Loading summary
Transcript73 lines
- [00:01]
Michael Stelzner
Are you feeling overwhelmed trying to keep up with all the AI changes? Trust me, I know how it feels. The marketers who are truly thriving today with AI have discovered something really important. They're not doing it alone. At the AI Business Society, we've created a community where smart, curious marketers like you share discoveries, troubleshoot their challenges, and celebrate breakthroughs together. Listen to what Marissa Shadwick had to say. Quote, I found my people. I love chatting about our experiences with AI and supporting each other on our journey towards the future. Unquote. Stop navigating the AI revolution by yourself. Join our community of innovators and let's grow together. Visit socialmediaexaminer.com AI to learn more. Welcome to the AI Explored podcast, helping you put AI to work. And now, here's your host, Michael Stelzner. Hello, hello, hello. Thank you so much for joining me for the AI Explored podcast brought to you by Social Media Examiner. I'm your host, Michael Stelzman. This is the podcast for marketers, creators and business owners who want to know how to put AI to work. Before we move on to today's show, I just want to share with you how exciting times are right now for us at Social Media Examiner. More than six months ago, in October of last year, we launched our AI Business Society. And we have been recruiting some of our best guests, is what I like to say, into the AI Business Society. And they come in and they do training every month. We've got live office hours where we answer people's questions multiple times a month. And then, of course, we also have a really active community in there. If you love this show but you wish you could be more interactive and actually interact with me, I am always there whenever the trainings are there, bringing clarity to all the guests. Similar to what I do here, you may want to check out our AI Business Society. It's our exclusive membership, it's economical, and it's really designed to make your life a lot easier. So on today's podcast, this is going to be a fun show because who does not want more time? You're going to hear me say that in the interview. Who doesn't want more time back? What if you could use Chat GPT to assist you to regain time back so that all of a sudden now you have an extra couple of hours a week or maybe even an extra couple of hours a day. What we're going to talk about with today's guest is exactly how to do that so that by the time you're done listening to this, you will Be able to use that tool that you use all the time to help you get your time back. And not just get your time back, but do things with excellence. Today I'm going to be joined by Kerry Weston. And I really think you're going to enjoy today's show. If you are a regular listener to this podcast, be sure to follow this show on whatever app you're listening to us in so you do not miss any of our future episodes. Let's now transition over to this week's interview with Kerry Weston, helping you simplify your AI journey. Here is this week's expert guide. Today, I'm very excited to be joined by Kerry Weston. If you don't know who Kerry is, he is the host of the ChatGPT Experiment podcast, a show designed to help entrepreneurs and marketers understand how to benefit from ChatGPT. He offers one on one coaching, consulting and workshops focused on helping business leaders regain their time, money and resources. And Carrie, welcome to the show. How you doing today, Michael?
- [03:50]
Kerry Weston
Thanks for having me. Absolute pleasure. Doing great. This is an honor for me. It's been a while. I've been listening to you forever. So to have me on is awesome. So thank you very much.
- [03:57]
Michael Stelzner
Well, I'm very excited to have you on. Today, Carrie and I are going to explore how to regain more of your time with ChatGPT. And who doesn't want to regain more of their time? Now, before we go there, I want to hear your story. How in the world did you get into ChatGPT? Start wherever you want to start with your journey.
- [04:14]
Kerry Weston
Yeah. So there's probably three key pieces, Michael, that I'll share because it's a long journey, because I'm old man. I think the first thing I just want to say, it's all about curiosity. That's been kind of my life is just being curious and getting into trouble that way. First thing I remember being in college, first paper I passed in, I got a big old F on it. I didn't keep the paper. I wish I had, but I had a red F circle around it. I can see it in my, my head now. And I didn't get an F because of what I did as far as content. I got an F because of how I did it. I hand wrote the paper and she said, you can't handwrite in college. And I said, I don't know how to type because I'm 52 and my path through school was there was never any computers. Right. Like computer was one thing in the library. And at 18 or 19 years old. And this is where I'm going to probably divide your audience. Half of your audience is going to nod along with me right now and the other half's going to look at me and say, you're ancient, man, you're a dinosaur. But at 18 or 19 years old, it was cheaper and probably more common to own a car than, than it was a computer back then, right? I mean, you just didn't have one. And so it was literally sink or swim, man. She said, well, you better learn. And my curiosity started right there. Like there was no class, right? There was no structure. I just went and jumped in. So that was the first thing. Second is, I always had a goal of having a business before I was 30. And that's about as far as my business planning went. I just went on my own business card on my own desk. That's it. Like I had no reason for anything else. And I remember walking into an ad agency in my town. I said, this might be interesting. He wasn't hiring. But I talked my way into a job because I wanted to know, if I go this path, what do they know that I don't know? And I talked myself into a job. It was a great experience. And what I learned was they were just trying to figure it out too. Like all these people that we think made it and have all the answers, they don't. They're just using their experiences to make the next best decision, decision. And so that gave me courage, I'd say. On a Monday morning, I woke up one day, I resigned from the agency. And I woke up and I decided I was going to have a business. And so I was self employed. And for about three minutes. That was a great feeling. About minute four, I realized I was really unemployed, man. Like, I didn't have a client. A business card doesn't get you any revenue. I had to go figure this out. And so here we go. Curiosity again. And so I owned an agency for 25 years. I had a business partner for 20. And there's a major moment, there's collisions in your life that pay dividends. And I always thought I'm either gonna succeed or I'm gonna learn something. Either way, the experience is gonna be valuable to me. And I talked my way into a coaches show at the University of Maine up here. So they were looking for a vendor and I, I had no reason even being at the table to be considered, much less do it. But I guess I said something worthwhile because they brought me back for a second interview. And that led to an agreement And I remember the day we signed the contract. I've never produced a show, never produced a commercial. I had a camera though, that's about all I had. And they signed the deal and I went into the parking lot, Michael, and I threw up because I'm like, now what do I do? Like they said, yes, figured it out. And that was a major turning point for the agency. It was a major curiosity stuff for the agency, rev, all that kind of thing, Right. And I've come full circle. So I had that agency for 25 years. This fall I decided I really wanted to invest more in what I'm doing now, which is coaching and consulting one on one. Being a problem solver. Didn't want to have the day to day responsibility of running a company anymore. So here I am, man. And. And you asked how did I get into Chat GPT? If you've been in the agency world or if you've been in the business world, you know that tasks take time. And I remember being in a workshop where it's like three years ago and someone pulled on the big screen, they talked about this thing called Chat GPT and I never heard of it. And he had some sort of prompt and he said, you know, I run this business. This is my audience and this is my goal. Could you create a 12 month social media editorial calendar for me? And you've lived in the social media world. You know how overwhelming this is like an anxiety exercise for a lot of businesses. Like, how do I use social media? What do I say? How do I do it in 20 seconds? And in front of me in this table format with headline format, tips, links, everything. The whole thing just, just popped out. And you're probably a Bugs Bunny fan growing up, you know, the whole aruga, like you just got the big thing. Like, my head's just going nuts, man. Like, I have got to figure this out. Because what I saw on the screen was what took our agency weeks to do, right? And it was probably a better, faster start. And at that moment I was hooked. I've been digging it ever since and I've been sharing my experiences with podcasts with clients and anyone I can talk to.
- [08:56]
Michael Stelzner
So how long have you had this podcast, the ChatGPT experiment?
- [09:01]
Kerry Weston
Two years now.
- [09:01]
Michael Stelzner
Okay, cool. So you got started with the podcast pretty quickly, right after ChatGPT came out within a couple of months because we're about two and a half years in as of this recording. So. Okay, first of all, love your story. Your story is such a common story that I hear from so many People, there are so many former consultants and former agency owners that are like, oh, I got to figure this out. If I don't figure this out, I'm going to be in trouble. Or they've already gotten rid of it and they're like, they have time on their hands to do this kind of stuff. What I love about your show is every couple of days there's a recording from you talking about some cool feature or cool strategy or cool technique that you've discovered leveraging ChatGPT. So I want to ask you this question for the marketers and entrepreneurs that are listening to this podcast. Why should they use ChatGPT to speed up their work? Said another way, if they do what we're about to talk about, well, what's possible?
- [09:57]
Kerry Weston
It's a great question. And the answer I give folks when they ask me something similar is we all have software tactics and tools that we use that we're burdened by trying to use all the features and trying to learn all the features like these. This thing About Photoshop, there's 1052 things you can do, maybe use 5. But we're overwhelmed by all these productivity tools. ChatGPT and other tools like it. It's not software. Like, we're not learning how to use features and functions of software. It's really a productivity partner, right? It's something that if you just use at a very basic way, you can be so much more effective, save time. Like you talked about in the intro. It's not something that requires you to master all of these functions and features in order to get any value out of it. From day one, out of the box. It can be a productivity partner. And you know that that's really, really powerful. I'll give you three specific reasons. There is nothing that can cause anxiety to a lot of people like a blank page or a blinking cursor. And you need to create something from it, right? So, you know, the ability of fast forwarding that and not just starting with a blank page is awesome. That's something that, that can really add some creativity, some brainstorming. You get to spend less time on busy work, you know, and we, we pay people. I told you I had an agency. I would pay people for their expertise, for the value they're supposed to bring. But I would. Or I would hire them for that, but I would pay them to be busy more than valuable. Oftentimes because there's so many things happening all at once. We find ourselves, even as owners, I mean, we spend too much time on busy work rather than the value of the focus that we should be on. And chatgpt and tools like it allow us to, to spend more time on the things that we add value and bring our value and then listen, just less stress and working faster. The ability of just getting through concepts and if thens and what happens and how do I do this and give me all these different perspectives. Things that would take teams to have, you know, in the agency days or in any workday is right in front of you. There's just a lot of things here, but I guess the umbrella that I would say it's probably the most amazing productivity partner I've found in my business more personally love it.
- [12:04]
Michael Stelzner
Okay, so let's get into the details now before we, anyone listening, Decide to use ChatGPT to help us regain our time. What are some things we need to consider? What are some things we need to be thinking about before we actually start working it?
- [12:19]
Kerry Weston
Well, I think the first thing, and this causes a lot of anxiety with beginners, is trying to figure out the perfect, whatever it is, the perfect prompt, the perfect project. What I share with people is maybe it's not for you, but let's go try it again. Let's be curious. So the one exercise I would say that works remarkably well, just open it up, it's just a blinking cursor, right? And it's waiting for you to interact with it. And just say, this is who I am and this is what I do. This is my company or this is my job, literally, how can you help me? And that one question right there, when I say that it's a productivity partner and not software, it's two way interactive. And that's what really impresses or surprises or amazes people, is that it talks back to you. So when you say, this is who I am, this is what I do, this is my company, this is my job. I don't know if you can help me, can you? You know, and then it comes back with an answer, Boy, that's all you need to get started right there. Like you just need to know if it understands who you are, what you do, and then you take it from there. So I would say the first thing you would do is simply just, just ask it, right? So share some basic information and just say, how can you help me?
- [13:25]
Michael Stelzner
Beyond that, when we were prepping, you had talked about how there's a couple different phases or things or concepts that tend to trip people up. Ideation, you know what the others are. You mind going over those real quick?
- [13:37]
Kerry Weston
Absolutely. So for most Projects, and we're talking computer projects or just projects in general. I think there's three phases that we can lump most of our activities into. The first one is that ideation phase, the brainstorming phase, the organization phase of what are we doing and how are we doing it or why are we doing. It's all the planning. Let's skip the second one for a second, go to the third one. The third phase is where we're perfecting it and we're editing it and putting the polish. This is basically where we're putting our spin, our insight, our expertise to play the middle. That second phase I call the busy middle. And unfortunately that's where we spend most of our mental bandwidth, most of our mental energy, most of our time trying to go from ideation to finished product. And oftentimes, and I know I was like this in the agency world and in business beforehand, you're so exhausted oftentimes, if this is meaningful to you, to go from ideation to that first draft, that you don't spend as much time as possible on that perfection phase on phase three, which is the polish, that's the insight, that's your angle to make it the best it can be. And so it's good enough, right? We rush it up. Perfect enough we used to call it. And so the busy middle is really where the value comes into play with folks that are in this world of trying to go from nothing to something where ChatGPT, I found better than anything I've used, can condense from a time point of view, that busy middle, so you can spend more time in ideation and then shift that busy middle time, that busy task time to more of the perfection of the pause phase so you can put more of you into it, right? And that's really where the value, I think comes into play when we use a tool like ChatGPT.
- [15:15]
Michael Stelzner
You know, when I think about myself, I have this phrase I use internally called the flywheel effect, which, which I don't know where I picked it up, but I feel like ideas are a dime a dozen, they're free, they're everywhere, and you can get them non stop from AI. But once you grasp on an idea and say this is the idea that I want to go with, then I feel like there's that important where you have to pull really hard on the flywheel, you know what I mean? To kind of start it spinning. Imagine like this massive wheel, right? And you got to give it a good pull, but once you give it a good pill, then it starts moving and then all of a sudden I feel like everything else falls in place for creatively minded people myself. And I kind of feel like that's the busy middle. Right. Because it's like once you kind of have a sense of what you want and why you want it, then it's a question of now how do I actually begin this darn thing, right? And how do I actually noodle with it and stuff? Is that what you mean when you say busy middle?
- [16:10]
Kerry Weston
Yeah. And listen, maybe you can relate to this, maybe your listeners can too, no matter how many times you've done something. I personally have been in the position where I know I've done this a hundred times, but I'm still anxious about how do I go from nothing to that drought. Like how do I actually move momentum? How do I. When you said that flywheels thinking price is right, you know, that big. Yeah, yeah. And I use the same analogy with a coal engine on a train. You know, that first mile is a lot of energy, a lot of effort, but once you get going, the momentum carries you forward. But yeah, that busy middle is all that messy work. Like we all have these ideas. Circus music, whatever. I kind of resonated as a. Or think of it as a puzzle, upside down puzzle on a table. It just dumped it. Now you've got to make sense of it, you've got to organize it, find the edges. And that kind. That busy middle is all that energy and goes into play. And honestly, it's where all that time goes. And once we do something, we can probably do it again in half the time. But most of the time we spend too much energy in that busy middle where we could be expediting. And that's what this tool really helps you do.
- [17:10]
Michael Stelzner
Okay, so let's assume someone listening right now has an idea. They're sitting on it because they don't know what to do about it. Right? Yeah, they've picked the thing. For me, for example, it's going out and hiring someone. I've done it before, it's been a little while and. And I'm stuck. You know what I mean? Getting started for someone else, it could be something completely different. So where do we begin then? If. If we want to regain some of our lost time and we want to use chat GPT to do it. And maybe we've embraced what we just talked about, that there's the ideation phase, which some people might be needing help with. There's the perfection phase, which is like the last part of the polishing. And then there's that huge chunk of the pie in the midd where there's all the hard tasks, like, what do we do now? What's the next logical thing that we would need to do?
- [18:00]
Kerry Weston
I'm really glad that you use the hiring analogy because one of the things that I've shared, I learned this early on is that technology can, I'll use anxiety and cause anxiety. It can make people feel uncomfortable. And it's a computer. And computer gets in the way of being curious and courageous to do things. And artificial intelligence, that whole AI thing sounds so technical and kind of divides you into those that know how to program and those that know computers and those don't. And so early on when I started teaching and coaching this, I just got rid of artificial intelligence as the meaning of the AI acronym and I got into amazing intern. And so I think if you could just picture ChatGPT as an amazing intern. And I usually point to an empty chair and I say, listen, if you had someone sitting next to your desk who was simply just there dedicated to help you 247 had amazing capability, right, but needed your guidance, what would you have them do? What, what is in your personal or professional life, not driving the kids around and mowing the lawn, but on your computer, what are the things that this person could do? And that mentality shifts a little bit. So like, oh, okay, I see. So if I. So I do this a lot, right? So find that one task. What I mean by this is find that one thing that you do a lot of that maybe you could delegate or maybe you could make easier to that amazing intern. Just find it. Start simple, right? Find that one repetitive task and then again, talk to it. I think a lot of folks that haven't had success with ChatGPT or haven't tried it or afraid to, it's because they don't really know the next steps. And when I say just talk to it, literally have a conversation and share it. And when I say the amazing intern, it would be like, let's talk about social media posts or proposals or whatever. It takes me too long to do this strategy report of this proposal. I want your help. Do you think you could help me? Right, it's going to come back and say yes. That's the one thing I say here with ChatGPT is it never says no or I don't know, it always gives you an answer. So we need to be, we need to be mindful of that. But I always ask it, what do you need from me? And I'm talking to Chat GPT as I say this, what do you need from me to better understand or best understand what I'm asking so we can be successful together? That's really where I start, is I talk to that intern just as a person sitting in my office. I want you to do this. I'm not going to put the expectations on you that you know exactly what I mean, because it's not an ESP or a mind reader. But here's what I'm thinking of doing. I want your help, right? Maybe give it a little background. I create these proposals for new projects, whatever it might be. Can you help me? What do you need from me? What questions do you need? What additional information do you need from me that would help you do your best work? That's exactly where I would start. And then I would just go on. You know, it's going to ask you back and forth. And then it's just like that person, you know, you want to give them examples. Like, here's some examples of what I'm looking to do. Here's the background, here's why I'm doing it, right? Like, here's the perspective or the reason. Here's the audience, right? Here's who I'm doing this for. Here's what success looks like. All of those things that if you had a new person sitting at your desk, again, back to the intern, you'd want them to understand. Where most people fall down here, Michael, is they have this intern come in and they say, go create my proposal. That's all they give it. It comes back with some generic garbage. And they say, oh, this CHAT GPT thing isn't for me. I don't know what all the hype is about. You would never expect a person to succeed that way. And I give that same advice with the intern that is ChatGPT.
- [21:25]
Michael Stelzner
First of all, you gave a task of a proposal. I want to dig in on that a little bit. But before we do that, I think it's wise for us to talk about kind of the types of tasks ChatGPT is ideally designed to do? Because my guess is it, it's not any task, even though it's capable of doing any task. Does there need to be certain kind of repetitive, structured steps to the task in order for it to be successful? Like, let's just talk about that, because just. Or maybe it can do anything. I don't know. What's your thoughts on that?
- [21:57]
Kerry Weston
So yes and no is my answer. And it all comes down to what I call attribute sets. Okay, so I just mentioned that Example of the proposal. Like, here's an example, here's the audience, here's what success looks like. Here's the reason we're doing it. Early on, I realized that I needed to create what are called attribute sets for me to be successful long term. And here's what that means to me. Okay, I know I'm going to be doing a lot of writing, so why don't we get to the point where you understand. Let's find the recipe of my voice, tone and style. I want you to understand my voice tone and style. And so I'm creating pieces of the pie. Think of this as a recipe for success. So I have writing samples. And so early on I said, hey, Chad, here's 10 of my blog articles. I want you to analyze it because I wrote these. I want you to analyze my voice, tone and style, and maybe even structure. What are the patterns that you see? And it brings me back that list. Now, here's an important tip. I never, ever use ChatGPT as the storage engine for anything. Every time I have something of value, I get it out. I store it in an external document.
- [22:58]
Michael Stelzner
Okay, why is that?
- [22:59]
Kerry Weston
A couple of reasons. Yeah. Because I never trust it to bring back the exact same thing over and over again. And those that have maybe used it will see that sometimes it gets tired, ChatGPT will get tired and start to get lazy. And it'll take a really complex thing and start to summarize it at times. But I also want to make sure that I own my own thing. And so all this information that's valuable, I'm going to have a backup. So I just put it in the document somewhere and I'll bring it in later or use it later. But I want to make sure I have it. I'm not trusting software for this. And so voice, tone and style is one attribute list that I would have. Another attribute list that I would have is my own audience, my best fit audience. So if I'm selling a product, providing a service, we call these Personas, Right? Who is your best fit? Right. Fit audience? And I usually get an attribute list of demographics and mindsets and frustrations, what they're feeling and what they're needing. And that list serves us really, really well. And then the other list, which is what does success look like? So if you've created something over and over again, let's go back to the proposal, the ability for you to model that and say, hey, here's three or four proposals. I want you to understand what they are and why they are the way they are. As far as a structure point of view, could you analyze this for me? It's a great analyst. So if you give it additional information, writing or attributes or documents, it can analyze it and create recipes. So I think of this as really a recipe structure. Once I have those, I can do some amazing things. So the reason I gave you that answer, Michael, as you said, is it good for specific tasks? So the answer is yes. If you can define something in structure like I did with a proposal, then you can do it over and over again. But it does need context in order for that proposal to be relevant. So you're going to need to provide answers to questions or what's in the proposal. It will know the structure, but you'll need to provide the details. But if you're not doing something over and over again and you've done the work, the homework that I just laid out, if I know my mindset, my voice, tone and style, my audience, all these things, and then I can put all that together and have a really, really fantastic strategy partner to do something one time, do a one off. I could problem solve an issue. I could have it be a CEO with me or a co CEO, have it be an accountability partner because it's taking all of these things that it's learning about you, that you're building these lists, right? And it's coming together to do creative things for you.
- [25:24]
Michael Stelzner
So let's talk about this proposal. Give me an example. Let's go through like an actual use case here.
- [25:31]
Kerry Weston
My favorite thing to do is to record calls with prospective clients because that transcript I think is the new gold. Transcripts are the new gold when it comes to AI. And here's why I say that, because I've taken my proposal, or what I call a recap memo and I've modeled it in ChatGPT. And by that I mean I've given it, let's say five or six that I've done through the years. And I say, listen, here's the recap memo, here's my proposal. Can you please take a look at it so you can understand the structure? I want you to understand what I do each time and why, even though the information may change, the way I do it is consistent and I want you to understand and it will. So it'll build out the recipe of the proposal. Now it's what I call an empty mansion. There's no furniture, there's no context, there's nothing. But it knows the structure. And so with that I can say, okay, chat here now. That you have the recipe from my proposal. I'm going to take this transcript from a sales call and give it to you. I want you now, now that you understand not only what the proposal is, but why the information is where I put it. I want you to score over and take a look at the transcript, grab all the information, put it in the proposal. Because you know how I write, you know, you know my style, find the relevant information, create that proposal for me so I can take a look at it once. I have a sales call and I have a transcript from the conclusion now of a sales call to the first draft of a proposal, two minutes. And then I can edit it and make it mine. But two things are happening here. I can spend more time focusing and being present on that sales call because I'm asking meaningful questions to pull out details because I know it's being recorded. And because ChatGPT not only knows what it looks like from a structured point of view, but the reason and logic behind the sections. It knows what to pull up from the transcript and put it in the right order. So it's a huge productivity exercise and I'm getting the words right from the client's mouth to put into a proposal, so I can say how I can help.
- [27:36]
Michael Stelzner
Okay, so let me just kind of try to reverse engineer some of what I believe you've specifically been doing. So it sounds to me as if you have created a. My guess is you started with a simple thread inside a chatgpt and you gave it some examples and you asked it to study the examples and. And event essentially to reverse engineer the elements of it that you then put into a custom GPT. Am I close to accurate on that?
- [28:01]
Kerry Weston
Yes, for both the literal structure, what goes first, what goes second, what goes third. But I also want to analyze what am I putting in there for information and why. What's the reason for that section to be there too? Yes.
- [28:12]
Michael Stelzner
Okay, so talk to me a little bit about that part of it. How to. How do we, like, ask it to analyze it and provide that context and all that stuff? Because you have to train it, like you said, if we're treating it like this intern, we're going to have to kind of somehow not just have it like reverse engineer what it is, but are we going to go back in and paste that into like a Google Doc and then add a little bit of the why behind why these sections are here so it understands it?
- [28:36]
Kerry Weston
Great question. So here's one section, literally from my recap memo, which is issues and problems. You're dealing with because if people come to me, they have an issue or a problem.
- [28:44]
Michael Stelzner
Right, right.
- [28:45]
Kerry Weston
And so I've trained it to look at this proposal. And I have a section that says the issues and problems that we discussed. And there might be four or five paragraphs or bullets or whatever it is. And what I've done is I've asked it to analyze what's in that section. Like, what do you see as the reason?
- [29:02]
Michael Stelzner
The transcript of the transcript.
- [29:04]
Kerry Weston
Okay, I'm actually in the proposal. Oh, yeah, okay, right. I'm in the proposal. Like, why do you think this exists? Just, just like I would do with an intern. Like, read this and tell me your thoughts. Why do you think this exists? What do you think I'm trying to convey? Why do you think I put this in here? Right. These are the questions I'm asking Chat tv and when it gives me the answer, I say, yeah, exactly. Right. So again, back to your reverse engineering. So as part of the recipe, when we go to, to extract information from the transcript, I want you to find those nuggets. I want you to find areas where we talked about problems and issues and frustrations and things. And then this is where it's going to go, Right? And it's okay, great, I understand this section, section 2 is called Problems and Frustrations. And these are the things I'm looking for in your transcript. And this is the reason. So I'm going to go find them, bring them in here and organize it accordingly.
- [29:50]
Michael Stelzner
Now, when it's giving you those answers to those questions, I'm assuming you're pulling that into your master document and making it all clean and then ultimately becomes a system instruction or whatever they call it for the custom GPT. Is that right?
- [30:04]
Kerry Weston
Yeah. And here's the beauty of ChatGPT. ChatGPT is the best way to use ChatGPT. And here's what I mean by that. You can go through, stick with this proposal section by section. Do you understand what this is called? Do you understand why it's here? Do you understand what I put in it? Yes. Okay, great. And we'll go section by section, right? And we'll get done. And I'll say, great, now that you understand everything about my proposal, I want you to give me the instructions that I can use to work with you next. Right. To pull that information from a transcript and put it into this thing. And it goes. Okay, again, back to the intern. This is all what you do with a person, right? You just, you'd coach them and make sure they understand and trust and Verify. But then at the end, I said, okay, great. You. You understand now, because we've had this conversation what this document is and why I do it. Now, if you were to do this on your own, right, Mr. Intern, Mrs. Intern, can you tell me the steps you take to put this back together again? And it will say, okay, great, here's exactly what I would do. And it lists out. I mean, ChatGPT is doing this for you. It lists out. This is what I would do. This is what I would go find. This is why. And you just perfect that. Yeah, that looks really, really great. Right? So there's three things you can do with that information. Once you've got that acknowledgement right, you can, one, put that into a document and bring it in time and time again into just a new conversation in ChatGPT. So you can use those instructions as you need it. ChatGPT has a function called projects, where you can create. It's literally called projects. And you can create instructions for the project, and you can paste those instructions into the project so it knows what its job is. Or you can create what's called a custom GPT, which is the most specific and the most focused and probably the most advanced tool inside ChatGPT. So it knows its only purpose on earth is this. And we'll do this over and over again. And you can give it all the instructions and all the clarity you need, but having CHAT GPT understand it and give you those instructions is a beautiful thing.
- [31:56]
Michael Stelzner
I love it. Okay, so so far, what I'm hearing you say is you start with a nice long dialogue with ChatGPT and you tell it what is it you're trying to achieve, and you ask it what does it need in order to be of assistance to you ask it a series of questions. And somewhere along the way, we've also developed these sets, as you refer to.
- [32:16]
Kerry Weston
Them, of the attribute sets.
- [32:19]
Michael Stelzner
Yeah, the attributes test, which might include things like ideal customer, or just give us a quick couple of other attributes that might be in there.
- [32:25]
Kerry Weston
Frustrations, goals. Do they own a business? Are they frustrated with anything? Do they. Do they trust you? Do they not trust you? Have they been burned?
- [32:32]
Michael Stelzner
Okay, but what beyond the customer set, is there other sets? Remind me of the other sets that might be in there.
- [32:37]
Kerry Weston
So three things I use is my voice, tone and style, my customer set, and then the whole purpose of my company. So I want you to understand my company, my audience, and what I do.
- [32:45]
Michael Stelzner
Are these ultimately, like one paragraph kind of summaries that are going to be.
- [32:48]
Kerry Weston
No, these could Be they're huge. I constantly am updating this based on new experiences, but I bet each one of These is a 8 and a half by 11, detailed with paragraphs and bullets.
- [32:58]
Michael Stelzner
Okay, so we're taking these sets that we've talked about just now and we're developing those out presumably on our own. Or are we using AI to help us do that? Or is that mostly just institutional knowledge that you're translating right into it, like a Google Doc?
- [33:12]
Kerry Weston
Again, I always use ChatGPT first to do things with ChatGPT because it's such a great objective partner. And so here's one exercise I do with clients. I grab a website URL, okay. And it says, hey, I go to ChatGPT, I start a new conversation and I say, I want to build an attribute list of my ideal best fit customer. I'm going to give you the homepage of this client. You tell me what you see and you tell me what you think the attributes of a best fit. Right. Fit client are or customer are. And I'll tell you, Michael, it's 90% dead on out of the gate. And so I just take it from there.
- [33:49]
Michael Stelzner
Okay, so we've got these attribute sets that we were just talking about, right? Like voice, style, ideal audience, and even what does success look like? I think was another one you talked about modeling.
- [33:59]
Kerry Weston
Yeah. Modeling though.
- [34:00]
Michael Stelzner
Yep. And then we've got those off to the side and those are things we could probably use over and over again, I would imagine, especially the ideal audience and your voice. And then we've. We've had this thread now, for example, I want to hire a project manager. Okay. For a project that I'm working on. Walk me through, like, what you would do if you were me. Like, let's just talk me through it a little bit.
- [34:21]
Kerry Weston
Yeah. So the first thing I would say is, do you have a basic job description?
- [34:24]
Michael Stelzner
Yes.
- [34:25]
Kerry Weston
Okay, so I love this question. So most people put up a job description on indeed or wherever and on a website, and then they wonder why they have a hundred applicants and 95 of them are just no one they'd hire. So I said, okay, let's take that job description. And you've got your productivity partner in Chat GPT that knows about your company and your client and your business and what you and has all this knowledge. Right. So take that job description, upload it, because you can upload files directly to ChatGPT. And I say this, hey, chat, here's a job description. It's got features and functions and jobs responsibilities. I want to make this Compelling to the perfect candidate. Okay, So I want you to think about. Because you know about my company, because I'm assuming we've done this work about, you know, my company and you know what I do and you know, my best fit audience, I want to hire the best or I want to attract the best candidate possible. So here's what I'd like you to do. Think about the position that they are in right now, whether they're working or not, and what do they want to see in a job? What do they want to see in a company? What answers are they looking for? What information can we give them right now about the job, the company, the description, their. Their position that could make this a living and breathe. Like, let's make a story out of this job description. Let's talk to them directly. Let's get real niche down. So instead of just having generic bullet points with jobs and responsibilities, let's talk to this person. Let's bring this job description to life. And I call this the empathetic position. When we take the empathetic position of either an applicant or a customer and we start looking at what we're saying and presenting and putting out there through their lens, right? This goes back to Marcus Sheridan's. They ask, you answer. We're being obsessed with what the customer or the applicant needs wants. Let's put ourselves in that position. When we do that, we bring that job description to life. And. And I've talked to dozens of HR managers on this one. We already know what the first 15, 20 minutes of every job interview is going to be. So let's get that out now. Let's just. You probably have these questions, so let me give you these answers, and not just in a four word bullet. I'm going to give you thorough, objective, honest conversation. So to I guess succinctly answer your question, let's bring that basic boring job description that we all have to life through the empathetic position of the applicant and give them what we know they want to see so we can inspire the very best applicants to want to work in our company.
- [37:01]
Michael Stelzner
Perfect. Now, let's assume we actually don't know how to find the applicants, but we want ChatGPT to assist in that part of the process because maybe it's been a while since we've gone out and hired someone. How could we also use chat GPT in a scenario where we don't actually know what the steps are? Like, let's just say I don't know where to go post the jobs and how to actually recruit the people. But maybe chat does you understand where I'm going? Like, this is an example of a task that's like not very common, right? And maybe in the past I had someone else doing this for me and they're not here anymore. Now I'm doing it for myself. Could my AI assistant help me in this particular situation?
- [37:38]
Kerry Weston
So some of these features are available only in paid. So there's paid and free versions. So what I'm talking to you now is about paid. And it's 20 bucks a month and it's temporary, right? You could get in it for a month and get out for four and come back in if you want to. But in paid, they have a search feature. So you can use ChatGPT as a search engine or as, you know, another search component. So what I would do in that scenario is I would literally say, just like that, intern, right? I would say, hey, Chad, I've got this amazing job description. This story I want to put up, but I don't know where to put it. It's been a while I haven't done this. Maybe the person that used to do this left. I don't know what to do now. Could you help me? Could you share with me where I should be putting these? This job description. Are there resources? Are there applications? I don't know what to do, so can you help me? I would literally say that. And then it's going to go, yep, I can help. And then it's going to come back and give you all kinds of resources.
- [38:30]
Michael Stelzner
So the reason why you're going to use search in this particular case is because it's going to go out and find the most relevant information out there. Is that correct? Or do you feel like even if you didn't use search, it would still be able to answer that question?
- [38:41]
Kerry Weston
I think both are true. I think it still could. But however, the data set in Chat GPT and this and listen, something's going to change that. I know by the time this podcast interview is over. But you know, it's got a limited. It's not real time. So chat in and of itself is not real time data set. It might be up until like 2023 or 2024 at this point, I don't know. But the reason I use search is because that enables the power, the reasoning power and the understanding of ChatGPT to go connect to the Internet so it can get up to date in real time information. And so it would go out and search. And the great thing is, is it's bringing Back sources now. So it's sharing with you. It might go out and find some articles, literally. It might go out and search for you and find articles that were written two weeks ago that say the five best places to post your janitorial jobs to get the best result. I mean, may find stuff like that for you and bring it back and then show the source. But the reason I would use search is to tap into real time information as much as possible.
- [39:36]
Michael Stelzner
Okay, perfect. So let's get back to something you said earlier, which is projects versus custom GPTs. So we've covered custom GPTs pretty extensively on this podcast. They're not super complicated to create, but projects is something we've not really talked about. So can you kind of give us a pro and con? Like I can tell you right now, the con to custom GPT is it has no memory and it only remembers what's in its instruction set. It's like you're talking to someone who just met you for the first time every time you talk to it. Right. Is that true?
- [40:05]
Kerry Weston
I don't think it is. I think the memory has gotten much, much better.
- [40:08]
Michael Stelzner
Do you think that's because of the General memory of ChatGPT? Okay, so what's the benefit of a project versus custom GPT?
- [40:15]
Kerry Weston
So custom GPT, if I remember right, were out of the gate first and then projects came in later.
- [40:19]
Michael Stelzner
Yeah.
- [40:20]
Kerry Weston
And custom GPT were kind of like this marketplace idea that OpenAI had. And then I think it got a little complex and they were going to monitor. So here's the one thing. You can do similar things, right? You can provide similar resources and similar instructions. But for me, I find that organization of your past work is probably the main difference. I'm sure there's some subtle differences, but organization of your past work, meaning if I open up a custom GPT and do something, that conversation just falls in line with every other conversation that I have in my sidebar. So if I use it religiously on a regular basis and I've got a hundred conversations, that conversation just kind of falls in line with the other 99 conversations I have. With projects, it's more of a visual organization. So think Google Drive or OneDrive or whatnot, where you have a folder system. So this is the way I look at it, and it's the easiest for me to understand. The benefit is I create a project, I give it the instructions, and I can give it resources and all kinds of stuff, just like custom GPTs. But if I have five conversations where I've used that project, when I click on project, I see the conversations that are directly related to that project visually. So from an organizational point of view, I can quickly see the previous work and maybe build upon it, whereas a little bit more hunt and pecking. And when I use custom GPTs, because that's the one thing, listen, it doesn't do math. It doesn't do math very well, and it doesn't do time or day very well. Custom GPT as a whole. But it also doesn't organize things very well. It doesn't organize conversations, doesn't name them. The search feature is still weak. And so I think from a directly relatable, everyday practical benefit, the ability of using projects so you can organize the work you do under it is huge.
- [42:00]
Michael Stelzner
Is there anything that projects can't do that custom GPTs can do?
- [42:05]
Kerry Weston
The one answer I know to be true, and I'm not gospel, I'm not an expert on all of this stuff. I've just made a lot of mistakes. I've learned from those is I know that custom GPTs now have APIs to go out and plug into other things. You can expand your capabilities from a programming point of view. And I know projects don't at the moment.
- [42:21]
Michael Stelzner
What do you generally use yourself?
- [42:23]
Kerry Weston
Actually, after talking to you, you were telling me about projects I've been exploring the past week more and more. I use custom GPTs more because I built more of them.
- [42:31]
Michael Stelzner
I see, Right.
- [42:32]
Kerry Weston
So my brain lives there. But I'm starting to get into that. I'm starting to use Claude projects a little bit more again since talking to you, you opened me up to that and I heard a couple of your guests talk about other things. So I'm expanding my horizons. But I'll say from a pattern point of view, I'm an accustomed GPU world.
- [42:46]
Michael Stelzner
Okay, so system instructions. We Talked about how ChatGPT can kind of create this big set of instructions. Do you have any tips on, like, since you use custom GPT so much? Because so much of what we talked about could simply be just be a chat thread. You know what I mean? Like, so much of what we just talked about could just be like this evolving chat thread, but maybe some people want to actually, you know, they've done it enough times that they should make a custom GPT. Just give a couple tips on like what people need to know about creating custom GPTs, or maybe advanced tips on like how to make it work better or whatever.
- [43:20]
Kerry Weston
Yeah, and I'll share with you at the end too. I've got some resources that will double down on this, the advice I would give you is use ChatGPT to create, Polish and improve upon the instructions that custom GPTs have. It thinks like it wants to think. And what I mean by that is sometimes as humans, we miss things or we give it more information than it needs and so it can succinctly shrink. It can really shorten things inside the instructions themselves. So I would say this. Start with the instructions, whether you have ChatGPT create them or not from a custom point of view, use them and then share your experiences with Chat GPT and say, you know, I've used your instructions. This is what I'm getting. I don't like this, this and this. Can you help me? And it'll modify. So yeah, I understand that. And it'll modify, just simply replace and go. So I guess the tip I would use is use ChatGPT to create and improve upon the instructions that you put into a custom GPT.
- [44:16]
Michael Stelzner
Is there any other tips that you have about custom GPTs or chatGPTs as a whole that you want to mention?
- [44:21]
Kerry Weston
I do. There's an exercise that I do in training where I ask everyone in the room to raise their hand and they typically do, and they'll raise it halfway and then I'll say, raise your hand as high as you can and they'll do that. And then I'll ask somebody, how come you didn't do that before? How come you didn't raise it that high? They'll say, well, I was just doing what you asked. Said, well, chatgpt does the exact same thing. And so the three words that I love to share with folks is tell me more. So anytime you ask ChatGPT to do something, it's going to give you that equivalent of raising your hand halfway. If you say tell me more about a certain section, it's going to go deep. So never take the first thing it gives you as maybe all it knows or all it's going to give you. The tell me more is going to get you into the deep dive and there's a lot more there behind you. And then I think the last thing I would say is if you're ever unsure, simply ask it. You've heard me mention a few times during this conversation. Just have a conversation with it. It's amazing. If you ask ChatGPT, can you. Or how you. Or can you just ask it, can you help? Or how would you do this? Or how would you approach it? Or whatever it might be, it will give you answers. And even if the answers aren't perfect, back to your first question as to why should folks consider using it? To have a brainstorming partner, to have a productivity partner, to have a sounding board. To have something that can hear what we say and give us back our thoughts, organize our thoughts and help us be better. That is an amazing ability for someone that lives on the computer as part of their work.
- [45:48]
Michael Stelzner
Kerry Weston, thank you for your insights. If people want to follow you on the socials or interact with you there, or they want to connect with you you in your business, where do you want to send them?
- [45:56]
Kerry Weston
You can find me on LinkedIn under Kerry Weston and you can find the podcast and all my contact information@chat GPT experiment.com and you were asking about best tips and custom GPTs. There's a couple free guides on the homepage there to help you model step by step, both custom GPTs and best practices for the things that we talk about today.
- [46:14]
Michael Stelzner
Carrie thank you so much for your insights today.
- [46:17]
Kerry Weston
Hey, thrilled to be here.
- [46:18]
Michael Stelzner
Hey, if you missed anything, we took all the notes for you over@social mediaexaminer.com a60. Also, be sure to follow us here on whatever app you're listening to. And if you've been a regular listener of this show, we would love it if you would share it with your friends, if you would give us a review on whatever app you're listening to. And if you want more shows, we've got the Social Media Marketing Podcast by me and the Social Media Marketing Talk show. This brings us the end of the the AI Explored Podcast. I'm your host, Michael Stelzer and we'll be back with you next week. I hope you make the best out of your day and may AI help you become more successful. The AI Explored Podcast is a production of Social Media Examiner. Just a quick reminder before you go. If you're ready to become indispensable in the age of AI, the AI Business Society is your solution. Join now and secure your discounted membership by visiting social mediaexaminer.com AI I can't wait to see you inside the AI Business Society.