Today, I'm pleased to welcome David Gonzalez, a trailblazer in digital marketing and the innovative mind behind the Internet Marketing Party. For over 15 years, David has been a pivotal figure in the digital marketing industry, hosting impactful events and mastering email marketing strategies through his agency, Lisbur. Everyone knows Dave for his ability to forge influential endorsement deals and for his unique mastermind experiences. So join us to gain insights from David's extensive experience in blending business success with life's joys. Let's get started. Welcome to Using AI at Work. I'm your host, Chris Daigle. Each week we'll be learning how today's business owners, entrepreneurs, and ambitious professionals are getting more done with smart use of tomorrow's tech. Let's get started. Hey buddy, welcome to the show, Using AI at Work. Hey, thanks for having me.
unknownYeah.
SPEAKER_00So, Dave,
one of the David, one of the things that I am most curious about is we're going to talk about a number of things, but the one I'm most curious about is the fact that you've got optics on a lot of people's businesses. You're not necessarily a consultant, but people they seek your counsel a lot when it comes to things or people that would support their current challenge in their business. So it's a unique position. Not everybody gets to see behind the curtains in a lot of businesses. So I'm gonna, if you're cool with it, I'm gonna ask you without you don't have to name any names or anything, but I'd love to hear maybe not only how you're using uh generative AI in your business, but some of the use cases you're seeing with the numerous people that reach out to get your advice. Sure. Yeah, I think uh a good way to start would be walk us through a typical day of in the life of David Gonzalez businessman and where AI is being used and what kind of impact it's having for you. Okay.
SPEAKER_01Yeah, for me, my an average day is I run I run um uh an umbrella business that has different uh businesses underneath it. And one of them is Internet Marketing Party, one of them is the three-day think tanks where I bring in entrepreneurs to uh share with me their biggest business challenge, which is I think is what you were reflecting, where I'll curate experts around those challenges to serve as their board of experts, their therapist, and almost like an ayahuasca session because we don't allow any suggestions to be made for the first hour. It's only clarification questions. And inside of doing these every month for the last 13 years, we started doing them two years into Internet Marketing Party. That's what you're referencing, I think, is seeing behind the curtains on some of the biggest names in our industry. And so um, yeah, part of what my business is is um putting together uh speakers and events for internet marketing party for the think tanks, and now for the three-day think tanks, which I do in a big mansion. Then I've also uh do some uh list monetization in the survival tactical and um health and I mean wealth space, health, weight loss space, and I've got a couple of endorsement deals. So there's a lot of what I do in my business that has to do with leveraging my relationships in a win-win way. And so my days tend to look very different, but if I had to put my finger, like if I had to say like the 80-20, tends to be a lot of um business development calls, a lot of clarification and alignment calls, and a lot of uh my my team, my my assistant manages my email inbox, so clarifying emails that she doesn't know quite know how to respond to, andor text messages. There's a lot of texting, a lot of clarification of emails, and a lot of phone calls, and that are all around deal making, um, aligning priorities and aligning vision for whatever it is some of my clients or colleagues in business are working on, either uh or or any projects that I've got.
SPEAKER_00Does that is that is that too vague or is that yeah, and just to help everybody who's listening kind of get an understanding of where you're coming from, what your paradigm is like, and that sort of thing. So it sounds like you're spending a lot of time in ideation or clarification with others rather than you know, a solo producer or solo creator kind of vibe. So it at any point, that's an exciting uh opportunity to explore AI's application. Are you using generative AI in any of that output that you do?
SPEAKER_01Yes. Um I I actually
use it if I had to categorize it, I use it for um writing copy. Um and I don't mean copy as in creative copy, I mean as in um tweaking. Sometimes I just I have a lot on my mind and I'll just speak it out either to my assistant or to I'll talk, I'll call a friend like you and say, hey, this this isn't quite sitting well with me. Do you mind if I record me talking out with you what I need to say to somebody that I need to cancel or change a deal or um or or create a deal that doesn't yet exist that um and and so I'll I'll take that recording and then give it to neither in auto recording or whatnot, give it to GPT with an instruction that's or a claude that says, Hey, I I've been struggling or grappling with this, and and it's you know, it's I I the best way to say that is I think um Blaise Pascal had a quote, I would have written a shorter note, but I ran out of time. And so a lot of times I feel like you know, there's that saying like, I don't know how to define obscenity, but I know it when I see it. Or I don't know how to make a beautiful designed website, but I know it when I see it, kind of thing. I feel like a lot of times uh GPT for me is is like taking an idea, a concept, or uh something that's not quite well formed. That I there's people in my network that I respect that I want to give them a shorter note, but I you know, and I want to respect their time, I want to respect their attention. And so I'll use interesting, I'll use the LLMs to take this nebulous idea or this not well-formed concept and turn it into something that is very tight and concise. Um that's one of the ways that I use it as is for copy in that sense. I also use it for analyzing uh data and spreadsheets. Um, I've also started using um personalized uh LLMs or GPTs where um I'll take in any one of my businesses a lot of high-level information. Um like like one of the things that I'll also do that I don't know if everyone does um is I whenever I work on a project using uh one of the GPTs, uh our or Claude, I'll what I'll do is I'll I'll I'll revisit. Like usually if it takes like 20 minutes to an hour and a half, I'll say, hey, analyze my language here and distill out what are my rules so that the next time I work on a project like this with you, I'm not starting from zero. So I have my tone. Like let's say I was I was I I struggled a lot at how to how to get this, let's say I wanted to invite 20 people to a specific thing. The first thing I would do is I would say, I would use it for copy. So I might call you, Chris. I'd be like, yo, Dr. D, let's can you hop on a call, a call for 15 minutes and I'm gonna record it. And we'd jump on and I'd say, hey, so I'm gonna be doing this thing in March, and let me tell you what I've got cooking, ask me some questions, and then at the end of it, you'd be like, Oh yeah. I, you know, and at the end of it, I'd distill that and I'd say, write me a text or an email inviting people to this, and then it would send me its version, and then I would say, here's my version, and my version would be like, Hey bud, how's it going? Uh hope you're having a great, hope you're having a great Christmas. Hey, wanted to run something by you. You got a few minutes, and then I would I would generate like a little a whole little like campaign, if you will. And then I would say, here's your version, here's my version, extract and deduce what I did so that there will be rule, there's a rule set. So the next time I want to do something, there's all I it's already it's starting from a from a v2 or a v3 rather than a v1 or a v0. And so that's a secondary way that I that I use it, is I I will like it. I literally have a folder on my Google Drive that's called AI Info. And it's it um, like I said, I have an umbrella company, and so I've got my some like very high-level explanation of what list revival is. I've got a high-level explanation of what internet marketing party is, the history of it, the team members, the monthly events, how what we charge, who are members, what's what you know, different jargon that we use, like speaker think tank and things like that. And now I have one for three-day think tank. So it's like a document. And so what I did is I created a personal LLM whereby um I fed it all of these PDFs into a system. So now whenever I want to ask it a question like, hey, can you explore with me a better possible pricing model for I am party monthly? If I don't have to explain to it what's I am party, what it what does it mean by monthly, what's our I don't have to tell it what our current pricing model is. Yeah, yeah. It's starting from a, you know, it's almost like having somebody on your team that's worked on your team for a year or two or three.
SPEAKER_00That's got context.
SPEAKER_01Yeah.
SPEAKER_00Hey, so uh what is the tool that you're using?
SPEAKER_01Is that did you build a custom GPT or are you using I actually just have GPT four and in GPT four there's um if you go um if you go to chat, you know, uh chat GPT four or the openai.chatGPT or whatever it is, on the left hand side there's a I mean at least they did for a while. I'm custom instructions and it's pretty straightforward, you know. It's just yeah, just hit it and create your own. Um I have I have a friend that that was you know CT from Opus, the soundbed, he was playing with it and and asked him if he could show me how, and he's like, dude, it's so easy. Like it'd be like me asking you, you show you saying like show me how to drive a truck instead of a car.
SPEAKER_00Yeah, it's or just to write an email or something.
SPEAKER_01Yeah, yeah, yeah.
SPEAKER_00That's a great tip for everybody. So custom instructions, if you're not currently leveraging that, if you go to Google and type in how do I build custom instructions for my Chat GPT account, I I I've done the same and it this is pretty smart though. So just to recap, what you what you're doing is you're taking the transcription or audio file of your actual language and training the model essentially on how to use your tone as closely as it can represent.
SPEAKER_01There's a small distinction there. I'll take a recording to give me a first version. Because, you know, let's even if you were like we've been on this call now, what, 11 minutes, right? If you were to take the recording of this 11-minute uh podcast and say, you know, transcribe this and chapterize it and give any hot tips and action items, it would probably take this 11 minutes and turn it into something that was one paragraph that someone could read in 15 to 30 seconds. So then if I said, okay, take that punchy paragraph and turn it into podcast notes, and then it would say, Oh, okay, David Gonzalez, founder of Internet Marketing Party, interviewed by uh Chris Dagle, founder of um chief AI officer, and then it would say, you know, Chris join Chris and David as they talk about, you know, how David's you leveraging this business for his his monthly events, I mean LLMs, and then it would, that's it, right? But then I would say, well, that's not how I sound. Here's how I would write what you said, GPT for I would say, hey guys, or I would say, you know, strap, strap your seat belts because you've got, you know, whatever it is. Yeah, yeah. And then it would go like, oh, I see, you write in a much more casual, fun, uh like organic style. And then I would say, cool, set give me those rules and one or two examples, and then that I put into my AI folder on Google that I then call David, I call that um, I call it like your prompt library, essentially. AI. Nice, yeah. It it's called David Tone or something like that. David, yeah. And so So that's relevant. Go ahead. Uh yeah, and so it literally says you're you start out, you know, you're very uh casual in your writing, you refer to people as hey homie, hey bro, hey bud, hey, hey, you know, how's it going? Like you start with it with an opening question, you're cordial, you're inviting, you're engaging. It was if you think about the way that I write versus the way that um Mike Norman writes, right? Like we're we've we've very great friends, but he writes real different. And he's like, How are we doing? Right? Actually, you and Mike Norman talk more alike than I talk to you or Mike Norman. You know what I mean? Okay, you know what I'm saying? And there's it, if I was to look at if GPT gave you uh a prom, you know, a uh an a text or an email to send out to 15 or 20 of your entrepreneur friends, there's a way that you would edit it that would sound like Daigle. And there's a way that Mike would edit it that would sound like Mike, and there's a way that I would edit it that would sound like me. And if I did three to five different examples of that, I would then show it to GPT and I'd say, pull out, what did you do that I didn't? And what did I do that you what did I not do that you did? And then it would it gives me those rules, and I say, Okay, cool, turn that into a rule set. And the next time I'm gonna write an email, I would say, here's what I want to say, here's my rule set, go.
SPEAKER_00Interesting. So the the one of the challenges that I have when using the models to create content or emails or whatever, it's the having it define the tone. Because for for new users, if you're listening to this, a lot of times if you ask it to write something, it it's the output is what's called synthetic data. It's and it reads like synthetic data. It doesn't have the warmth or the nuance or the whatever. So by doing something like what David's talking about, you can get the the beginnings of a rule set, as Dave defined it, with that teaches the model how to generate output that's more like that's gonna require less work on your part because when you get the output, you don't copy and paste that anywhere. That still needs human augmentation, it still needs a human in the loop to give it that that natural feel. But the closer that you can train it on your tone by doing exactly what David just talked about, the less massaging of the output from GPT will be required because it'll be more closely aligned or held even better than how you would have said it yourself. Right.
SPEAKER_01Yeah.
SPEAKER_00So that's awesome.
SPEAKER_01And some of the examples are like I just opened it up. It's like communication style and preferences. It says direct and sincere approach. Begin with an immediate acknowledgement or context setting, address the matter at hand without hesitation. For example, for instance, mentioning, hey, I've I have people waiting for confirmation conveys directness and context. So instead of just saying, like, hey man, are you gonna be able to make it to the next three-day think tank? I would say, hey, I've got a couple people that are you know, that are waiting, and I've only got two spots left. Like, are you in? You know, like that way it's yeah, yeah, and not like, hey, so how's it going? You know, blend a form blend a formal and casual tone, use a mix of professional language interspersed with casual terms like epic, homie, take care, mono, or examples of casual tone, strong personal touch, start with a direct address. I often use terms of endearment, familiarity, narrative flow, understanding of empathy. So that's those are some examples, yeah.
SPEAKER_00So that's essentially like a prompt in your prompt library that you use to very quickly uh start getting output from the model that's more in line with what you're looking for. I think that's awesome. Yeah, yeah, I hadn't considered doing that. I'll tell you, man, every time I do these podcasts, I'm like, oh, I know a lot about I use it all the time. I always learn something really cool from every single guest that we have. It's awesome. And this is a way of tone has been an issue, right? Like I've been, how do I how do I not have to, how do I get a better output from the model? And what you just described is how I'm gonna do it from now on.
SPEAKER_01Yeah, and it's actually kind of fun because it makes me feel like I'm doing something. I don't know about you, but sometimes, you know, when let's say I was to come visit you in New Mexico and I stayed stayed
at your place for three or four days, and I was like, hey, how do you do this? Like, isn't I don't know about you, but I find it gratifying to show someone who's intelligent and who I care about how I do something that I do naturally well. So when I when I interact with with one of these LLMs, I enjoy training it and going like, hey, here's what you did. Here's what here's how I'd say it different. Tell me what you did that I didn't and what I did that and what I didn't do that you did, you know, I the the quadrant. And then it splits that out, and I go, cool, now turn that into rule sets and give me some examples so that then and then I'll take all of that and turn it into something. So I think it would be an enjoyable process for you to literally tell it to do something, like maybe like that's a great idea. Your next how to reach out to 10 people in your network to invite them onto the podcast. Like say, hey, there I run this podcast, I want to invite 10 people in my network. I want it to sound kind of cool, uh, I want it to be simple and easy for them to say yes. Give me your best um text invite. And then if it says, if it makes like a four-paragraph text invite and be like, okay, here's the way I would do it. It would say, Hey buddy, run running a podcast. Uh I'd love to have you on. You down? Question mark. That's literally what a Kiki would say, right? And then it would go, oh, cool. So your your text messages are rarely going to be more than two sentences, and they're going to be very to the point. And then and then you would have it give you three or four examples, and then now you've got it.
SPEAKER_00So this process that you did, did was it modeled on anything, or it was just a trial and error as you played with GPT? Uh no, because that's a that's a prompt pattern that you just defined. That's like you know, good.
SPEAKER_01I I feel really um like kind of, you know, when somebody gives you a compliment that you weren't expecting to receive because it was just like, oh, that's just how I did it, you know. Um I think I interact told you about Joshua Pellis here, um, that was working with Evan. He stayed with us for a couple of days, and and he's really, really into, you know, um AI. And he I showed him what I was doing, and he was like, dude, you're like using, you're doing naturally what we so I feel like I think um I've always been a big fan of frameworks and understanding the structure and the sequence and the order of excellence and the way people do things well, and much like I referenced earlier, how I I tend to be a good fan of design, but I and I I tend to be a fan of really well-organized and decluttered spaces, but I don't naturally have the knack to design things well or stay decluttered and organized. So I whenever I have a tool like the LLMs that that naturally work that way, I feel like it it's basically like I don't know about you, but there's been times when it's five or six o'clock and my assistant went home and nobody Nobody on my team is working anymore, but I've still got the gumption and the desire to work. Yeah. And I've already called all my homies that I brainstorm with. So I I almost feel like I'm I'm alone. Like it's like, oh, I've like I've got to do some project that requires two or three people, but it's only me. And I've had that that unmet desire for so many years that I feel like a lot of times um the LLMs end up being like a dream come true for me. Like, oh my God, I can continue working with someone who's got an IQ of 155, where Elon's got like a 150 or 160, depending on who you talk to. Like, like I've got a work buddy that I can talk to and that's smarter than me and that I can bounce ideas off of. And so that's how I think about it. I think about it as like, ooh, here's someone I can finally engage with that isn't gonna ask me for a raise and over time and doesn't get sick.
SPEAKER_00Yeah, yeah.
SPEAKER_01Yeah.
SPEAKER_00So how how did you like what what was your introduction to using the LLMs? Um who turned you on?
SPEAKER_01Um you know, it seems like one day it was just they were just here. Um you know, I wish I could say I was like some of my friends that have been using them for years, but I didn't really catch on until like I would say it was within the first few weeks of when everybody was talking about, like, you know, but my my uh friend Josh Whitten, uh from Make Soil, and you know, he's really he's been involved in AI for a lot of years. Um he stays with us when he's in town, and and he uh I heard him talk about anthropic and uh open AI, and and he was where I learned about Claude versus Sydney versus GPT
versus Palm versus Llama, and I was like, oh shit, like there's a lot of and he turned me on to the idea that there's different ones, like some have the amount of amount of tokens that you can feed them, the amount of information they can give you back, and then this is before it's now it's become a lot more normal, but it was like no, some have the ability to access the web real time, some their web access stopped as of 2020 or 2020.
SPEAKER_00So you use a lot of the models, you're not just like a um like a Marlborough man, you're not just a GPT guy.
SPEAKER_01I was using Claude more than uh I call it T3 or T4 because it's just fewer syllables. But ever since GPT 4 came out, with the ability to create my own LLMs or uh GPTs, and since it's it can surf the web and it's um but I often will compare results. Like I'll take a prompt that is well done in GPT-4, and if it's a meaningful enough project for me, I will take it to Claude and see what Claude gives me different because sometimes there will be one of the you know, to answer your original question about how I use it, I already said copy, I already said, you know, uh work buddy, I already said, you know, um that it give one of the other ways is it'll give me it helps me organize and sequence and structure information because um there's things that I've been doing for so long in my business that I only know how to do them second hand, like second nature. Um, but that doesn't necessarily mean that I would be qualified to teach it in at Stanford or Harvard, because there's a very ideal way, like in accounting, it would be generally accepted accounting principles, right? So there's a formalized, or if you're in chemistry, there's a way to do proofs and there's a way to do uh abstracts that if you don't follow that system or that model, you're you're you can still get a good result, but you're not following generally accepted practices. And so a lot of times I like using uh the LLMs to to take what I do organically and formalize it to allow me to see gaps that sometimes show me like entire new worlds of like, dude, I never thought about doing that that way. That's easier, that's faster, that's better, or that's more thorough. Um yeah.
SPEAKER_00I'm interested in this. Tell walk me through what that would look like because I I'm like you, and in any professional that's been out there, they have that unconscious competence with what they do.
SPEAKER_01It it's actually explained. It's gonna sound a little repetitive, but what I'll do is I'll say, um, so let's say I was I was gonna host uh the three-day think tank, right? And I've I've done that, I've done that before. I would say, I would, I would talk it out with my assistant or with someone. And a lot of times I'll leverage phone calls, and then I would I would pop it out and I'd say, hey, I'm hosting this event. Here's the structure. I'm I'm charging 10K, I'll have 10 people, it'll be in a mansion, but and uh and and then I'll say someone has asked me for a proposal to be an event coordinator. Write me a proposal. And now instead of because if like let's say you were gonna host uh chief AI officer conference, right? Um there's some obvious things you would know to do. You would know, hey, you need to get a venue. Hey, you should probably have somebody that's gonna run it, like as an event coordinator, you should probably get some sponsors, you should probably make a list of your attendees, you should probably have a general outline of itinerary. Beyond, oh, you should probably have food considerations. Uh, and then you'd probably want to do some some cool factor, right? Like everyone that comes gets something, right? Whether it's an experience, we're gonna go exotic car racing or blah, blah, blah, right? But that's that's the general category, right? But then if you were like feed that to AI and say, hey, act as a world-class event coordinator. Um, here's what I have. Turn this into a um uh a high-level itinerary uh overview for anyone on the team to know and be able to use. And it might say things like, oh, where's your performa? Where's your budget? Um, and are is this all in one spreadsheet? Are you using Trello? Or he's like, Oh shit, I never thought about using a project manager. And oh, and by the way, here's a pre-done template of how to do an event. And I was like, holy shit, like I didn't think about um, you know, how am I gonna do invitations? Are they gonna be mailed? Are they gonna be emailed? And and oh, and by the way, here's a timeline by this is your latest time. Oh, and hot hotel room blocks. Didn't think about you know what I mean? And so it's almost like I imagine if you had someone, imagine if you had Michael Jordan on your team and you said, Hey, I want to win an NBA championship, and you're like, here's what I think needs to go into it. And he might say, What else? Yeah, you might say, like, well, where's your draft pick in this? Oh shit, and like it's gonna point out things that you wouldn't ever thought of, and then it'll and then and then it and some of the things you'll be like, yeah, I don't need that.
SPEAKER_00You know, this is interesting because uh strategy is uh leveraging the models for strategic purposes is of great interest to me. The the content and all the
like the tactical stuff is it's very helpful. However, the value that you get from making those calls to friends and saying, hey, let me, you know, let me just run something by you and give me your feedback kind of thing. Not everybody has that, right? Like they they got a couple buddies at work or something like that, or they go home and talk to their husband or wife and get get opinions. But to be able to use the models that the way that you're doing for strategic application is not something that's talked about very much. Most people are looking for let me give it input, let me give it output, okay. Next task, not necessarily like okay, GPT, let's have a sesh here, let's get down on and like really get into it. So um, and that idea of the event coordinator is fantastic. I mean, I've put on events a lot of work, a lot of work, huge reward, but a lot of work. If I would have had access and done exactly what you just did, time, money, mistakes, all that stuff would have been mitigated significantly. So that's a great tip. Yeah. Now I want to kind of shift gears to, man, you got, I mean, I've met some actually Joe Stoltey was a podcast guest already. Uh you you've got some uh high achievers that are high ethics, high integrity people in your your network. Are you hearing their anecdotal use cases of AI and their businesses? Obviously, Joe's got AI driven, it's a whole business, but what about some other folks?
SPEAKER_01Um anything come to mind? I'm hearing people use it for, and I've started to use it for data analysis, um spreadsheets and finding stats and data that I wouldn't have seen on my own. Patterns and stuff, yeah.
SPEAKER_00Um we've done that. We we've got a a project in a different industry, and we were able to take a lot of the, I don't know, maybe I don't remember how many records, but maybe about 10,000 um user profiles from a CRM for that industry, right? And we fed that into, and it it had a lot of data points for it, but we fed that into the model and we said, identify you know, as many different avatars, customer avatars from this data set. And the space is the cannabis space. And you think, oh, okay, it's people that want to buy cannabis. Unbelievable the output that it gave me. It identified obviously the like the holistic health individuals and all of the categories they would be interested in in buying, right? So this was built around we've got access to data, the individuals want to monetize the data through sending offers, you know, list revival kind of vibe. Um and it identified, I think we had six or seven different avatars from that spreadsheet, essentially. And now that opened up, I think we ended up with about 50 different potential advertisers. And the average individual would get access to a list and and any industry and go and just look for the similarities, like, oh, these people like fishing, let's sell them fishing rods, right? But what it opened up was these people also drive Cadillacs, these people also subscribe to, you know, Time magazine. It it opened up all of this, these opportunities uh just from that data set that again, we all have a bias, right? We all look at things through a certain worldview, and we have defaults and models like you're talking about. This allowed for me to um evaluate it without those biases, without those uh default frameworks that I would normally, you know, like the the thing to a carpenter, every problem could be solved with a hammer kind of vibe, right? We all have those patterns, and this broke those patterns. And I'll tell you another supplemental to that, and I mentioned it in a recent podcast. I was going through Alex Hormozy's hundred million dollar offers. Fantastic book. If you haven't gotten it and you own your own business, worth a read. And he he introduces this concept of the Grand Slam offer, right? It's so much value in his process of going through it. And one of the things that he did was he talks about how I I use divergent thinking, right? And in the GPT model, that would be okay, I've given you this data
set, give me 10 ideas that you know are you can just say use divergent thought to come up with 10 ideas, and what it's gonna do is it's breaking out of your biases or your frameworks or defaults, and it's pulling in ideas for products, pricing models, you know, whatever that you would you would have never thought of, even with your own team, because on an internal team, there's a culture. That culture is a bias or it's a way that we do things. It's like having that outside consultant come in and say, Oh, we look at it this way, you guys look at it that way. Here's my ideas. So that's pretty interesting there. Um is AI, I know that when you do these think tanks, how how big of a topic is using AI in business or uh concerns with AI or anything?
SPEAKER_01I feel like um the original heyday has of of like it being something everyone was talking about has normalized. Um and I think there are people that are using it, but because our community tends to be people that are looking to to share the cool the cutting edge cutting edge, yeah, that I think it's not being talked about as much anymore, but I don't think it's because the people aren't using it, but they don't want to sound like they're like, oh, have you have you seen the late to the game? Like, hey, have you seen the side Tesla Cybertruck? And it's like, bro, that that was two weeks ago. Like, and it's like, no, no, no, but have you actually driven in one? It's like, no, I haven't. Like, dude, it does, you know, but like people don't they're not they're not making that initial, like I think people have already kind of in their mind assumed that oh, ai's good for writing copy, it's good for uh reading your email, you know, just different different uh sentiment analysis, yeah, yeah, yeah. And um so I I I don't hear many people talking about it as much.
SPEAKER_00Um interesting. So for context for the listeners, David lives in Austin, uh a tech and entrepreneurial hotbed. So it that would make a lot of sense. I would imagine that somebody that was listening to this maybe in a place that didn't have such a an identity associated with, you know, uh entrepreneurship and tech and cutting edge, they might not be experiencing that, but that's a great point. And it doesn't surprise me because uh especially in the marketing space, you do want to be identified as a thought leader. And if you're talking about something, if you're evangelizing on something that came out a year ago, it doesn't necessarily the narrative might not be that you're on the cutting edge anymore. Yeah. Yeah, I see it as 2023 was the year of awareness, right? Like people, what is uh you I mean, even even though GPT 3.5 was only released about a year ago, you were very early to not just mess around with it, but to apply it and kind of crack the code. The experience you had when your buddy came through when Joshua came through and and you know commented on the the logic of your prompting, like that that doesn't happen overnight. So it's obvious to me that you've been using it probably longer than I have. Um I get great results, but I didn't and and also I kind of defaulted to I used Claude because of the context window, but now that GPT can do that, I just it kind of seems to be my default. I still pay for premium versions on all the models, and I don't but I've kind of gotten my default now. Yeah. Does your team use it to you know in their workflows?
SPEAKER_01Yeah, my my my team uses it for sometimes I'll I'll I'll like just yesterday. Um I I found well last week I found out my my assistant who ended up being my event coordinator for the last three-day think tank won't be available for the next one. So it kind of put me in like, oh shit, I've got to find another event coordinator. And I reached out to somebody who was a very, very high likely potential candidate for it. And we had a phone call. By the end of the phone call, she's like, Yeah, can you send me a proposal? And I was like, sure, sure, I'll send you a proposal. And it's like, I don't know about you, but sending proposals tends to be uh has a lot of emotional charge for me because like I it's on something I've never done or I've never created a proposal because you don't want it to be too formal, you don't want it to look too amateur, amateur, and then you also like don't know what the best price is, like you know, and and and like how how much detail to add. And and so I asked my assistant, hey, would you I always feel guilty like asking my assistant to do shit that I hate doing because it's almost like, hey, will you clean out the latrine? And she was like, Yeah, sure, I'll do it. And the way she's like, I was like, Yes, thank God for Cole. And like 30 minutes later, she's like, It's in your inbox. And I looked at it and I was like, Cole, this
is amazing. Like, I felt humbled, I felt small, I felt like, and I was like, How did you do like did have you did you model a proposal? She goes, Oh, I just took because I recorded my call with with Olivia and she asked me for the recording. I actually recorded my call with her, and then I recorded, I I also talked to somebody else about being an event coordinator. And yes, and I recorded both of those calls and I gave them to her, and she took those those recordings, you know, the transcriptions of them, fed it to a to GPT, and said, Hey, my boss is uh looking to hire an event coordinator. They've both asked for proposals. Um, here's the here's the description of the event, here's the phone calls he had. Um, can you turn this into a proposal? And it, I mean, like I I wish I could, you know, do a quick screen share um to show it.
SPEAKER_00So if she didn't have that tool or wouldn't have been trained to think to use the tool, that would have probably been an all-day or a multi-day effort. She got research, you gotta get some models.
SPEAKER_01It would have, it would have been such a and I would have been a hundred percent okay with her doing it because I it would have been like, uh, I don't want to do it, you do it. Uh yeah, I'll text it to you if you want to take a quick look. Yeah. Like if you want to take it and just see what it came back like, because if you look at it, you'll probably go, holy moly, this is this is good. Like, I might want to keep this as a as a template for a future event.
SPEAKER_00Did you train them on your style of using the models?
SPEAKER_01I have heard I've done it very with my assistant much more intentionally and hands-on, but it's also been what's the word I'm looking for? Uh it's been through osmosis. Organic. Organic and through osmosis. And then my wife uh serves as head of operations on our team. Um, she's smarter than I am, but I've given her points of like how I use it, but I I didn't train her. I just me, I just encouraged her strongly to use it. I've got a virtual assistant. And now you'd say she's she's a convert?
SPEAKER_00Oh, yeah, she uses it all the time. Um so let me ask you, I know that you you do a lot, but you have a small team. Has your team by them using GPT and other models, has that allowed you guys to actually open up bandwidth, do stuff that had been on the back burner or had been forecasted like you know what's a blessing and a what I kind of consider a call curse of of these LLMs?
SPEAKER_01I kind of feel like it's kind of like when people are like, damn it, this microwave's taking two minutes to make me a baked potato. Like what used to take an hour to make a baked potato, you can do in two minutes, and now if if it takes you can't make a baked potato in 30 seconds, you feel like it's taking too long. So as much as I want to say that like GPT and uh these line LLMs have opened up more bandwidth, I think we just normalize to this is the new normal. Um and and it's yeah, I'm not proud of it. It's but it feels like okay, that like now we can now we can move faster. We don't look at it as yeah, like finally we have a car, is what it feels like. Yeah, like if you know if you lived in Western society and and like you got a car and you didn't have one, you'd be like, well, finally. It doesn't feel like, oh my god, now that we have a car, we're so much more blessed.
SPEAKER_00Taken for granted.
SPEAKER_01Yeah, yeah. That's yeah, yeah, unfortunate, but that's I that's how I feel, and I know our team we don't look at it as like it's not this promise of now I only have to work three days a week kind of thing.
SPEAKER_00Entrepreneurs just fill it up with something else.
SPEAKER_01Yeah, and and I uh you know, in this day and age with the there's so much shifting with the economy being Better and worse for different people and the the
future not looking as certain. I I don't see it as being ungrateful. I see it as like, hey, I better I better take advantage of these advantage of these these unfair advantages or new advantages because I don't really know what the future looks like. And there's some parts of my business that are still doing well, but I would be foolish to assume that they will continue to do well without AI impacting them. And sometimes it's even indirectly, like we run an email business and the ISPs have AI at a level that we don't understand, and they can and and and like from what I'm hearing, like uh affiliate offers aren't converting as well because the AI models at the ISPs are starting to notice, like, oh, this is clearly an affiliate offer, and then they're gonna put it in the promotions tab, whereas before, like, and and or this is the same copy that everybody's using, or hey, it was written by AI, or hey, you know, and it's like there's subtle nuances that are making that are having a trickle-down effect of certain things that have always just run a certain way aren't being as interesting, profitable or effective. And so yeah, it's kind of a more of a philosophical conversation, but yeah, I I I I feel like like leveraging these tools is is safety and and it's it's important. It's not like okay, now I can coast.
SPEAKER_00Safety more like it helps ensure the viability of your enterprise.
SPEAKER_01Uh-huh. Yeah. Not maybe not safety, but it's uh it's insurance. It's more like uh what's the word when you enriching. Enriching for increasing the probability of when thing when the dust sets up. I don't know when how long it's gonna be before dust settles. I I think we're at the very beginning of a lot of things looking really different for maybe a few years.
SPEAKER_00Well, I hope that that uh works in both of our favors and all of our listeners' favors. Hey, one final thing, Dave, David. Um there's people that will listen to this that will this this kind of like, oh my gosh, I didn't know it could do that, right? They haven't made that that aha or that oh moment where they've had the experience, they've interacted, they've gotten output, and it's made their mind go, right? For somebody that isn't uh familiar with the use of them or or maybe has only casually used them, what advice would you give them for getting for gaining that comfort level and that like use comfort?
SPEAKER_01Uh I'm gonna borrow Evan Pagan's advice, and it's not direct advice, but I'm gonna borrow what I learned from him, and that is there's um technology, when we think of technology, we think of something that's like a machine that physically can do something we can't do as well or as easily or at all. Um and it's you machines usually have moving parts, right? So there's there's biology, which is prep previous to technology, which is just plants and animals, and um uh uh so there's biology, which is the study of living things, and then there's technology, which is the study of things of machines that can that can we can leverage. Okay. And then there's um he considers computer science as being separate from technology because computers technically don't have moving parts. And it's easy to think that one builds upon another, but artificial intelligence is a separate category than biology, it's a separate category than technology, and it's a separate category than computer science. It is what he's referring to as, and I don't think it's his something, but I think it's something he he's borrowed or learned, and that's intelligible, which is the study of intelligence that is separate from biology and it's separate from technology, it's pure intelligence because it has an average of 155 IQ, as I mentioned earlier. So if you think of, if you make the mistake of thinking of artificial intelligence as being technology or being computer science or being an extension of us, then you're not you're doing yourself a disservice by thinking like when in the 30s or the the you know early 1800s or late 1800s when George was Henry Ford said if I had asked people what they wanted, they would have said a faster horse. And it was like it was something completely different he built, which was a car. And so this is complete, like a car isn't a faster horse, it's a completely different thing. Um, and so artificial intelligence and these LLMs and these systems, they stand alone, and to be thinking of them as like, oh hey, um like I'm gonna take a picture of my refrigerator and have it give me recipes. Like, that's such an underutilized use. Um like think of this as if you had an extraordinary think of who your favorite business book was written by. Mine was by uh Keith Cunningham. Another one would be MJ DeMarco, right? Or Chris Chris Chris Christensen, right? You think of these legends, geniuses of business, titans of business. You can literally feed that book as a PDF. You can go and find that book as a PDF, feed it to your LLM, to your, and then say, give me the consulting that Keith Cunningham would give me. Here's my business problem, here's my business case. Or if you're like looking at starting a new business, you can go and talk to somebody that has that kind of business and say, hey, would you like to partner with me? Or is a way we can collaborate or tell me the pros and the cons of starting this business, record that, and then probably they're gonna say, they're gonna say so much that by the end of it you'll be like, I just feel like I was drinking from a fire hose and I don't even know how to categorize what they said. You can take that recording, give it to GPT, and then have it tell you like distill it down to something you can read, right? And then go like, and then take that and then give it to one of your best friends or one of your business colleagues, which is one of the things I did. I was thinking about going into another email business. I distilled it down, then I took it to a friend of mine who's a genius, a business genius, and he literally said to me, David, with the amount of startup capital you'll need, you would be emptying out your savings, and you've never run this kind of business. And it doesn't sound it sounds like it's a like a 50-50 chance. So I would recommend you not do it. Me knowing to not do that business saved me priceless weeks, months, maybe even over a year of deliberation, consideration, and then years, and then potentially the future of my business, because like I don't have an infinite amount of savings, and I was about to pour it into this thing because I thought so, like hell yeah. Does that make you know I yeah, what what did you hear me say, like in your own words, because I that'll help me.
SPEAKER_00I I heard you say, and it follows um some research done by Pew. They they researched a bunch of adults in the US who had heard of Chat GPT, and of those who had heard of Chat GPT, 90% had used it for entertainment purposes only. And what you're suggesting here is that for anybody that's new to these models, sure, it can it can build you a seven-day menu based on a picture of you. It can do those things, but that is not that's like um that's not even scratching the surface of how how much impact it can have on your life. Because what you talked about just now, that would have taken you on a completely different trajectory in life. And the recovery from that, should it have been unfruitful is impossible. The recovery would have been all of the time, and all of that time means when you're saying yes to that thing, you would have had to say no to other things. It's just bandwidth limitations. So that using it instead of picking out a keto diet based on the picture of your cupboard, you did that to make a major life decision, which may not have seemed like a major life decision to you at the moment, but upon reflection and output from working with the models, it became apparent that oh my gosh, that could have been a real shit show. Yep. Right?
SPEAKER_01Yeah.
SPEAKER_00So that's that's awesome.
SPEAKER_01Yeah, and then I don't know about you, but I like setting quarterly goals. But sometimes it's very, very, very challenging to decide what to make my one quarterly goal when there's so many so many things pulling attention. What becomes clear for me is I can say, hey, I'll say, hey, can you what questions like here's literally, I said, taking my quarter, my Q4 goals into consideration, give me some ideas about how to kick off tomorrow. No, no, that that was I already I said, I I need to find it, but I basically there's one where I said, based on what you know about my business, because remember, I have already fed it all my stuff. What questions do you need to ask me to help me decide where my quarter, my Q4 goals should my what should be my top Q4 goal? And then it asked me, it was like, oh, well, I think it would be smart if you were to tell me what your profit and loss was. Uh so here's what I said. I said, I I said, you're a business, you're a world-class business advisor with several multimillion dollar startups founded under your belt. You're advising the found, you're advising me, founder of Simply the Coolest, to determine what my Q4 goal should be. In spite of that, of the fact that it's already November uh November the 7th, because I was I got behind. I said, you're set on giving me a clear North Star that I can orient to and shoot for the rest of the year. Use the following notes to ask me what questions you would need to come up with the best ideas. And I gave it all my business, you know, what's going on, blah, blah, blah. And then what it came back with is said, oh, here. I would I I would need to know what are your um what are your profit margins? Where did your money, where did the majority of those profits come from? Um, is there anything going on in the marketplace that would make this, you know, is there anything you're working on that you hate doing? Is there anything you're working on that you're excited about? Um, are there any, you know, it almost went through kind of like a SWOT analysis.
SPEAKER_00Yeah.
SPEAKER_01And by the end of that, I was like, oh, it it made me think about some things. It made me realize, like, shit, I don't know.
SPEAKER_00It's making me do the same right now.
SPEAKER_01I don't know what my PN, like, I don't know what my margins are. I think I do, but I don't know. And you know how that thing works, right? Like, how many times if like if I asked you, if I told you you'd have to give me a hundred thousand dollars cold hard cash, if you were off by five percent on what your margins were for 2023, would you know what they are?
SPEAKER_00This has been a year of investment rather than earning for me with the startup. So um I haven't I haven't been doing any sales, and most of my revenue is from residuals, kind of like you.
SPEAKER_01So, but let's even say in your investing, but I but under these circumstances, even even with your investing, not so much, even with your investing. So let let's say you've like made three to five investments. Could you do you know, could you prioritize them which has had the biggest ROI? Like from one to one.
SPEAKER_00I have a good finance team. So if you let me make a call, if you let me phone a friend, I can tell you tell me the answer.
SPEAKER_01I'm asking whether you know it. And that's the kind of thing to where, like, you know, if you it's important if you fed that information to GPT-4 and you said, hey, here's here's what I did this year, here's what the results were. Give me some insights that I might not see on my own. And it's gonna come back and bam, bam, bam, bam, and this, and congratulations. And man, you really need to not do the third thing because it looks like you did well, but these other things you should have doubled down on instead of do this, you know, because that return is not good. You know, it looks good because you've the volume you put into it, but on a percentage, you're less than three percent.
SPEAKER_00You know, you know, this what you're doing is something that should be common practice in business. Any business owner or consultant, coach, solopreneur, whatever, should be doing exact that exact act uh activity. And I'll tell you, it reminds
me of I saw a quote and it said, within a few years, a doctor that makes a diagnosis without having consulted AI could be sued for malpractice, right? It's a projection, but they're indicating that's how that's how powerful the counsel of the models is to an ultimate decision. If a business owner is not doing something similar to what David just talked about, they are relying on their own intellect to help them determine the next best path forward rather than consulting with you name it, Lee Ayacoka, um Franklin Delano Roosevelt. Any any any thought leader that you can think of can be accessed in these models. In Dave's case, he David's case, he used a generic, you're a successful business consultant who's had many uh many multimillion dollar startup successes. But if you have a specific uh thought leader that you respect, plug them in and say, how what would they tell me to do that in this situation? I actually did. I told the models now.
SPEAKER_01I told it to to tell me what after he gave you the generic one, I said, okay, now here's the situation. What would Keith Cunningham say?
SPEAKER_00What would I've used Keith before?
SPEAKER_01Yeah, what would uh and then sometimes it just gives me like really high-level data, and I'll be like, no, no, no, no. Like, dive in. I like I'll even get kind of smack it on the wrist a little bit sometimes, and I'll be like, come on, like, like in sometimes it'll give me like 18 things, and I'll be like, it's too much. I can't I can't digest that. Distill it down to the top three that I can that are the most important that I can take the most action on. And by the way, talk to me like like a like like don't give me just principles and like apply it into my business, and it'll go, okay, sorry, and then fix it. Because a lot of times it'll just be it'll just it's almost like you're reading like a textbook, and you're like, come on, don't you like you're phoning it in. Like, don't just like weave in, weave in these principles into my business and into what I'm what I focused on and into what today is. Like today is Thursday, December 28th.
SPEAKER_00David, this has been a a fascinating uh insight into how you're using it. And you know, we've got the chief AI officer certification program, and we teach a very we teach some unstructured, but we teach very formal prompt patterns and like things that you can use to get predictable output. But what you're doing, you're not using training wheels, right? You're you're up in the air with this stuff.
SPEAKER_01And what I mean by that is that I mean I to me it just feels like like I've been I've I've gotten like some of my natural ways of being because I didn't have this before. I I I I I I want to say part of it is because I tend to be a verbal processor, I don't like writing, and I love to talk things out with people, and I there are not enough people in my world that I can talk things out through without feeling like I'm overstaying my welcome. And because of my background in personal development and stuff, like I have a sensitivity where it's like, okay, this person's just being courteous at this point, but I still have another two hours to talk things out through, right? So I'll like, yeah, sometimes I'll even just go into my garage and just get hit a recorder and just start talking about things. And yeah, and I didn't feel like I could do that before because then now I needed to give it to a human. Now I could just talk things out and give it to GPT, and it'll distill like what I took 30 minutes into like five minutes or two minutes.
SPEAKER_00I haven't tried that, but that that would be like a divergent activity for me because I'm not much of a talker, but I'm actually going to give that a shot. It's a great bit of advice. Hey, so we've done an hour, David. I really appreciate how much time and how much information that you've shared on this.
SPEAKER_01Oh, that's fun.
SPEAKER_00And you you do some cool businesses, and I'd love for you to tell everybody about how to find out more about you real quick.
SPEAKER_01Yeah. Um, just shoot me an email at David. It's simply the coolest, uh, just like it sounds. Um, I my I I the the internet marketing party is something I do in Austin every month, and and I I get I get to I get to polish up the website because it's not something that I feel really proud of right now. And then three-day think tank is um where you know something that I it's I just did the first one in December and I'll be doing the next one in March. And I've literally one of the things on my agenda is to get the website up and running because I just sold all of them with phone calls and text messages. And then list revival is all done through referrals, and so we don't even have a website for that, and that that's a multi-million dollar business that doesn't even have a website. Like I lean too much on my relationships. I I'm the internet marketer that doesn't do internet marketing.
SPEAKER_00Yeah, it's working. So for anybody who has the opportunity to uh attend one of the internet marketing parties, it's something that David's been doing consistently for over a decade. Um, some of the top names in uh digital marketing, SEO, paid advertising, and these aren't just like, hey, I'm doing a good job. These are the the top names on any topic related to uh an online business, monetizing an online business. David's had them as a guest. So if you make it to Austin and it lines up, you may just see me there. I'm quasi-regular, even though I'm not in town. So David, thanks so much, buddy.
SPEAKER_01Thank you for having me, Chris. Appreciate you.
SPEAKER_00Thanks for tuning in to using AI at work. Don't forget to subscribe for more conversations about how to use AI at work. And a special thank you to our sponsor, Chief AI Officer, for empowering businesses with AI education and training. Visit their website for free AI readiness assessment and AI strategy guide to help you get started using AI at work. That's www.chiefaiofficer.com. Thanks to our producer, Evan December, for making this episode please. Follow us on Twitter at the handleusingAI at work. And visit www.usingai at work.com for free resources to help you harness AI in your role.