured in weeks with, with ai it’s gonna be measured in like an hour or two, but during that onboarding phase, you’re going to explain, you know. Who you are, who your business is, who your target audience is, what your brand voice is like, what are the common repetitive tasks that just need to be done all the time.

And so with Chat t and Gemini and some of the other models and solutions, you have a couple of different places where you can go to customize ’em. The first level is memory, where you go into the settings and you go into memory, and there’s just a field there where you can put in basic information. I’m Mike Colton, the founder of [00:05:00] the AI hat, also working at Agora Pulse from the chief storyteller.

I help businesses. Here’s my website, here’s my social profiles. That kind of basic information, the kind of stuff that you might want the AI to know at the outset of every conversation. When you are talking to the ai, you can also tell it actively, Hey, remember this? Mm-hmm. And it will also append it to the memory for you.

You don’t always have to do that, which means you might want to periodically go back in and check the memory and see what else it’s added there that it’s trying to remember that. Right. Maybe sometimes you don’t want it to remember those things. Yeah. But the second area, the kinda the deeper level is what I call custom instructions.

Now these are called different things by the different models. Chat. GPD calls ’em custom gpt. Gemini calls them gems. Claude calls them projects. I use Maggi a lot. You use Magi. You’ve talked about on the show a lot. For folks who don’t know, Maggi is a wrapper for all these other tools. And at Magi you can create custom instructions called personas, which I love because with a Magi [00:06:00] persona, you can start one chat with a persona.

And have one of the LLM selected. And then as your conversation evolves, you could switch gears and start talking to a different LLM in the same conversation. Whereas obviously today, if you start a conversation with chat GPT, right? A very high level strategic conversation about a marketing campaign. But then you get to the point where you actually want to create the marketing materials, and you’ve been told that Claude’s a better writer than chat g pt.

Well, that you’ve gotta copy and paste all the pertinent information over to Claude and start over. Mm-hmm. You could skip that with Magi, right? But in all these cases, the custom instructions operate exactly the same. You name it something. And then you put in the text for the instructions. And that could be just specific step by step things that you want it to do when you start that conversation.

It could be additional context. You could attach files. So you might have a custom instruction just for your brand voice, or I’ll address some details in a minute about how some very specific custom instructions could be set up for very specific tasks where you want to do [00:07:00] the exact same thing over and over again.

And that’s the only purpose for that instruction.

Jeff Sieh: That’s very cool. And I wanted to, this is gonna throw you probably a, a curve ball a little bit, but I think it was yesterday or the day before, open, I just came out and said their memory go and they didn’t even, but they said it, it remembers everything, which that’s a big deal.

The problem that I have is a, a lot of us I know are the same way, and I know you probably do the same thing, Mike, is we don’t just use chat GPT for us, we use it for our clients as well. So. What are your thoughts about this long-term memory thing? ’cause the first thing I thought was like, that’s great, but I also don’t want like the, the clients that I have to start using my voice on their stuff.

And I know you can flip it on and off, but do you have any solutions or ideas, for that, that actual problem?

Mike Allton: Yeah. Well, for, for those listing, if you’re not really clear, or maybe you hadn’t heard this before, there, there’s several different layers of memory within these systems. The, the first layer that we’re all familiar with is within the individual chat.

You start a chat with chat GPT, [00:08:00] and you can go back to that same chat a week, a month later. And it’s gonna remember what you talked about previously. To a point. Mm-hmm. Now all these chats have what’s called a context window, which means the more lengthy the chat goes, there’s gonna hit a limit at some point where now it no longer remembers the stuff that you talked about all the way at the beginning of the conversation and stuff will start to fall off.

So that’s a bit of a problem that they’re trying to address. So the next level would be the actual memory setting that I mentioned a moment ago. Now, this next level that’s, it came out with Gemini a couple of months ago. Mm-hmm. They started doing this and now OpenAI has just released their version of the same thing where it will.

In theory, remember everything across all your conversations. I haven’t tested it actively. I’m kind of more of a Gemini user than, than a chat GPT user. I, I suspect it’s not gonna literally remember every single thing, but maybe it has the ability to search them all on the fly. Mm-hmm. I’m not sure, but your point is really valid for people who are using this [00:09:00] professionally, particularly for clients or multiple businesses, or maybe like me, they’ve got multiple hats and they don’t want to intermingle the things that is gonna require a technical solution of some kind.

’cause we’re not going to remember to toggle it on or toggle it off. Perfect. And if, and if the solution really is that they just instantly search all the chats. Toggling it on and toggling it off isn’t gonna really matter. Right. the solution that I’ve seen that works the best is actually built into Maggi already.

They have workspaces. So each individual workspace can be devoted to an individual business or client. Like I work full-time for ago Pulse, so I have an ago Pulse workspace, right when I’m doing my AGO Pulse podcast prep for instance, which we could talk about as a, as an assistant that’s in the ago Pulse workspace.

I’ve got a different workspace for my personal stuff, right. So that those aren’t intermingled at all.

Jeff Sieh: Yeah. I know we have project folders that we use in chat GPTA lot and so I’m hoping, and like I said, this came out I think [00:10:00] yesterday, yeah. That I’ll be able to say, Hey, don’t use this folder. ’cause that’s where all this is.

We’ll see, Dustin actually has, some, a great point and he, you know, somebody who knows, he says memory is actually a slippery slope. Less helpful than most things could talk at length about this from technical aspects to pragmatic aspects. So, but I know that’s, that’s a concern especially for a lot of marketers and.

Even, you know, I don’t want to co-mingle stuff that I’m doing. ’cause I have a very specific voice that I use. I’m asking it business questions. I’m using it as like a business consultant or a copywriter. And some of that stuff is way different than what I want for my, clients. But, we have a lot of people who love Magi.

Chris Stone says Magi personas rule, and then, and, and Dustin takes offense to the term wrapper. Um, oh, you called it the wrap. Now I’ll, I will say this, I said this from, from this, Justin,

Mike Allton: you should tell me what to say instead. ’cause that I’ve been using that term like okay.

A long time. So I even

Jeff Sieh: said from this stage at Social Media Market World, I said, you know, I don’t recommend anybody buy. A lifetime or a, or a year, subscription to any [00:11:00] AI platform except Magi, because it does what you Said that you can go and test out all these different things and he’s always adding new APIs and stuff.

So, by the way, I, I did, I had brought this up when you were talking, but if you wanna find out more about Magi magi.co, uh, slash Jeff, uh, get you there. alright. We’ve got a lot to talk about and I’m, I’ll get distracted ’cause this is so fascinating to me. when we’re talking about, you know, matching a brand’s style, I mean a voice or style.

’cause that’s one of the things a lot of marketers or solopreneurs and everybody is like, they excited about, like, I can, they can write my voice. So, but a, a worry for small businesses. And is that this AI generated content won’t sound like them or reflect their brand personality. And I know you have done a lot of work with this.

So can you talk about some practical ways to address that? Like how you can teach these AI systems to adopt your tone? Is there certain instructions that you use? You mentioned custom GPTs. What are your favorite ways to do that, Mike?

Mike Allton: Yeah. First and foremost, you, you absolutely have to [00:12:00] have what your brand voice is documented, and it needs to be documented in a way that the AI can access that documentation.

So that’s either having a custom GPT or a gem or a persona that you can reference at the start of the conversation because you know you want to create some content and you know, you want it to sound like you. Whether that’s just a personal email or a marketing campaign, you know, whatever the case is, you’re gonna want to make it, make sure it sounds like you or you have it.

In a PD file that you can simply attach to that conversation, which is a great workaround for all of these. You don’t always have to create a set of custom instructions for a task if you can attach the appropriate context to the conversation. So I have my personal voice documented as a PDF. I also have it as, as some custom instructions, but that means if I’m ever in any kind of conversation, if all of a sudden I need to shift gears and I don’t have the ability for whatever reason to mm-hmm.

Call that from a custom instruction, I just attach the file and say, okay, [00:13:00] I want to write this. Use the attached file to make sure it sounds witty, authoritative, casual, yet professional. I mean, all the things that, that kind of make up how I personally sound when I’m writing or I’m speaking. Now, if you don’t have that, then you need to back up before you even get to the ai.

You need to have that documented. Now, I will say, if you’re a Magi user, Dustin’s really helped you out here because he has a persona called. I think it’s just brand voice. Mm-hmm. Dustin can correct me in the chat, so you can just start a new chat inside of Magi and choose from the dropdown of Personas, the brand persona, and it will ask you questions and walk you through.

But if you’re not using Magi, and shame on you if you’re not, but if you’re not using Magi, the cool thing about AI today that is unlike every other technology we’ve ever had in the past is if you don’t know what to do, you can just ask the ai, right? So you don’t have to know how to create a brand voice document.

You don’t know have to know how to create a target persona. You [00:14:00] can just open a new chat in whatever LLM you’re using today and say, Hey, LLMI am so and so working for this brand, or me, me personally, and I want to create a brand voice document. Ask me questions, walk me through this with the end goal of having a complete set of instructions.

For myself or for the LM in the future. Mm-hmm. And in a nutshell, what it’s gonna do is it’s gonna ask you for, you know, what is your brand? You know, who are you working with? And most importantly, give us some examples. That’s the best way to train the ai. I’ve been blessed, I’ve been a writer all my professional career, so I have literally thousands of documents.

I have all kinds of great blog posts that it was easy for me to just pile in there. That might not be you. You might not have a lot of written content, so you might wanna think kind of outside the box. What about emails, social media posts, particularly LinkedIn. think about what you are hoping to accomplish through this.

Is this going to be for professional communication or you just want something that you can be able to write personal correspondence that sounds like you, and let that kind of guide you as to [00:15:00] where you might be able to find some of those resources, but you’re going to want to come up with, depending on length, I would say five to 10 could be less if it’s longer.

Might need more. If they’re shorter pieces, like if everything’s two 50 words or less, you’re gonna probably need 10 to 20 examples. But once you feed the ai, all those examples, it will get a sense for how you prefer to communicate and talk to your target audience. And then it’s gonna summarize it for you so that you can read it and say, oh yeah, I never thought about it, but yeah, that’s exactly how I sound.

’cause that’s what most of you’re gonna find. Mm-hmm. You’re gonna be pleasantly surprised, I think, what the AI infers just by reading your emails or your blog posts. Oh, okay. Jeff always likes to insert some humor into everything he does. Jeff knows this. I’m not telling him anything he doesn’t know, but there’ll be some other aspects about it.

You know, Jeff’s, you know, fairly colloquial in in his talk. He uses short sentences, whatever. It’ll pick up on all those observations. So you review that and they say, [00:16:00] fantastic, thank you. Summarize all this for me and. A single blurb that I can then add to a PDF file and or a custom instruction. Mm-hmm.

That’s how you’re gonna wrap that up. And once you’ve got that done, particularly as a small business owner, this is what’s super valuable. If you’re a professional marketer and it’s your job to write content, you probably don’t really need this, frankly, but if you’re a small business owner, that’s not your job.

You’re doing it because you have to. Right. In addition to juggling all the other things that you have to juggle as the owner of a small business, and so being able to quickly turn to AI and say, look, I, I need a Facebook ad, or I need, an email invitation to an event, whatever, and you just feed it the details and the context, then it’s gonna be able to give you that piece of content that’s gonna sound remarkably like you wrote it from the start.

And that’s where we start to real see real time savings. One, the AI is. Doing a lot of the heavy lifting for you, right? You’re, you’re [00:17:00] doing important things, don’t get me wrong. You’re deciding what needs to be done. That’s where the human still in the loop. But the AI does the heavy lifting and it does it with minimal revisions.

That’s where a lot of people I think, got frustrated in the early days of Right, got GVT or even before Chap HCV Jasper, there were great tools like Jasper, but you know, you’d say, Hey, or conversion ai at the time that I was using it, Hey, write me something in the voice of Gary Vaynerchuk and it wouldn’t sound like me or Gary Vaynerchuk.

Right? So that wasn’t good for anyone. Yeah. Uh, but, but this will come out remarkably well if done well. if you do the brand voice documentation up front, well then you won’t need to keep revising it.

Jeff Sieh: So Mike, I, I haven’t asked you this, and this is something that came up when you were talking about it because I, you, like you said, have been a writer for many years.

In fact, that’s kind of your skillset. That’s kind of what you put your, I mean, your value behind is like, I’m a good writer and you use like, history in your, ’cause you’re a history buff. You’d weave that into your blog, postings. And like I said, you’ve done it for a year. [00:18:00] did you struggle with AI and because it’s so easy now to plug stuff in and get content out.

I mean, sometimes it’s slop, but you know what I mean, like, ’cause a lot of bloggers were like, this was a big deal and they kind of took offense. It’s kinda like when Canva came out with the designers, got all in a a, a huff. Did you have that initial like, I don’t wanna use this. I’m a writer. This is cheating.

I mean, how do you kind of navigate that as a blogger and now you’re this AI kind of specialist, so I’d love to know how you kind of put those things together in your mind.

Mike Allton: My struggle wasn’t that at all. My struggle has been that I recognized a decade ago. Mm-hmm. You know, when, when we were on Google Plus together, that video was the most effective means of communication.

Despite the fact that I don’t really enjoy creating video, I don’t enjoy creating podcasts. I know it’s more effective than written blog posts. And I’ve seen the decline of consumption of written blog posts over the past decade. I used to have a blog. I mean, technically it’s still [00:19:00] there. Mm-hmm. Devoted to blogging and content marketing.

I no longer publish to that site. I don’t push that content at all because people don’t wanna do it anymore. ’cause they frankly don’t really have to unless you really enjoy it. So my issue wasn’t so much the, the morality of using AI to create written content. When I used to be a writer, my issue was simply the fact that I recognized I needed to devote most of my time to creating video and audio content.

And I just don’t have the luxury to write as much as I used to. Mm-hmm. Right. And I have to be okay with that because I have to go and create the kind of content in the kind of mediums and channels that my target audience wants to consume it.

Jeff Sieh: Right?

Mike Allton: So it’s more about them than it’s about me. So the good news for me is, yeah, I can, I can lean on AI to help me create the slew of content that I’m creating between basically two jobs, right?

I have five podcasts. I’m doing customer interviews to create sex success stories. I am speaking on [00:20:00] stages, appearing on podcasts, doing a weekly newsletter. I’m, I’m still a content machine, right? Just most of it’s video. And I’m able to weave AI in a bunch of different places to help me streamline those processes and create the bandwidth to do all that.

Jeff Sieh: Yeah. And I think the, the key thing for that is you developed your storytelling skills through writing. That doesn’t go away. Like you can now take those skills of storytelling and move it to, like you said, where your, your, your people are consuming that content. So that’s what I, even with image creation, like that’s kind of my jam, is like, how to use AI to create images.

It’s the same, how do you tell a story with images that is not changed? Like it’s still the same way that we’ve always had to do that and use those mediums together. And so I I, I love what you said about that, but it was fascinating because I knew that you. I mean, we used to, to joke that you were writing two blog posts with one in each hand on your keyboard, and like, because you were cranking out so much content, and it was really, it’s been really fascinating to see your evolution, embracing [00:21:00] ai, but using it as, as a storyteller and not so just much as a blogger anymore.

So I think that’s really, really cool.

Mike Allton: Yeah. There, there was the running joke because it was during one of our manly Pinterest tip shows. Right, right. I was taking notes while we were interviewing, Gary Goldstein, the producer, pretty Woman, and had a blog post go out like probably 10 minutes after we ended.

Right. I mean, it was disgusting how fast that went. It was, it was like, how did you write that so fast? Yeah. And that was before ai. Yeah. So, but yeah. Yeah. Yeah. That was, that was definitely pre ai. Mm-hmm. But to your point, you’re absolutely right. Like I, I just gave a keynote here in St. Louis at the Business Expo earlier this week, and I’m using my love of history and my storytelling skills to wrap a couple of open loops.

I started off by talking about my podcast situation. Mm-hmm. I went from one show to three shows, just to five shows and how sick that is and, and how overwhelming it was until I learned how to use ai. And then I quickly transitioned to telling the story of Ada Lovelace, one of those historical stories that you mentioned about she was eventually [00:22:00] the inventor of the algorithm in the first computer program two centuries ago.

And I weave that into the conversation about how is AI impacting business today. Mm-hmm.

Jeff Sieh: Yeah. See, and I think those skills don’t go away. They can be enhanced and you can reach more people now because of the things you’re able to do with ai. So, speaking of old school, people that we know from back in the day, peg Tic is saying, Hey, step and Mike Yeah.

Peg is here. So that’s great. And, Dustin is singing your praise and saying, Mike is a great and highly detailed writer, and because of that, his AI output is phenomenal. It will only enhance those skills we already have. yeah. And he, he jokes. Yeah. Mike is blogging right now is our favorite thing to say.

Yeah, that’s very, very true. so I, I want to really quick, ’cause we’re talking about customizing ai. I wanna jump back to what something you said about using those custom gpt and some of the custom things in Gemini, and I wanna make sure that people understand, ’cause that sounds really intimidating.

It sounds like you’re getting in there and messing with the code and all this kind of stuff. Can you kind of talk about the ease of [00:23:00] use when it comes to kind of customizing these AI models? You don’t really need code. You, like you mentioned, you can ask chat GPT to help you, like walk through the steps.

So just kind of reassure people, I guess when it comes to customization.

Mike Allton: Yeah. There is no coding here whatsoever. We’re not talking about building a custom WordPress install where you’re gonna go into the theme and the PHP file and the CSS file, which I can do, but that’s not what we’re doing here.

Mm-hmm. We’re simply manipulating the existing fields. So, let’s take Gemini for instance. ’cause I, I love Gemini. I’m in there every single day. I love being able to talk to Gemini and so on. So in Gemini, the custom instructions are called gems. And when you first open up Gemini, you’re gonna look in the lower left hand corner and you’re gonna see a little field that says Gem Manager.

And you’re gonna click there and you’re gonna see any gems that you’ve already created. Or if not, you won’t see any. Except I think they have some default ones, kinda like Magi does as well. Right? Right. And you will simply click a button that says, create a new [00:24:00] gem. And just like all the other systems, they’re all kind of copying each other, just like they all now have a deep research function.

Yeah, they all have custom instructions, which is a good thing that makes it very ubiquitous. there you just put in a name, whatever you want. So whatever purpose you have in mind for this gem, put it in there. My go-to Gem is what I call my AI chief of staff. It’s an assistant that has been trained on everything it needs to know about me, my business, my challenges, my goals for the year.

So I can fire up a conversation anytime of day with my AI chief of staff. And it already has that context. It’s like an, it’s like a work buddy that knows me really, really well and I can lean over into the next cubicle and say, Hey Joe, what do you think about this? So you give it a name and it could be like a real name too.

Like I’ve, my AI chief of staff is named Clue, but you know the, you could name these functionally whatever you want. And then there’s just simply a field for instructions. And if you worked with the AI in a chat to develop a custom gem or a custom GPT, then it’s gonna give [00:25:00] you the output to just copy and paste.

Into that field. If you’re starting from scratch, then you can put in whatever you want. Now here’s a shortcut. Don’t start from scratch, right? Chet, GPT has a store called the custom GPT store, the GPT store, and that’s where you can find all kinds of gpt that you could adapt to yourself. My very first GPT was based on, my friend Dan Sanchez or Sanchez, he created a, a tool called My Showrunner.

It’s at my showrunner.com. It’s for people who do podcasts, and they wanted to streamline the podcast production. And that was the first time I’d really wrapped my head around what a custom set of instructions were. And he just lets you download it for free. and it’s basically a, you, you fill in the blanks, the details for your podcast and pop it in there, and it’s a tool then to help you build your pre-show documentation and you run a show.

That was the first time I’d ever tried to build a custom GPT, and it was remarkably easy. This really isn’t hard. Folks, you just need to know [00:26:00] what it is that you wanna do. And again, so if you’re not sure what you want to do with it, go browse the GPT store and you’ll see all kinds of examples. Or do one of the things that we talked about today.

You know, build a GPT just for your voice, right? So that if you need something created, you can tap into it there. So that’s the instructions field. And below that, in virtually every one of these, there will be a file attachment field. And this is where you could attach a file, that might be additional context, additional information.

like for instance, maybe your brand already has brand guidelines set up. Maybe it’s not exhaustive, but if you’ve got something, we’ll start there and attach that, and then add what other additional information you need in the instructions, and then save it. And, and that’s it. That’s, that’s all the programming you’ve done.

That’s all the customizations you had to do. ’cause now every single time you start a new chat, you can just select. That GPT or GEM or persona in Magi and you can have a conversation with the LLM, [00:27:00] with the AI system and it will know everything that you put in that instruction.

Jeff Sieh: That’s very cool. And you can go and cus you can change things easily too.

Like if you add, you can add more content, you can, you know, oh, I shouldn’t have said that, that’s throwing things off. I can go adjust that really easily in those gpt. Easy.

Mike Allton: Absolutely. Yeah. That, that one show runner that I first started with has now evolved to five. I’ve got one for each podcast, one for each show that I’m running, because then each showrunner has the unique data details of that show, like the name, right?

The target audience. All my shows for Agora Balls are focused on very specific target personas, like hospitality, social media managers. That’s one of the podcasts. Mm-hmm. That’s how niche it is, right? But they’re all, structured essentially the same. I haven’t worked step by step through a process where I collaborate with it and ask me questions.

I give information, we go back and forth and at the end we get a document that we build together. That’s a pre-show document that I share with each [00:28:00] guest that has the description of the show, the topic, and the questions that I’m going to ask it now over time. ’cause now I’ve been using these things for right, almost a year now.

I have evolved how I want them to act. Like the first, you know, dozen times I might have found myself going back and forth and, and asking it to redo certain sections over and over and over again, and finally realize, oh, I should just go back in and edit. Read the instructions, right? Yeah, yeah. To make it more clear from the start.

Mm-hmm. Rather than having to tell it that over and over again. But that was easy. I mean, once it dawns on me to do stuff like that, it’s, it’s five minutes and it’s done.

Jeff Sieh: Yeah. And I think that’s the thing is people don’t realize that it’s okay to go in and change things. You can, you know, copy it and, and, and move it.

So, Chris, Don, I see your question. We’re gonna get to that in the next segment because that is very, very important. But I wanted to, you mentioned Dan Sanchez. I think it’s the same Dan that was at Social Media Marketing World. I think he works for Social Media Examiner now. He had a great, thing as well. and that GPT you were talking about, there’s a great one that I always use for images that I wanna let people [00:29:00] know about. I think it’s, it’s gatory, it helps you, it gives you actually like four different images if you have trouble like figuring out like, okay, I need to make a graphic for networking instead of that same old.

Image of the overlay of people coming together, what I want something different. And so you go in there and you say that it’ll give you ideas for it. It’ll actually give you midjourney code if you wanna use Midjourney to copy and paste over there. He’s got one on making consistent characters. He’s got one on, how to, for video prompts, text to video prompts for like SOA and some of those other ones.

So, and it’s all free. So go check those out. Those can save you a ton of time. And then you can take them and actually make them your own if you wanna customize them a little bit later. So, great stuff there. Mike, I wanna put, have you put on your, like your, fortune teller hat, and I want you to kind of think in the next couple years, I won’t even say five years, just the next couple years, how do you see the evolution of AI customization coming for these, for small and medium sized businesses?

What are the things that you really would like and where do you, and where do you think like, practically, practically it’s gonna [00:30:00] go?

Mike Allton: Well, where we’re at the cusp of right now in terms of this overarching AI revolution. Like, like, like Jeff mentioned, I, I, I’m a historian, so I’ve studied the industrial revolutions.

We are currently in the fourth industrial revolution. If you’re not familiar with, it’s, it’s not just new tech for marketing. It’s going to be a, an shift across the entire society, which is going to impact jobs, it’s gonna impact how we do things. And where we’re at right now with this AI revolution is what a lot of people refer to as the agentic.

It’s where the AI systems are now progressed, where they have some limited capacity for in independent and autonomous action. Probably you guys have seen some of the demos of like chat, GT’s operator, there’s some other systems out there where you can say, Hey, I want a reservation at a restaurant downtown tonight.

It’s gotta be, you know, outdoor seating, anything but Italian, whatever. Right? And it’ll go in, it’ll try to do that for you. It’ll, you know, and it will surf, a bunch of different [00:31:00] websites to surface the restaurants that fit that criteria. And it’ll start to ping to see which ones have open tables for this evening, which is nice.

In theory, I haven’t seen a lot of examples yet where it’s truly saving me like a ton of time. but that’s because we’re just at the cusp, right? And, and that’s what’s gonna happen over the next couple years. These systems are going to evolve to the point where you can have a natural language conversation, but what you want to accomplish, and it’s going to be far more important than a dinner, a reservation.

It’s going to be, you know, for a personal consumer use, you know, I want an entire vacation mapped out where we’re gonna stay, where we’re gonna go, here’s who’s coming, how long we’re gonna be there, how we’re gonna get there, you know, plane ticket, reservations, you know, all the things. and it’ll be given to you.

Just like a travel agent might in business, it’s going to be things like, I want an entire marketing campaign mapped out for Q4 about around this theme, right? Promoting this product, you know, these kinds of things. And it will be able to go [00:32:00] and do the research, find what the competition is doing, what the market looks like, all kinds of other benchmarks that might need to be brought in that would take somebody hours probably to do and use that to feed the strategy.

I’m doing this today by using Gemini Deep Research for the initial research on whatever it is, and then pulling that data into a new conversation with my AI chief of staff and saying, okay, I did a deep dive. This is what, you know, St. Louis businesses are struggling with as an example of a recent research.

And this is what my competition is like in St. Louis. Again, another example of recent research. How do we help them and how do we differentiate ourselves from the research? And these are. You know, 20, 30 page PDF reports that it’s able to ingest in seconds now that I have it.

Jeff Sieh: Mm-hmm.

Mike Allton: And help me collaborate on, on a, on a strategy as we move forward in the next year or two, the capability of these AI agents will be that they’ll be able to, first of all, do all these things [00:33:00] without having to go back and forth with different chats.

Right. They’ll be able to do them more autonomously and think through, okay, at each step of the way, is this really what Jeff wanted? Is this helping Jeff and his business? Is this aligned with all of our other goals and information? And that’s just the first step. The next step will be when these agents are able to actually communicate with each other.

You will someday, and by someday I mean like within the next couple years, have an AI agent that’s responsible for your paid advertising. Mm-hmm. And a different AI agent that’s responsible for your content perhaps, and then a marketing agent that you’re able to communicate with. To collaborate with all those.

So if you’re kind of feeling like you’re talking about an AI workforce, that’s exactly what’s happening. That’s where we’re moving. And this isn’t just me saying this. If you pay attention to what the folks at Google DeepMind, the people who are behind Gemini or Sam Altman, and the folks at Open ai, mm-hmm.

They have mapped out their various levels of AI intelligence, and this is exactly what they’re moving towards, is AI that’s smart enough to do a [00:34:00] lot of these things autonomously and eventually smart enough to play the role of entire organizations or departments, right. Semi autonomously or fully autonomously.

And it’s incumbent upon us, I think, as marketers and more importantly business leaders to recognize that’s the direction we’re going with this technology. Mm-hmm. You will soon be able to hire an AI agent for five, 10, $20,000 a month. Who’s going to do the role of. What would’ve been multiple people.

Right. And doing They can work all the time. All the time. Yeah. Yeah. 24 7.

Jeff Sieh: Yeah, for sure. So it’s interesting ’cause I was thinking about, you know, the biggest hurdle now, even for like those places where they’ll go and like book tickets for your, you know, what venue or a restaurant, whatever, you’ve gotta hand over your credit card and there’s like a, there’s still this like, ugh, I’m giving my, but we do this already, like with meta Facebook ads, we already give it as card and it’s using the AI to try to figure the best thing [00:35:00] out.

It’s just where that money transfer is gonna happen. You know? Is it gonna be, like, in the beginning Apple knows my credit card information, it’s in the Apple Wallet. Google Wallet knows my

Mike Allton: credit card information. Yeah. So it’s, it’s just going to be a mind shift mm-hmm. In our perception of what these machines can and can’t do.

But eventually, you know, as humans, we’ll all have, I think a personal AI assistant, that quite frankly. Yeah, within five years we’ll probably be an actual physical assistant, a robotic, being of some kind that has access to all this and can, you know, be that assistant that can actually do the stuff that we want it to do.

And then that truly saves us time.

Jeff Sieh: So, Gary has this a great point. He goes, then we’ll have to figure out how to make money if we’re not AI savvy. And here’s the thing that I was thinking about actually this week after Social Media Marketing world, processing everything. My job, like, I do a lot of video editing, a lot of podcast editing.

I think that will be gone in five years. I don’t think I’ll need, you’ll need to hire an editor. Now you may have somebody supervising it and stuff, but already, [00:36:00] like, if I could use AI agents to just throw it at a podcast and get it done, and it’s gonna, it already can pick out the best clips for me. It can do all sorts of some stuff right now, but that it could take and it’s, we’re not that far off for it to do all that stuff for me and I could just review it or whatever.

you know, we all have to start thinking about what our jobs are gonna look like. Of this stuff coming. I mean, it’s a little bit scary, but we can also think of like, we can be so creative now. We have so many cool tools now that we’re able to, like, if I didn’t have to edit anymore, what could I do with that time instead?

Like, I’m excited about it now. I understand this freaks a lot of people out. If they’re building up a, a, a, a business or you know, copywriting or that kind of stuff. I mean, it’s scary as a blogger. Like if you didn’t have this mindset, this could be potentially, you know, like crippling for you. But I think we need to think of it as a way, as an opportunity.

And also we can’t really put it back in the bag. Like, we’re gonna have to deal with it. We can whine about it. We can gripe about it, but we’re gonna have to deal with it. So [00:37:00] very, very interesting. and I knew Chris was gonna come back with this. He says I disagree ’cause I know, I knew he was gonna do that.

I, I do agree. So, so let me, let’s bring this up, Mike and we’ll talk about it. He gets disagree. Jeff, when can was launched, everyone said graphic designers will be eliminated. That is not the case. So I agreed to you to it with a point. Chris, here’s the thing that I like to think of. We used to have blacksmiths on every corner.

Because we all had horses. we still have blacksmiths. I mean, we have, we used to have horses and we had a farrier come out. We don’t have as many blacksmiths. They’re not on every corner anymore. They’re still there. So I think that’s what’s gonna happen with editors, graphic designers, I think to a point has already happened.

what are your thoughts, Mike? Am I off or not?

Mike Allton: I, no, I, and I do think there’s a distinction between the role and the tool. Yeah. Right. You know, we we’re, you and I are thinking about individual people and their roles and their work and their future. Canva is a very specific tool, and I’ll have to, I, I, I have to say from a historical perspective, there is not a tool today that will last forever.[00:38:00]

Right. No. It’s, it’s not possible. so every tool’s days are numbered. Canvas days are numbered. Now are they numbered to this point where we won’t need them in 12 months? No, absolutely not. you know, within five years, will Canva be what it is today? That depends on Canva. Right. Will they shift like Netflix?

I think the answer is resounding yes. Because we’re seeing that today. If we look at the Canva feature launches, they just had their big event. They do one every single year. They’re weaving more and more AI into their tool. Right. And I, I think within five years they will have evolved beyond pure design, creation and editing, because they will see the same thing that you’re seeing, Jeff, is that, you know, the, the basic AI tools are replacing a lot of their basic functionality.

And we’ve, we’ve seen this with technological advancement time and time again, going back to the human part of it. Mm-hmm. When ATMs were introduced, many people predicted the end of bank tellers. We did see fewer bank tellers per branch, but what we also saw was more branches and more [00:39:00] tellers overall.

Right. Because their roles evolved from basic counting cash to more advanced financial transactions. So over the, over the all thing, we’re, we’re not talking about job elimination. It’s job evolution into something newer, bigger, and better.

Jeff Sieh: Yeah. I knew this was gonna bring up good stuff in the comment.

So, Gary says, over on YouTube, he says, chat fatigue gave wrong or misleading answers to 35% of financial questions in a new test, according to entrepreneur, don’t rely at completely. And then he goes, legal, financial and medical are all dangerous fields to rely on it. at this present time, I would say yes.

Like, and there’s a great, I can’t remember the article. It’s like, why, AI can’t do math very well because of the generative thing. There’s this whole science behind it. But it was the same reason recently that, you know. AI couldn’t generate a full wine, gra glass to the brim. Like it wasn’t, it was impossible because you look at all the images out there, it was trained on, it’s always a half glass of wine.

Now, with the new open ai, you can do it full to the brim or to the bottom because it’s using different kind of [00:40:00] math and algorithms to do that. So I think eventually, like I accountants, like QuickBooks is already starting to roll out AI stuff. they’ve been doing it for a while. I would love to be able to do that for, for accounting, but I, I think we all are gonna have to be careful, but I think things are gonna evolve and we’re gonna have to, to roll with the punches and, and see.

But I think it’s an opportunity instead of a. A thing to get, uh, stressed out about. But anyway, this isn’t a production show. Yeah. but go ahead.

Mike Allton: Well, that’s right. Let, lemme weigh in real quick though. Yeah. On what Gary said, because I wanna share a word of caution for every single person watching and listening.

Don’t let statistics like what Gary shared and not, this isn’t a dis on Gary at all. ’cause I, I’ve seen other kinds of, of performance benchmarks and metrics in the past, but do not let that lull you into some kind of false sense of security. Oh, AI can never replace me because it can’t do X, y, or Z today.

You’re going to be really disappointed in 12 months. In 16 months when all of a sudden AI can actually do that 10 [00:41:00] times better than it can today because that’s the rate of progress that we’re seeing. The models over the past 12 months have all become reasoning or thinking models, which means they take the time to think through things.

So now they can actually start to do complex mathematics and they can start to do calculations and they’ll start to be able to do financials. Right. And honestly, 12 months from now, I would expect that 35% to be, you know, almost no errors whatsoever. They’re gonna be correct 99.99% of the time. Don’t let that lull you into false in security.

Jeff Sieh: Yeah, and I do, this is, I love my audience. I love you guys who show up every week. It’s so much fun. And Gary says, if I’m pouring it for the white wife, we’re talking about the wine glass, it’s to the brim. Uh, preach brother. and then, um, and this is a good point by Chris, he goes, I think doing video and podcast editing is a job you’ll lose if you don’t understand how to use all this stuff to set yourself apart from what everyone else is doing.

That is true, I think across industries like, and even Mike, he, you know, he’s using his skills like we talked [00:42:00] about, his storytelling skills that he has crafted throughout, all his writing career. And it’s not throwing them away. They’re, no, they’re still super applicable to whatever he’s doing because he knows how to craft a story and we things in it and teach AI how to do that a little bit quicker.

So I think that, that, I think we’re not disagreeing on everything. Chris, I think things are gonna change rapidly, faster than we can comprehend. But I also think the way to tell a story, whether it’s podcasting, video editing, whatever, is still gonna be a skill that people will pay to do because, the creativity is hard to scale in a lot of times.

And so I think your point is valid. So anyway, back to an earlier question Chris had that I really wanted to bring up because, it goes right into the, next. part of our show. He goes, Mike, are you using AI agents? And what are some practical ways for creators to lean into a gentle ai?

and we had this conversation before the show that we need to classify the difference between [00:43:00] AI assistance and agents. So can you kinda walk us through that? Because I think there’s a, it’s, it is very confusing when people name it different, like operator is, like you mentioned for chat gt. Yeah. So can you talk a bit about the difference between AI assistance and AI agents?

Mike Allton: And, and the first thing I’ll say is confusion is normal in AI today because there are a lot of terms that are used interchangeably and there’s no, you know, world Book of Standards. This is what this means. Nobody agrees on what artificial general intelligence or a GI truly means or looks like. There’s just a bunch of kind of similar definitions floating out there.

AI agents are very similar, particularly honestly, once the marketers get their hands into these things, right, they’ll call anything an AI agent if they think it will sell them more subscriptions to their tool. So it it, it behooves us to be a little bit careful there. So when I talk about AI assistance, I’m talking about custom instructions.

I’m talking about versions of the large language models that have been [00:44:00] set up by us to perform specific tasks, whether it’s just that one thing that you have to do every day, analyze reports. And give you a summary that you can email the team that’s an AI assistant, or you know, to much more broader, looser things like my AI chief of staff, who’s obviously an AI assistant for me, but not built specific for any individual task.

Those are AI assistants to me now, AI agents is something very different. That’s what we were touching on a moment ago, where they are AI systems that are deliberately programmed by the manufacturer, whether it’s the LLM or a SaaS solution like Salesforce, where the AI system can operate at least somewhat independently and autonomously and completing multi-step tasks.

So I think that’s a crucial point, that multi-step task. This isn’t just sending something, via Zapier to chat g pt, and then having it translated. That to me wouldn’t be an AI agent. It’s cool. that might [00:45:00] be really useful. I use these systems to translate almost every day at a gorum impulse because we’re a global company.

Mm-hmm. And I’m doing interviews a hundred percent in English, but we want those same interviews to go out in French and German and Spanish. So we’re using ai, which is really good at translating today. But an AI agent is something that’s going to be able to think through a process, perform multiple steps, and then come back to you with the results or when it stops.

So to answer Chris’s question, deep research is my favorite AI agent application today, and here’s why. It’s an AI agent. I go into deep research and I’m using, again, Gemini, consistently for this because with Gemini, I’m already paying 20 bucks a month to Google. I was already paying for the extra storage.

And then I realized last year I just. Needed to switch the subscription to Google one. Yeah. And now I have complete paid access to all of their professional AI tools for the same cost. and I like it over deep research from chat GPT because there [00:46:00] you’re limited to 10 searches a month, if I’m not mistaken.

I’ve seen some people report, you know, better results, which is cool, but if it’s only 10 and I wanna use this every day, that’s a problem. Right. So with Gemini Deep research, I type in, in very natural language. I wanna know more about this topic, I wanna research this. And whether it’s, a personal research topic, or, or something for a marketing campaign or something on behalf of a client, like maybe I need to research, you know, market trends in St.

Louis and come up with three customer retention strategies for small retail shops. That’s really specific, right? And I might spend, you know, hours pouring over blog posts and studies and case studies and so on to, to surface something, which all I want at the end is to be able to give my client good advice,

Jeff Sieh: right?

You know,

Mike Allton: I’m not necessarily paid by the hour. So me being able to substitute a tool to do all that research for me is extremely handy. So I tell it what I want, [00:47:00] and the tool comes back and says, okay, I understand, Mike that this is what you wanna research. Here’s my research plan. And you get to look at it and say, yeah, that looks pretty good.

What I really love is that these things are so smart. They often insert ideas into the research plan that you didn’t suggest. I was on a, I was actually on a call with a client and, he’s in the event sector and the call was mostly about ai. But one of the things that he, he talked about in the call was that he wanted to kind of raise his awareness in the event industry.

And I said, well, you need to start showing up on podcasts. It’s a really smart approach, to raising your awareness. And I said, you could use deep research to research all the podcasts in the event space and who the podcast owner and, and contact might be. And so while we’re on the call, I put that in there.

I wanna know all the events, all the podcasts in the event space. You know, what, what’s the name of the podcast? And, and who’s the host of the podcast and like a website if, if they have it. [00:48:00] And in the research plan, Gemini said, you know what else would be really cool if you knew what their requirements for a guest were.

Or if they even accept guests. Mm-hmm. Like, oh, that’s smart. Yeah. Because if you’re doing podcast outreach, first of all, don’t embarrass yourself by reaching out to a podcast that never does guests ’cause, right. That doesn’t look, it looks stupid, obviously. Never listen to it. Right. Right. And secondly, you know, if they’ve got published requirements for the guests, well, why not pull that into the report?

And so then again, it’s age agentic, meaning once you say yes. To this research, it will independently go out and look at 10, 20, a hundred different websites. Mm-hmm. Scan them all, pull that information all together and then publish it as an entire report right there on your screen. Yeah. And you can download it as a PDF.

You can now have a conversation back and forth with it. You can say, oh, okay, that’s cool. what about this podcast? How often have they published? Or whatever questions you might want. [00:49:00] Or maybe it surfaces another angle that you didn’t think of before and you wanted to do some additional research. You can have it do that.

with Gemini you can either download as a PDF or what I usually do is push it to a Google Drive folder. Mm-hmm. and now it’s a Google Doc that I can share with the client. That’s awesome. Or you know, if it’s for personal use, reference that in other chats.

Jeff Sieh: That’s really cool. And I, so because of the thing I did last week at Social Media Market World using the new ’cause, it just had like broken three days before that this new uh, model came out.

I had to upgrade to the $200 a month Jet GPT plan because I needed to get slides done. but one of the cool things is it gives me access to the deep research, their new model and I can and do some really cool stuff inside of there. And I wanted to use it ’cause I had, this month I had to buy anyway, but I used it actually for this show with Mike and I actually showed him the results.

’cause I thought it was really, really cool. ’cause I know he’s e nerds out about this stuff too. But what it did is it walked me through actually when it was trying to get information from his website, it would actually show me like, oh, this is blocked. And I wanted to send it to Mike because I know that he’s big into [00:50:00] SEO and you know, matters what things are being, crawled and Googled.

But it would walk me through that in that kind of, those steps that we talked about for agents. And it was really fascinating to watch it. And it took like 20 minutes now. It asked me beforehand clarifying questions of what I wanted to do with this research, which was very, very helpful. Another thing that I’ve done that I usually do for my shows, and this is a a, a really practical example.

I’ve had guests like that have been on, Chris Stone and, Jim Fuses show, like they’ve been on the show. I love their show. It’s great stuff. We have some crossover audience, but I wanna ask my guests questions that haven’t been asked before. And what I’m able to do is use these agents to go in and research.

I’ll say, go look at all the YouTubes that they’ve been on, you know, ’cause I can read the transcripts and I say. This is what I wanna talk about my show. I don’t wanna ask a question that’s already been asked a bunch of times before. I want it be just for my show. And it does that. And I’ve had numerous guests say, you know, no one’s ever asked me that before.

Which as an interviewer, that’s what you want. Right? So [00:51:00] being able to do that, which wouldn’t, it’d be impossible for me to do the amount of time to do that kind of research I can do now to make my show better. And I think that’s really, really cool as creators that we have these tools that we can become better interviewers.

So I just wanted to share that, ’cause I thought that was a lot of fun. So

Mike Allton: I think it really illustrates too, just how rapidly these systems are evolving because a year ago, almost a year ago to the day I interviewed Chris Penn for my podcast. Yeah. He was one of the first episodes that that dropped out.

And this is how. Intimidating and nerdy it is to talk to Chris would be, he’s in the conversation. We were talking about podcast interviewing and this is obviously before any kind of deep research agent was out, right? This was a year ago. April, 2024. And he made me let him share his screen so he could put me to SHA because he’s like, yeah, you don’t ever want to ask somebody who’s been on lots of shows, you know, the same question.

So what I do is go to YouTube and download the transcripts from all their shows [00:52:00] and write this custom python script that pulls ’em all in the text, trans transcripts, and then I feed that into chat g pt and I have haven’t analyzed it. Make sure I don’t ask these questions again. So there were a lot of programming steps.

Yeah. That even though he demoed it like this on the podcast interview, I’m still like flabbergasted. It didn’t really comprehend me how you would actually do this. ’cause it’s just second nature to him. Don’t need to do that today. No. Just ask deep research to do it.

Jeff Sieh: And I remember trying to actually get a, a full hour transcript inside of Chat Toi to analyze, and I had to do all these workarounds where you would split.

Remember, we have to split up the things and then combine it back in and hey, don’t forget this part. And now, I mean, that wasn’t that long ago. So anyway, to our points that this stuff moves fast, it moves fast, and it’s almost mind boggling to think what it’s gonna be like in six months, like what we’ll be able to do.

So, very, very cool. yeah, so, Justin says, just drop a, a link to Magi, [00:53:00] YouTube, link into Maggi and you’re done. So he’s always, yeah. So if you guys don’t know and haven’t tried out Maggi, go to maggi.co. Do slash Jeff. he’s amazing. And another shout out to our folks at Ecamm ’cause we are at the end of the show, because we’ve nerded out so much about this.

We gotta do a part two with Mike, because we didn’t even get to the last section. But Ecamm dot com slash Jeff, we appreciate them for sponsoring the show. And Mike. I wanna let you have the, floor because you have so many great resources on your website and everything. So tell people where they can find out more about you.

What you got going on the floor is yours, my friend.

Mike Allton: Well, thank you. So if you wanna connect with me, just go to the ai hat.com/connect. You’ll find all my social channels there, all my podcasts and that sort of thing. So whatever it is that you’re hoping to gain, whether you wanna click with me on a particular social channel, or reach out because you might want me to help your business level up your use of ai, that’s the way to do it.

Now, I mentioned that AI chief of staff as essentially a work buddy for me. I actually documented how to create your own AI work, buddy. I [00:54:00] initially thought it was gonna be a prompt, and then the prompt got so long I thought, no, I’m gonna put all the details in a PDF file. And now the prompt is just one paragraph.

So you download the PDF file for free@theaihat.com slash ai work, buddy, all one word. And you’re going to take that paragraph prompt that I give you and attach. The PDF file. And it’s basically gonna have your LLM of choice ask you questions to learn about you. And once it gets through asking all the questions, it’s going to give you the entire set of instructions that you can then copy and paste into your new custom instruction.

And boom, you’ll have your own personalized AI work buddy that knows your role, your responsibilities, your challenges, your goals, even your communication style and preferences. And you’ll be able to turn to them anytime you want to for free.

Jeff Sieh: See, there’s a reason I’ve been friends with Mike Alton for such a long time, is because he is super smart.

He, is on the cutting edge of all things creative when it comes to social media marketing, AI stuff now, which flowed perfectly into Mike’s skillset. So make [00:55:00] sure you guys check him out because Mike, has been around for a while and there’s a reason, there’s a reason why folks like Mike have been around and continue.

To add value to their audience. And, uh, with that, thank you guys so much for day. Sorry, we, nerded out a little bit. Didn’t get to the, the final section, but I’m sure we’ll have Mike back on. Thank you for Peg Rich Graham, Dustin Stout, Chris Stone, everybody. Gary Stockton, of course, everybody who, stopped by, in the chat and gave us great questions, great thoughts.

We appreciate you wouldn’t be able to do the show without, you wouldn’t be able to do the show without our friends at Ecamm. So make sure you guys go to Ecamm dot com slash Jeff. And with that, we’ll see you guys next week. Bye everybody.