Master ChatGPT Prompting & Skyrocket Your Productivity with AI Expert Jonathan Green

Executives, entrepreneurs, and professionals, this episode is for you. Learn how to harness the power of ChatGPT and unlock its potential for increased productivity and creative breakthroughs. Join us as we chat with AI pioneer Jonathan Green, host of the Artificial Intelligence podcast, who will share his expert insights and practical tips on mastering ChatGPT prompting.

In this episode, you'll discover:

• The secrets to perfecting your ChatGPT prompts and turning ideas into high-quality outputs.
• Proven techniques for maximizing your productivity and creativity with AI assistance.
• In-depth discussions on the value of premium AI features and their benefits for creators.
• Expert guidance on navigating the growing GPT marketplace and utilizing it to your advantage.

By the end, you'll be equipped to:

• Transform ChatGPT into your trusted digital partner.
• Boost your brand and stay competitive in the ever-evolving digital landscape.
• Unlock the full potential of AI and revolutionize your work and life.

Whether you're a seasoned entrepreneur or just starting your AI journey, this episode is your roadmap to success. Don't miss this opportunity to learn from a leading AI expert and take your productivity and creativity to the next level.

Listen now and unlock the power of ChatGPT!

Enjoy!

CONNECT WITH JONATHAN:

Visit https://servenomaster.com/master for your free gift. One secret ChatGPT prompt to make them a top 1% user in 5 minutes.

Listen to Jonathan's podcast, The Artificial Intelligence Podcast, wherever you like to get your shows!

Links:
Roberto on Instagram http://www.instagram.com/robertorevillalondon
Tailoring Talk on Instagram http://www.instagram.com/tailoringtalkpodcast

Credits
Tailoring Talk Intro and Outro Music by Wataboy on Pixabay
Edited & Produced by Roberto Revilla
Connect with Roberto head to https://allmylinks.com/robertorevilla
Email the show at tailoringtalkpodcast@gmail.com

Transcript

Speaker 1: 0:00

You ready? Yeah, of course. Welcome to the Tane Ring Talk Show with your host, roberto Rivilla, a bespoke tailor, menswear designer and owner of Roberto Rivilla London Custom Clothing and Footwear. I activate your superpowers through the clothing I create and the conversations on this podcast. We will meet self-starters and creators to learn about their journeys, while they share valuable lessons to help you be the very best you can be. Please support the show and my guests by subscribing, and it helps so much if you take a few seconds to leave a rating and a review.

Speaker 1: 0:42

Joining us today is a trailblazer in online business and AI technology. Since 2010, he's used AI tools like ChatGPT and Mid Journey to build a digital empire, authoring ChatGPT Profits and amassing a 100,000 plus mailing list. He specializes in helping entrepreneurs leverage AI to replace their 9-5 income with successful programs like AI Freedom and Fractional AIO. A best-selling author of over 300 books and a celebrated podcaster, he turned a career setback into a thriving business from a tropical island Tailoring talkers. Let's welcome the digital entrepreneur who turned adversity into opportunity here to hopefully simplify all things GPT and AI for you. Jonathan Green, welcome to the show. How are you?

Speaker 2: 1:39

I'm good. How are you doing? Thank you so much for having me.

Speaker 1: 1:41

I'm like pumped, like you probably sent me okay. So for a bit of context listeners, it might be a bit echoey and I apologize for that. I moved house between Christmas and New Year. We started packing on Christmas Eve, we moved just before New Year's Day and it has been a stress fest the last few weeks. I do not have my new studio set up, so I'm talking to you from the kitchen island. It is a bigger house. It is much more echoey in here. I don't even know if that's a word. Jonathan is my first guest of the New Year's, but he I've been particularly excited about GPT has been my best friend recently because it has been helping me to get a lot of things done, but I don't think I'm using it in the most effective way that I could, and I'm sure a lot of you are probably nodding your heads right now and he's going to help us work all of this stuff out. Jonathan, thank you so much for joining me today. Now you are. Your accent does not betray where you are based.

Speaker 2: 2:39

Yeah, I've lived so many different places. I lived in Los Angeles till I was 10, then Tennessee till I was 18, then I bounced around Eastern Europe, eastern America, that I lived in England and then Wales and then Japan and all over the South Pacific. So I'm one of those people who, whoever I'm talking to, my accent will start to sound like them. It's not intentional, it just happens. So when I was living in Wales I had a little bit of a Welsh accent and it's just becomes. I think it's because when you go up in Los Angeles you have the accent of movies so people don't detect an accent. It's like, no, it's only because you hear it in movies all the time that you don't that it's the LA sound. But that's kind of my baseline accent.

Speaker 2: 3:22

And yeah, and of course I have a bunch of things I say because my wife's first language is not English. So my kids are all bilingual or trilingual and I have a tendency to have like a lot of mixed language stuff when I talk to them. So I do sometimes say words incorrectly in English because I'm saying actually the foreign version of the word. Like I say, I realized the other day I kept saying Skellington and I was like well, that's not a word.

Speaker 2: 3:47

It's not that I don't know, and it's like, as an author and a writer, people expect you to never misspell a word and never most pronounce a word. I'm like no, I'm a writer, not an editor. And yeah, it just all comes out because I've just been traveling for a very long time.

Speaker 1: 4:01

Yeah.

Speaker 2: 4:01

And you're based in the Philippines now. Yeah, I live in Southeast Asia, so I'm always bouncing around. Right now, we're living on a small little vacation island and we've just been here for a couple years and we're just every once in a while we move around. We're looking at a couple of new islands next, but we always end up on an island. That's why I like living on Wales. I live right on the coast and that's why I like Japan island nation, so I'm definitely someone who always wants to live on an island.

Speaker 1: 4:26

For the listeners, give us a little bit of a flavor of kind of how you got to be ensconced in. I don't even know if that's a word, but you know what I mean. I've really embedded this whole world of AI, chat, gpt and so on, because I didn't even know, when I was writing your introduction and I was reading through your bio and so on, I didn't even know GPT and tools like this existed 14 years ago, and I'm sure, again, a lot of our listeners are nodding their heads right now thinking, yeah, neither did we. We thought this was something that only came about in the last couple of years.

Speaker 2: 5:00

So I've been using a lot of tools that were heading towards AI. So they were kind of pre AI tools, pre predictive tools. So when I started out, a lot of the tools I use were not intuitive. So it was even when you set up a website, like now you can install WordPress by clicking one button. When I started out, you had to create a SQL database, create a SQL password, connect the two things together, do a manual installation, learn how to log into Linux. So I've always been playing with programming very rudimentary levels, like I tried.

Speaker 2: 5:31

I was doing programming when I was six or seven years old, like in the programming club in school, but never really it wasn't really AI stuff until about three or four years ago. So I was using a lot of tools and building an online business and pushing things that were not yet AI, and three or four years ago even the tools were interesting and cute but never really useful. So I was even doing a lot of interviews about some other competitive softwares. Like two years ago I did not use this tool for anything more than a rough draft. I was like nothing this tool outputs. Should you put your name on? Because it's good for rough draft? That's it. It's good for idea generation, but it's not ready for prime time.

Speaker 2: 6:08

Only with the advent of chat, gpt 3.5, which would came out around the end of 2023, 2022. There was a November 2022 that it came out that I finally saw content. I go, oh, this is something that's actually good enough to put out publicly. So in a lot of what I've been doing is building online business, but really the heavy AI focus has only come out in the last few years, and then the full time AI 100% focus has been in just the last a little bit more than a year. Just when it finally crossed that bridge that Rubicon regocates finally good enough for a regular person to use it and get usefulness from it. For a while, it was like you have to be a superpower user to get anything out of it that's actually useful. So that's really kind of in my journey as an entrepreneur is that I'm always looking for ways to kind of improve my process, to improve the tools that I'm using and to just deliver something better for my followers, customers, clients, whatever you want to call them like the people that I try to deliver for.

Speaker 1: 7:07

I've been relying on GPT a fair amount the last couple of months to get a lot of things done and, like I said, I'm probably not using it very well, but it has already made such a big difference in my working life and my personal life. I subscribe to it. So I paid the 29 bucks a month whatever to get GPT for, which gives me dally as well. That's a whole other conversation. By the way, I kind of the best way to describe it to you, jonathan. I kind of treat it like it's a personal assistant on my desktop. So take today, for example. I was very short of time because again, with this house move, it has just been chaos. I don't know how you and your family move yourselves around every couple of years. I'm just not coping very well at the moment. So I'd written, I'd done my homework on you, I'd written your intro. It was so long because you are so talented, there's so much you've achieved, there's so much that you do, and I'm like, oh my God, this is like two pages. I can't do this as a podcast intro. So I said to chat GPT. My guest today on Tainering Talk if you use what you've learned from previous conversations about the podcast is Jonathan Green. I'm going to pump my draft in and can you shorten that down to an intro? That's no more than 20 seconds.

Speaker 1: 8:20

And then GPT did its thing, and then I edited it a little bit and then that's what you heard me say five minutes ago, and that is pretty much how I use it. You know, whatever it is that I'm doing, I will turn around to GPT and I'll say, right, this is what I'm doing right now, or I need to do, or I'm about to do. I'm going to give you whatever here it is. Can you do this, this and this, and this is the result that I want. Hey presto, it produces something.

Speaker 1: 8:47

Now, it's not always exactly what I wanted, and so then I kind of have to ask the question again in a different way, or I might have a little mini argument with it. So some of my chat GPT chats you probably laugh at because it would be like, no, that's not what I asked you for. Listen, go back again to what I said to you in the first place. And but now I want you to blah, blah, blah, blah, blah, blah, blah, blah, blah, and then eventually I'll get to what I was. I was getting to. Do you hear that a lot from your own clients that come to you seeking help?

Speaker 2: 9:16

and with all of this stuff, yeah, I mean you're pretty advanced user compared to most of the people that I work with, because most people don't even get to the point where they're getting a useful output. The fighting with chat GPT that's normal. I do that all the time. You're going to always have that because it's not fixed. So we're used to software tools where if I type in, if you use a calculator, right, I put in one plus one, it will always be the same answer. Chat GPT doesn't work the way. It will give you a different answer every time because it's designed to be a little bit of a thinking machine, so that means you can ask a question and ask a question tomorrow and you get a different answer tomorrow. So sometimes it's going to give you what you don't want, and that's kind of the price we pay for it being a little bit smarter. So what you're experiencing is the most important thing people realize, and that is that chat GPT is a generalist. These AI tools, these text generators, whether using chat GPT or Claude or whatever name, a grok from Twitter or whatever the one new name they choose for the Google AI all of these are generalists, which means that they know a little bit about everything. The advantage of that right is. The same goes. It's the jack of all trades but the king of none. So it knows a little bit about everything, and that's because of the way it's trained.

Speaker 2: 10:33

So imagine if you're a tailor. You know about fashion. Imagine if I gave you a database of every single fashion design ever created, but I didn't tell you which ones are good and which ones were bad. Statistically, most designs are bad, right, which means that if I told you, pick from this randomly, and you picked a random dress, you probably pick one that is bad designed. You know whether it's a stitching problem or it's just tacky, or it's a design like a 1980s prom dress. And that's the first issue you run into. Is that? Well, because the database is so broad and it doesn't know how to separate good from bad. It doesn't have that ability. So that's why it needs the human interaction. That's why you're running into those problems. So it's very normal and once we understand, oh, that's how it's trained. But you can then say, hey, I want to work on a new design with you, right, I want to work on designing a new outfit. Here's the only designers I want you to pull from their databases. Here's the 10 designers that I most admire.

Speaker 2: 11:30

Now what you've done is actually eliminate everyone else. You haven't created a database, you've eliminated the bad, and that's how you can get much better responses. Most people think, oh, I have to train it or give it more data. It's like no, what it needs is less data. So it's the same thing.

Speaker 2: 11:45

If you look at every website ever made, most websites are terrible, right. If you look at all the everyone made a website in the 1990s and the AOL days and they were all like you couldn't turn the music off, it had flashing icons and it was like bright colors and it was like so offensive to the eyes. But there was more of those exist than good websites. So if we don't know how to wait good from bad, we're always going to end up with bad, because there's just quantity, right, most books nobody reads. So most poetry golly, like most poetry is bad. So if you say, write me a poem, of course it's going to write a bad one, because its database has mostly bad online poems, probably from Tumblr. So once we start eliminating that and going, oh, narrow it down to this data set, then you can get really good. So that's the first thing to think about.

Speaker 2: 12:27

The second thing is that we interact with chat between AI tools in a command response structure. We say give me this, this is my issue. And that's exactly what you did, which is normal because that's how they taught us to do it. There's a better way and here's the way, because when you do that, you have to ask the question in the perfect way, which is hard because it always changes the way it was. I want to write the perfect introduction for my guests today. I've already written an introduction, but it's too long. What information do you need from me? And it was like, oh, give me the one you already wrote, and can you give me examples of past introductions you've done that are really good and what it really needs? And this is how I do. All my creation with chat, maybe even more in creating custom coding for people, is an example of a good output, right, an example of a great introduction you already done.

Speaker 2: 13:13

And the raw data. So it would be my biography for this week when it has that. So it has raw data and it has a great output. Then you know, oh, I want it to look like this. You can compare when it makes something. Is it close enough to what you want.

Speaker 2: 13:25

So I like to ask in the form of a question, because it moves chat GPT into interrogative mode, which just means conversation mode. The way it's designed is that chat GPT never thinks you will ask a follow up question. It's designed to always answer your question with the assumption you won't answer a follow up. So if you ask a bad question or give a bad prompt, it won't tell you that. Because then it's way. It's like a genie with one wish. You can't waste that one wish. But as soon as you switch to ask me questions and say if you need more information, tell me, it goes oh, you've given me permission, yeah. And then it will tell you oh, I need this piece of data If you give me more of this. And it means you don't have to be a genius. Now chat GPT can be the genius. It moves the pressure on you from being a prompt engineer like coding prompts. All of that stuff makes me crazy.

Speaker 2: 14:17

I'm like I don't want to learn a new language because they change. This just happened a few weeks ago. Mid Journey just released an alpha version six and in their announcement there's one sentence that a lot of people missed, which is the prompting structure has completely changed. You'll probably have to relearn how to prompt, and I was like I can't think of anything worse to hear, right, after you spent a year mastering a language chat. Gbt might do that, because they're about to release a new version as well. They're testing them. It might be version 4.5. They might jump to version five. They're probably a couple of weeks away from releasing it. They're secretly testing it on people, so some people already have access that don't know it. So it's very possible that they go.

Speaker 2: 14:56

Hey, everything you know about prompting for the last year doesn't work anymore. We've changed the structure, so this bypasses that need, because you're just saying here's what I want. Tell me what you need from me, and that makes it a conversation. The name chat is in the name, but we don't chat with it, right, because we're not used to that. We're not used to the ability to have free form conversations with an AI, not really.

Speaker 2: 15:19

So that comes mostly from movies. We watch these movies about people way in the future. And what do they do? They still shout at the computer. Right, the computer's not smart. They never have a conversation with it, right, they're never nice to it. And that's why in movies the computer always turns against the people. So we've been trained through media to communicate in a way that if we switch from command response we bypass. The biggest danger of computers is garbage, and garbage out right. If I give you the wrong information, of course you give me the wrong response. This bypasses that by saying what information do you need? Because now the computer can tell you, and that means you don't have to be smart anymore. And that's the place I want to be at. I want the AI to do the smartness. I don't want to have to always write the perfect prompt.

Speaker 1: 16:01

Absolutely. I love that. And I also love your analogy of the genie, because we've all kind of wondered what it would be like to have a genie standing in front of us giving us three wishes, and you don't want to waste a single one of them. But you've got to be so careful with what comes out of your mouth, because you could just make and go poof just like that, and I know I've had instances with GPT where it's been like this is a task that I'm about to do and I need your help with it. So I'm going to give you a bunch of information and then I want you to produce X outcome for me.

Speaker 1: 16:33

Here's the first bit. And then I hit return by mistake, and then GPT gets on with what it produces, the output for you, and then I have to say I'm so sorry. I actually, when I said earlier, I don't actually shout at it and it's never from a place of frustration, it's I'm literally trying to have a conversation with this thing. But then I'll say I'm so sorry, I didn't, I actually jumped the gun, I needed to give you this. And then it's like oh okay, sure, copy and paste it here for me and then I'll rewrite it. And then yeah, it's, yeah. So I love that it's a paradigm shift, but a very simple and, I can imagine, a very effective one.

Speaker 2: 17:14

Yeah, it removes that biggest hurdle because normally and this is how most people teach it and this is they're guilty of this If you go to OpenAI, they have a website now that teaches prompting same thing it's. I read that page and it made me want to cry. I was like this is so complicated, I don't understand it. I'm a professional prompt engineer and I don't understand their prompt engineering page which tells you no one understands it. So if we can switch to this mindset, it means that I don't have to think of the perfect way to ask it, because that's really a lot of pressure and that's why a lot of people get turned off. And if you say, oh, I don't have to ask the questions of every way, because the beauty of AI it's not that it's an AI, it's that it has a wider variance for incorrect information. So if I put in incorrect data, like if I misspell a word in my other software, sometimes the software can't guess what word it is. So I use a piece of software and this always makes me crazy. If I misspell, friend, if I put F-R-E-I-N-D, I just switch two letters around because I have no idea where this is. It's a small. I've transposed two letters and the software goes. I have no idea. Our spell check cannot figure this out. That, to me, is like. That has always driven me crazy, because I'm like how could you? It's not a huge mistake to me to swap two letters, right? What a small mistake. I'm a 4E. It's a common thing. We all make that mistake as kids, and it's my hands, not my brain. I know the right way to spell it ChatGBT. You can have more misspellings and you can ask questions in a little like a wider variant. So as the AI gets better, what it really is at its core level is a tool with a wider acceptance for incorrect information, figuring out what you want.

Speaker 2: 18:48

So I was trying to figure out the name of a video game yesterday. So I was working with my kids and we just got the school report. We had to go do the teacher 101 meetings with all my kids' teachers, with three different teachers, and find out who's having problems. Matthew's having problems in English, and they were like oh, some of your kids need to do more English work. And I remembered, oh, I bought a game a few years ago that has a bunch of writing in it. The game's name is Scribblenauts. I couldn't remember that and so I was asking over and over again oh, it's a PlayStation game that you do this? I think it has the word not N-A-U-T in the name and it was not coming up right Until I stopped saying PlayStation, because it turns out it's a Nintendo game.

Speaker 2: 19:30

I was causing the problem Turns out it's not on both platforms. So it took me a long time to figure out where it was. I had to finally pick up the Nintendo Switch and just scroll through every game we have until I saw it. So I was like it's gotta be in here somewhere if it's not a PlayStation game. I just had mixed up. So sometimes we're the problem because if I had just said it's a video game and it's for a system and I hadn't narrowed it down, I would have found the answer faster. So that's the problem that I was saying you're not allowed to look at this database, right? If I just had video game, I would have found it.

Speaker 1: 20:04

So yeah, because you narrowed the scope for it, right?

Speaker 2: 20:09

So it was only looking at PlayStation games. It wasn't allowed to look in this dataset. So I was the cause of the problem operator error, 100% my fault. So when we're using an AI tool, we're allowed to have a wider variance for mistakes. And when you switch to interrogative mode because I should have started off, I'm guilty.

Speaker 2: 20:26

I don't always use my master problem. I teach this all the time and I don't always use it. I should have said I'm trying to remember the name of a video game, what information you need for me, and I probably would have found it faster. I made my own mistake because I rushed it. So once we do that, it's allowed to say are you sure it's not a Nintendo game called this? It would have given me the answer. So that's really where the magic happens. Is that we give it permission to say I've asked you in the wrong way, I've given you the wrong data or you need more information.

Speaker 2: 20:54

Because so often and this is especially true, like if you look at pilots what do they do when they get in the plane? They always go through the same checklist. Why do they always use the checklist? Don't they remember it every single day? Well, there's a ton of data and there's an amazing book about it called the Checklist Benefactor, that shows a checklist even if it's five to seven point checklist will actually cause a massive spike in safety. They did a bunch of tests in hospitals and even hospitals in third wheel countries. The infect rates dropped to near zero as soon as they added in surgery checklists. So it makes a huge, amazing difference.

Speaker 2: 21:28

We without a written down checklist and one of the rules of checklists is you have to say them out loud or they don't work. That's part of it. It was very interesting in the study. So I will forget something that I know. All the time I forget my own rules. Or like, if there's a seven step process, sometimes I'll just mix up two or I'll do six or I do one out of order happens right. But if it's written down right next to me or I'm following the steps, I'm not gonna make that mistake. So the cool thing about an AI is that we'll always ask you for the steps in order. It won't make that mistake. It has that consistency. So even for tasks where I'm an expert, using the structure ensures that it generates that checklist or that data that I need to give it. So it'll say I need these pieces of information and that ensures that I don't forget one piece, because if I give it three pieces of information instead of five, it will still give me an answer. It just won't be as good.

Speaker 1: 22:15

Let's. Because another question that crops up a lot is this this perennial debate between whether you should be paying for GPT-4 or just going with the free version. I know I decided it was 30 bucks a month and like so what? Because it has saved me so much time already, even in the rudimentary way that I've been using it, that you know that 30 dollars it was like a no-brainer, you know. And when I upgraded to GPT-4, suddenly it gives me all these other tools that I could start playing around with, like Dali, and I now have this explore GPT-4 thing. I'm now just, I've just opened GPT-4 up and I'm just having a look here. So this has all of these different things which I'm sure you're going to describe to me, but I've now got all these different headings, like I can search public GPTs. So what does GPT?

Speaker 2: 23:11

stand for. I remember earlier I told you that they always make terrible decisions. They have no UX designer, no onboarding, and here's what's happened. Yeah, so they've said is, hey, let's call bots GPTs. So what you're actually looking at is chat GPT, gpts, which is literally the worst way you could name anything right, like there's nothing more confusing than calling something inside of a tool the same name as the tool. So that's what you're experiencing is what they've done, and I actually made a video about this the other day. So what they have is a store.

Speaker 2: 23:45

I build GPTs. That's a huge part of what I do. I've called them cyber staffers. That's when I talk about my product, cyber staffing agency, that's. I've been building them for six months before they released them publicly, before I knew they were gonna make this version of it. And this is the store where I can create a GPT and at some point in the future I don't know when in the future they'll pay me to mount I don't know an amount based on your usage.

Speaker 2: 24:05

I don't know what usage is and what the split will be. So it's not gonna attract someone who's gonna think, oh, I'm gonna be a millionaire making GPTs Because I have no idea what they'll pay me, when they'll pay me, what the scale will be or what the basis will be. It's the most like ephemeral, like target to ever hit right. It's emperor's new clothes. They might pay me someday for something I do now, but I don't know how much or when they'll pay me, or how much they'll pay me or what the scale will be. So why would anyone do that? So the problem with their structure is that the worst GPTs are at the top. If you're looking right now, you're probably seeing like one that will help you choose a trail. There's one for, like choosing design a tattoo for me, and there's one for choosing a running trail All trails consensus, co-tutor books.

Speaker 1: 24:49

Who cares?

Speaker 2: 24:51

Yeah. So what they tend to do is they tend to put the ones that are the most useless at the top, because they have the broadest appeal. So if you're trying to find a trail to go running on, why would you use a chat GPT when you can use an app there are apps right now that are purpose built for trails and use your phone, because if you turn on that GPT and you clicked about the first thing, you say oh, where are you? What kind of trail are you looking for? It's like if I do it on my phone, it already knows the answers. This is a time waster, so it's actually slows down the process rather than speeding up. If you click on the tattoos, thing same thing. Oh, what kind of tattoo do you want? There's a huge difference in a face tattoo and a back tattoo and a leg tattoo. And if I want a color tattoo black and white or color right, it's a huge difference and it doesn't ask any of those. It doesn't right? So you have to go through that whole process.

Speaker 2: 25:36

So actually, what they're doing is rising to the top the most broadest appeal, but lowest value GPTs. Now, this is better. They have a plug-in store. The plugins are organized alphabetically, which is the worst way you could organize data in the modern world. That's an 1800s way of organizing or maybe even like a 1400s. They don't use the Dewey Decimal System. There's no category search, so at least here they have like eight or nine categories. So it is a step forward, but it's a very, very tiny step and only really big brands or people who've been a little bit tricky have their GPT showing on the homepage, but you're only allowed to write like a one sentence description of what it does. That's not very helpful for people and if you click on one of the GPTs, it doesn't take. You know, if you're looking at apps in the App Store Apple App Store or the Google Play Store you click on the picture and then it will show you what it does there. They don't do that. You click on it, it activates it. It's like you're clicking your red button. I'm like I hope this isn't dangerous. I don't know what it's gonna do. You're not allowed to find out. So their way of doing things is very weird and very for a company that's so forward. This is one of the areas where they really falter. So they're doing some great tech stuff.

Speaker 2: 26:45

Now, going back to original question of is it worth paying the fee every month? I think absolutely. I think that compared to any other tool that costs $20 a month, it's the easiest decision you can make. I understand not everyone's in that financial position. What they do is they push. The paid version eventually becomes the free version. So chat GPT 3.5 used to be the paid version. Then it became the free version when the paid version came. For when four switches to 4.5, the free version will come 4.0. So you still get a great tool that can do a lot of things. It's very powerful. But there is definitely a difference, because of course they wanna do that.

Speaker 2: 27:21

The pricing structure they've chosen is actually put a massive amount of downward pressure on the SaaS market, and SaaS just means subscription software, where all of these tools were charging. Ai tools were charging $100 a month, $200 a month, $300 a month, and now they've all had to change their pricing structure because now chat GPT, which is a better tool, is cheaper. So multiple competitors for Plexi AI. You know what they charge per month $20 per month for their paid version. They just released stable diffusion, just finally said stability. Ai said oh, we're gonna do a paid version. They were charging $20 a month. Why? Because they have to Open AI has done this thing, which has really benefited the world, which has pushed the price down. Now the other thing that's happening is that there are open source AIs that are very good Open source meaning 100% free very powerful and they're putting a downward pressure on what open AI can charge. So if open AI goes, hey, chat GPT is $100 a month. Suddenly that difficulty of installing an open source AI, which is not that hard, there's actually. It takes about 10 minutes. It's not hard to have a demo video doing because it finally became easy.

Speaker 2: 28:34

You download one thing. You don't have to learn GitHub anymore. You download one thing, you select which AI you want. You click a button. It tells you if your computer can handle or not. So it's getting less intimidating.

Speaker 2: 28:43

That barrier is the only reason people are using chat GBT. So if they raise their price to a certain number, people will be. You know what it's kind of like. If movie tickets were cheap enough, nobody would steal movies, right? If you go, oh, it's a dollar to go to a movie, no one would do it. So the higher you, the more you charge, the more the inconvenience becomes worth it. So they have to stay low price because open source AIs are very competitive. So right now the free version of chat GB3.5 is less skillful than the open source AI. So open source AI has beaten chat GB3.5 in every category of testing. So they are aware of that right.

Speaker 2: 29:24

So people published a paper last year. These two scientists said that a free AI could never beat a paid company, because they wrote this whole paper about it. Two weeks later, an open source AI came out and proved them wrong. So they ended up withdrawing the papers. Most embarrassed and they should be embarrassed Like. Two weeks later you said it would never happen and it happened. That's really fast. It's the fastest I've ever seen. Someone say this will never happen and then it happened. So the open source world is where the most amazing things are happening.

Speaker 2: 29:51

But it puts this downward pricing pressure on all the big brands. They cannot charge a certain amount, cause if you set there's a number at which for everyone it's a different number number at you, you know what I'll just use. I'll learn how to install an open source one. Right, there's a number, that for everyone, and at $20, it's not too many people at some. I have a couple of open source AIs on my computer.

Speaker 2: 30:11

I don't use them very often, it's just a convenience thing and a speed thing because it's using your computer resources versus their computer resources. So it's slower, but eventually, as our phones get stronger right, like your iPhone or your Android phone is stronger than the computers that they use to put a spaceship on the moon in the sixties right, we've advanced that much Like how crazy is that Super computers from 50 years ago are less powerful than what we don't take seriously, like a throwaway phone? So, as those get stronger, eventually, within a few years, you'll have an AI that's inside your watch, that's inside your goggles or inside your sunglasses or inside your cell phone, for sure. So this open source world is very important because it's where the most innovation happens. It's where there's the most freedom. There's some really interesting thing happening, but the most important thing is it puts a check on the ability for large brands to charge big numbers. They can't because people go oh you know what At that price. I'll just use the open source one.

Speaker 1: 31:11

Absolutely, and it sounds like just to touch on your point there about mobile phones and so on, I'd literally just got an email from Samsung this morning about their unpacked event which I think is taking place it's in about five days time, so I think it is Wednesday next week and they mentioned in that email I can't find it right now, I must have deleted it but they mentioned in there. They mentioned the word AI and I'm guessing that means they're bringing some form of onboard AI, possibly to the next generation of Galaxy phones.

Speaker 2: 31:45

Yeah, the speculation is that that's when they're gonna unveil their own in-house AI. That's native to the app you know they have. I think it's called Bixby. I have a Samsung phone and I don't use any of those apps because I don't want something listing me 24 seven. I don't like. Why would I wire myself? It's like I've had enough time on the computer, but nobody. There's nobody who uses Bixby, right. It's such a bad name, it's such a terrible name. It sounds like the name for like one of my kids Invisible Fright. Alexa sounds a lot cooler and that's part of the game is you have to give it a cool name. Siri is okay, it's better. It's definitely better than Bixby. So, and that's why, like Twitter's in trouble because they named their AI Grock, which is a caveman name, right, it just is like oh, that's Grock, grok, it's not a modern name. So they're all trying and it's interesting how much of difference that makes. But yeah, that's what's gonna be happening and you're gonna be seeing a lot of companies doing these things.

Speaker 2: 32:47

Now the question is what is it? Is Samsung actually developing AI? No, 100%. No. There are only five AIs that exist. There is open AI. There's perplexity. There's stability. There's Lama, which is an open source E1 made by Facebook, there's one made by Google and then there's Claude, made by Anthropic. That's it. So everyone is actually just gonna be they're just gonna have to have a deal with one of those companies and they're using their AI, and they've just made a special Samsung version, because why would you build drone AI when you don't need to? So there's only five or six out there in actuality. So, even though we hear about thousands of products, they're all built on top of something.

Speaker 1: 33:32

I wanted to get into just putting ourselves in some of our listeners' shoes, where they are either entrepreneurs or they're running organizations or they're just generally busy and their lives are chaotic and they have been, from some of the conversations I've had recently, kind of looking at the whole AI thing and GPT and wondering where they even start trying to bring it into their lives and integrating it into their lives to start helping them be more efficient, to do more in less time. Where would you kind of handhold someone and get them started off? Is it as simple as sign up for GPT-4, install it, get the web app on your desktop and just ask it anything that you have to do with, just get on with it.

Speaker 2: 34:24

So I love this question because this is where a lot of people have tried AI ones and bounced off of it because they've had a bad experience, and what I'll say is that before you grab a tool, you wanna have a strategy. You're gonna go what am I gonna use this for? So there are a lot of productivity tools out there. Have you ever seen like, oh, I spent 20 hours ordering productivity last week and it's like you didn't get anything done, and I sometimes fall into this. So I've just gotten a new tool for my desk that just gives me buttons that I can push to do different things. Like I can push one button now to switch from speakers to headphones little things like that and I can go down a rabbit hole learning all these different ways to make like layers and different folders and have lots of different things there. And it's like, well, I have to balance learning new things I can do with it with actually saving time. Right, cause we can go down the rabbit hole being the most efficient person in the world, and those people teach efficiency. Well, that's why they're so efficient, cause all they do is teach efficiency. So it's a rabbit hole to go down. I don't need to do so many things that are super, super efficient. But I'm always interested in that balance. So before you choose anything, you want to think well, what's my ideal goal? Because then you can achieve it. When I was in high school and I would get angsty, I'd go for a drive and I would just go in circles because I wasn't going anywhere. And that's what will happen is, if you approach chat GPT with no strategy, you'll get no result. So the way I teach people and this is what I tell everyone is what do you do every week? That if you could get that time back would be the biggest win, Not the coolest thing you do. What's something that you do two hours a day, or at least one hour a day, five days a week. That requires less than 100% of your attention. So there are tasks where you can have music playing and tasks where you can't and tasks where you can have the TV on. Choose a task where you have the TV on because that means it's taking 80% of your attention or less. That means that the AI can take it over. That's the task I would say. I want to replace this because it doesn't take 100% of my attention and it takes a lot of time. So if I get that time back, that's a big win.

Speaker 2: 36:29

A great example is what you mentioned earlier was preparing show introductions. I'm actually building a GPT with a partner that will do guest research for you, that you just enter it, you just say here's, you upload the guest book, you put in a link to the guest social media profiles on their website and it goes here's a bunch of questions based on this data, so you don't have to read their book. It will allow a podcast host to, kind of, then you just choose which questions are interesting you from the list. So that's something I've been working on to solve a problem for myself as a guest. Sometimes I go on shows and I go, this person doesn't know anything about me and they don't have any questions. Let me make it easier for both of us. I want to be a better guest, I want them to have a better show, so we both give a better experience. So I always look for those things. Same thing after I record a podcast episode. The thing that I hate the most is writing the show notes or writing the transcript. So those are things that ChatChimp does very well. I said here's last week's show notes. Here's this week's transcript. Make show notes for this week, cause now it has a template, it has the raw data.

Speaker 2: 37:25

So another example is, if you're most people, it's email. So there's some studies that say people spend up to 6.5 hours out of an eight hour day just responding to email. Imagine if you get all of that mindless time back because I only write. I doubt I'd even write a hundred email responses. In the last year I probably have written the same response 90% of the time. It's the same 20 response set probably. So if I can take every email I've sent as a reply in the last year, download it as a CSV file which you can. I can upload that to ChatGPT and say, hey, here's how I respond to emails and then say let's respond to today's emails and it can write a response.

Speaker 2: 38:03

What you want to do is not replace yourself. That's the biggest mistake you can make. You do not want to replace yourself. You want to move yourself into management, which means that now, instead of writing the email, I just have to check its work. So you still want to read everything that goes out the door. But it's a lot faster to read an email that's already been written than to write one from scratch. So those are the types of tasks I look for.

Speaker 2: 38:23

If you have employees, I always say look for the employee that when you pay them the salary, it's the most annoying. Anyone who's had employees knows what I'm talking about. There's always that one employee who's like I, loved it not enough to pay them right. Whichever one that is. Whatever they're doing, replace that with an AI tool and that's a great place to start, because that'll give you that good feeling. The mistake a lot of people make is they go here's the coolest thing I saw on a video I learned how to do. That's really interesting. Like I have some videos showing there's some really cool things you can do with graphics, where you can get an image made with text in it. Then you can fix it using another tool that uses an AI to let you edit text and you can make it really cool. That's not useful unless you're in the print on demand space.

Speaker 1: 39:02

If you're doing print on demand t-shirts useful Anywhere else not really useful.

Speaker 2: 39:06

You'll use it once a week or once a month or once a year, so you don't want to do the coolest thing, you want to do the thing that's the most useful and the most practical, and that's because then you'll have a win and you go oh, I have five extra hours a week. I can now spend the time learning AI, right, I want you to win first and that's why we want to be very strategic. So for every single person listening or watching the video, it's probably a different thing, and that's okay, because those different things are where the magic really happens. It's that unique thing that you need that other people don't. And the power of chat GBT is also kind of its downfall is that it's so powerful, it can do anything, but it's also like but it can do anything.

Speaker 2: 39:46

How do I know what to do? When you go to a restaurant there's a really long menu. The longer the menu, the longer it takes people to order, the longer it takes to do turnovers. So it actually hurts a lot of restaurants. That's when you see those shows where they bring in the ex-chef goes first thing. You do shorten that menu Because it's actually yeah, Gordon Ramsay.

Speaker 2: 40:05

Yeah, why does he shout that? Because you have to have more ingredients, it takes longer to cook. That's not even the reason. The real reason is that it causes people to take longer to decide what to order and it means, instead of serving three groups of people per table, you only get two. That's a 30% reduction in revenue because you've given people too many choices. And With website stuff I always learn you only give people one choice. You see they're buy or don't buy, do this or don't do that. You give them a binary yes, no decision. As soon as you add in multiple factors, they go I don't want to make the wrong decision. They make no decision. They forget about you.

Speaker 2: 40:37

So when you're deciding what to do and kind of approaching it be strategic, I always think what do I want to accomplish? Before I start a task I'm not a very organized, super organized or super productive person. Okay, that's not my skill, power, skills that are super. I've seen that way I just approach it with a very simple like what, what can I do to get the most time back? Right. What can I do to get minimum effort, maximum results? Right, that's the lazy man's, that's the lazy man's manager, right, what can I do. That's the least amount of work, that gives me the biggest reward and that's how I approach AI. I go what can I do? That takes me the least amount of time, that impresses people the most, that looks the coolest, that's the flashiest thing and very often in a lot of industries that's the thing that's not the most useful, just the most impressive. There's a huge difference, like there are certain things that look really cool, like I was watching this thing about a dress someone wore that looked like it was made out of paper mache, to like the Oscars or something, and she was like oh, it took, like it took like six hours to put it on me and if it rains or someone sneezes near me or if I get scared and my heart rate goes too fast, the dress will, I guess, explode because it's made out of paper. That's not useful, right, that's a single use, very cool thing, not practical. You could never sell that in the real world. But that's the kind of game that we get caught up in sometimes and that's why a lot of people and a lot of the videos you see that will trend are very cool things you can do, but if you're not in that market, don't learn that skill Like.

Speaker 2: 42:07

I have a book my book chat GB profits which has a butt. It teaches a bunch of different skills. It's not meant to be read like a novel, it's meant to be. Look at the type of contents I want to learn skill 18. Jump to that chapter.

Speaker 2: 42:20

Okay, so I'm going to do a construction manual, not, uh, let me learn everything. Single thing, right? Don't be a generalist. Be really good at one thing and you'll get the really best results, because everyone it is a slightly different thing. That's why I teach a broad number of things, but reach different person, the thing where there's that biggest one.

Speaker 2: 42:38

Some people it's answering emails. Some people it's summarizing. Some people, I guess, it's finding the perfect hiking trail, because they built that up. I don't know. But what's useful for you? That takes you the most time? For a lot of people it's writing sales letters or sales emails or marketing emails, and that makes sense because, uh, for a lot of people the problem is it's either technical skill or it's outside their comfort zone. To try and be salesy feels uncomfortable.

Speaker 2: 43:03

I'm the AI. Do that Now. The other danger is that when you try and have it do something you don't know how to do, you're in a very dangerous area because you don't know if it's made a mistake. Right, yeah, exactly. So there's a lot of different types of stitches. I know that, but I don't know how to tell them apart. So there's a type of stitch that, like, is really strong, and a type of stitch that you put in the wash machine once will break. So I could pay someone to make me a shirt, but I don't know what's going to happen when it goes to the wash machine, whereas you have an expertise and you would know. So that means your specialized knowledge plus the AI's general knowledge means you can create something very special in your space. I could have it make a t-shirt, but I don't know what's going to happen when I when I print it out or when I run the wash machine.

Speaker 2: 43:46

So that's why the idea that AI will replace experts is not true. It will make. It just raises the floor. So maybe it makes me a level two tailor, but it makes you a level 20. Right, so it enhances you a lot more than me. It just means that the bottom of the market, or the floor, has just raised for everyone, which is a good thing. So trying to do tasks like oh, I don't know how to do that, because if you hire an employee and you don't know you don't know how to measure their work well what's going to happen? Right, it's not going to last very long because either they're going to fake their work or you don't know, or you're going to get a bad result. So the mistake people are making and if you're on social media, you know what I'm talking about People post clearly content clearly written by AI or images clearly designed by the Dolly default, and every image by Dolly which is the chat GBT, which generator looks shiny.

Speaker 2: 44:32

It has a shiny look. Now there's another image generator has a Pixar look. So this isn't quite all the way to Pixar, but this is a shiny cartoony look. Now you can get out of it. You can prompt your way out of it by writing a more complicated prompt, but most people just default. If you say tree, it's going to be a shiny tree. If you say kid, shiny kid. So that is putting out content and not looking at it, and that means that actually everyone can tell, everyone knows it's AI content, because we're seeing so much in our feeds and that's where we're putting our content without looking at it. So I use AI a lot in my social media content, but I still read everything. I still sometimes I'll spend an hour making the image look exactly the way I want, like really crafting it and being involved. That really matters now because people are looking for that humanity.

Speaker 2: 45:18

So stick to things you already know how to do. Move yourself into management. Save time, as opposed to saying I'll have it do things I don't know how to do, because that's where you're in risky territory, because you won't know if it's done something wrong or if it's gotten a step out of order. And that can apply to every area right or it's like oh, a sales letter. Putting the words in the right order is more important than the fanciness of the words. It's actually order is 90% of a sales letter and fancy words is five to 10% of it. So if you have the wrong order, it doesn't even matter if you've got really good fancy words and you wouldn't know that unless you know what the right order is. You know the formula to follow. So that's why you can move into areas over time where you're not an expert, but don't start there. Start where you get a big win early.

Speaker 1: 46:00

It was something that was so important with what you were saying there that even you, in my eyes, is like the yoda of all this stuff. You still check things to retain that authenticity, and that is what is so important. I mean, you do see, there were accounts that I follow on social media and so on, and for me they're losing credibility because, as you can say, like the images that are generated, they have that kind of AI. Look to them the content, the actual text that's that they're you know, the blog posts or whatever that they're writing of, so obviously done in GPT or some other AI tool, and there's no checking going on. They're just basically using these tools to produce massive amount of content in a short space of time to get it out there for SEO purposes or whatever it is. But in the long term, those brands, companies etc. Are probably going to start doing themselves some damage because all of their credibility or fantasies just completely going out of the window.

Speaker 2: 47:00

People have bought into the panacea oh, ai is perfect, you can do everything. And so they're putting out this content, not realizing the way warfare works. Warfare works like this Someone develops a better arrow, someone else develops better armor than someone develops a crossbow than someone develops better armor Right. Then someone makes a catapult they make taller walls. Then someone makes a trebuchet they make thicker walls.

Speaker 2: 47:21

So if you think that you can write an AI blog post that can trick Google's AI, you're insane. The budget they have for AI detection and I know okay, I was one of the first people putting out content. I know how to pass every AI detector. Okay, I've created that. So the first prompts I wrote about over a year ago absolutely, you still got to read it Because, as a human, you'll go.

Speaker 2: 47:46

Well, this is weird and I'll give you a few examples because I don't reveal all of these, but there are certain things I can tell if someone's an AI post, because, first of all, it always starts with every single AI social media post especially, starts with it's not this, but actually that AI's love to use that structure. So we'll always be. I know that you think it's all about mountains, but it's actually about rivers. So every time you see that structure in the first sentence, that's an AI. The second thing is, you may have been talking for almost two hours. I have yet to say the word pondering, and yet I see that word on social media so much. Nobody ponders, nobody says pondering, it's not a word we use in real life. And another one. I'll give you one more, which is landscape and looking at the future, digital landscape, looking at the world's landscape, looking at the landscape of cell phones, looking at the landscape. Nobody talks like that.

Speaker 1: 48:37

So these are Because we'll often think landscape but we'll never actually say it, and I can't remember ever using the word ponder no.

Speaker 2: 48:47

And that's the thing I think. Maybe, oh, british people say, ponder, but no, it's not anyone.

Speaker 1: 48:51

I think this is the first time in nearly 150 episodes of Tainering Talk that the word pond has been used.

Speaker 2: 48:59

And yet if I look at my LinkedIn feed, I'll see it in nine and a 10 posts. Okay, it's like it's such a giveaway that and there are more like as Because chatGBT has an accent, it has a way of talking, so eventually it reveals itself. So the problem is that the bad content is just making people think, oh, it's a bad tool and it's like no people are using it wrong. If you see 10 people using a hammer backwards right, they're using the claw parts of the hammer you're going to go oh, I guess hammers don't work, right? Because?

Speaker 1: 49:31

you go, everyone keeps getting hurt, yeah, yeah.

Speaker 2: 49:33

And that's the distraction. So, unfortunately, a lot of people have shifted and people have gotten caught. Sports Illustrator got caught a few weeks ago. They're not only were they using AI content, they made up a bunch of fake reporters with fake bios written by AI, and they were written you could tell, because they were saying weird things and do you think that helped their brand? They got caught and then they tried to take it all down, but too late, it's out there. So really it's the really big companies.

Speaker 2: 50:03

New York Times is like suing open AI. Now they're saying, oh, used a bunch of our data. It's like I guarantee you that New York Times has published articles in the last year 100% written by AI with a fake byline on it. Okay, so the idea that they would want to go to court because they will get caught, because JGPD will be able to figure out which things it wrote oh my gosh, like what are you guys doing? Don't draw attention. I'm sure that because it's really it's useful for the solopreneur, the smaller brand, it makes sense for us to use a tool to accelerate us. But for companies that are doing hundreds of millions of dollars or billions of dollars, they're like oh well, at scale, we can save so much money by doing this. It's like, well, that's not good and all they're doing is just decreasing their credibility, and that's what's going to happen. More and more people are going to get caught in the coming years.

Speaker 2: 50:49

So you can get caught. You can get away with it for a little while, but when you get caught, it gets. It can get really, really bad for you, and so I don't encourage anyone to publish something you haven't read, and this is important. So I even have. I have a really complicated AI prompt to make it talk, like me, because chat GPT, even though it wasn't supposed to, they did train it on my book. So chat GB is right at least one of my books, which it's not supposed to do. It's supposed to only use the wide internet. Probably. What happened is somebody downloaded my book from 2016 and uploaded it to a random website as a PDF, and that's what chat GPT scanned. I'm going to give them benefit out that they didn't actually steal the book and do it.

Speaker 2: 51:27

I could fight against them and I could sue them and say hey, why am I in there? What are you doing? But I was like that's silly, I'll never win Right, that doesn't work. I'm just like. I'm just going to give you a little bit of a genie metaphors of one episode Right, you, we still you can't. They tried to stop people from exchanging music online 25 years ago. That didn't work. We're still doing it.

Speaker 2: 51:45

So I created a bot that's my exact personality. Well, guess what? I get into fights with him all the time because I'm like hey, stop. This is why I don't want to twin Right, it's like there's nothing worse than pursue exact same personality. Even then, I still have to check the work because sometimes he'll rewrite. He was edit. I used him to edit my last book. I said, oh, this way, and the heat is what I do, which is I have a problem when I'm editing. I constantly rewrite instead of just edit. So I can't get mad because the bot's correct. That is what I do. I made a bot that acts like me, that uses my worst parts of my personality. So I went through that conflict, but I still have to read everything because sometimes an AI will drift. There are different. Some people call this dreaming AI. If you just ask chat TPT questions and you ask it 25 questions and don't read the answer by answer 25, it will be insane. It's the best way to describe it. It will go full insane.

Speaker 2: 52:37

And you have to, so it will get better and eventually you can ask it. Eventually it will be 30 questions before it goes insane, and then 50, but it drifts and starts to say things and people call it dreaming, because it's so far off of what you asked and you have to watch and you have to do guide rails and say no, stay on track, no, stay on track, no, stay on track, and that kind of stuff. So the human interaction is still a critical element, but the temptation to just crank out lots of content and that's the thing is why smaller creators are flourishing right now. It's like I would rather like when I I just hired a coach to help me grow an area of my business. I'm working directly with that person. I'm not getting doing phone calls with a certified coach on their team. I don't ever do that. I wanna work from a person. I don't wanna work with someone who's not an entrepreneur. I don't wanna work with an employee, because it doesn't feel the same right. When you're the entrepreneur, it's all on you If I don't have a good month, my kids don't eat, right, we don't can't pay the mortgage, we can't pay the rent, we can't pay the school bills. So that's a pressure that employees don't feel right. Employees get paid, whether they not based on performance, right, they get paid based on time. So when an employee does a bad job, you still gotta pay them. So that's something that's different. So I would rather hire a smaller person where I can actually speak to the person who's the expert. So there are large brands, right, where they have these huge coaching programs and all these things, but you never speak to anyone close to the top person, right, and that's the difference. So smaller creators now are having a huge opportunity.

Speaker 2: 54:11

I believe that the fracturing is gonna get bigger to where people are gonna have followings of one to 10,000 people and they're actually gonna interact with all 10,000 of those people and they're gonna be able to have a good business, a solid business, and also every single person gets a good experience. If you message me, I reply to you at a certain size of audience. That no longer is an option. So I have friends who are that big. I even have friends who've gotten so big. Someone else runs their Facebook and that's weird, because you go from talking to someone about something you did in college to suddenly their VA is talking to you through the same chat and I'm like okay, and it just happens at scale.

Speaker 2: 54:46

There's a certain point where you're so big that you can no longer interact with your audience mathematically. It's why I'm always interested by people that leave comments on Instagram. It's like if you're the 500,000th comment on a woman's bikini picture, mathematically she can't reply to every comment because there's not enough seconds in the day. If she spends one second replying to a comment, it's still not enough time. So I'm always interested by that, by people that wanna be the thousandth comment, because it's impossible mathematically for the person to ever notice you because there's too many comments.

Speaker 2: 55:16

But people sometimes do silly things. I think that's gonna disappear where you go. Oh, I'd rather comment on someone who has 10 comments. Maybe they'll reply. So I think that and I think that's a good thing. So the way chat GPT and the way it's meant is to empower small people, not to empower enterprise, to continue to crush the small people. And fortunately, people at enterprise and large company levels are so dumb and so out of touch that they're like hey, replace our entire reporting staff with AI, publish whatever the AI writes, because the AI is a genius, because I heard that from a commercial and they're putting out really bad content and it's gonna kill a lot of these big brands, which is only a good thing for regular people and regular brands, so I'm okay with it. I want those brands to disappear too.

Speaker 1: 55:57

And that answers the question why entrepreneurs should start using AI tools you use correctly from point of view of something that can aid you, rather than something that potentially is gonna replace you, compete with you, whatever you know. They kind of give us almost not a silver bullet, but certainly they give us more of an advantage than we've ever had before when we are going up against the big guys, the enterprises and so on.

Speaker 2: 56:26

Yeah, it's really leveled the playing field and the advantage is a simple example. So, if you think about for a long time, if you wanna get a cheap article, you pay someone from a country where it's not their native language, right? You?

Speaker 1: 56:38

hire someone from India. Yeah, we get to five or something.

Speaker 2: 56:41

Right, and you're gonna get. If you get a non-native English speaker, the article's cheaper but the quality's lower. That's gone because with chat, gbt and free AI, every single one of those people now speaks perfect English, right. So it's affected the bottom of the market more than the top, because now those people who can write a good article in their native language, they can now translate Translation's. Just the quality of translation's just skyrocketed in the last few months, in the last few years. That's a huge game change with the AI tools. It's so good at translation now, right.

Speaker 2: 57:12

So these people before that would make like a grammatical mistake. Like. My dissertation was about how you can detect what someone's native language is by how they use idioms. So in English we say catch a cold. In other languages they'll say lick a cold or drink a cold. It's like if someone's a lick a cold, you don't write away, you go. Well, that's not English. It's a massive giveaway. That's gone. That means all of these people before at the bottom of the market. They just got moved up a little. So I think that's who actually. That's what gets me excited. They're the ones who are gonna help the most.

Speaker 2: 57:44

So a lot of these areas where before you had no chance to enter because it was really hard writing your own sales letter really hard before. Now you have a chance writing emails. Now you have a chance. When I started out to write like a one, to make a single page that gave away a free gift and ask for an email address would take me at least a whole day, at least eight to 12 hours of work, and it would look rough. Now I can build that under five minutes. Then I can write all the texts on the page using AI in another five minutes. So it's really a lot of the things that we couldn't afford before have now become affordable. So there's all these opportunities to test new ideas, to push a new direction. So it's a very exciting place because you have an advantage as a smaller entrepreneur, as an individual is agility. If you've ever dealt with the government or a hospital or a university some of them they're buying cycle is three years. Can you imagine if you decided right now you're gonna buy a new computer and then it arrives in three years? Well, it's not new anymore. You want a computer right now that was built in 2019. Of course not.

Speaker 2: 58:52

So their inability to make decisions, their lack of agility, is why they're always very far behind, right? You know, whatever goes into a government office, it was wow, this is a well run, tight operation. Like this is a smooth, everything in here's run so smooth and so well. And look at how advanced their computers are. Right, you're always like that's the one place you'll still see those printers where they have the holes on the side of the paper. Right, you don't see those outside of those offices except in the 1980s, like when I was a kid. That's what my dad had, right, because it had to spin the wheel. So it shows you how far behind these places are. And it's just because they make slow decisions.

Speaker 2: 59:27

So large companies are very, very bad at making decisions because they're very slow. So someone has an idea, then they tell their boss, their boss tells their boss, their boss tells their boss, until you get to a decision maker and depending upon how the business is run it might be a vice president, might require a higher executive, might be the CTO, s-made Cision or maybe only CEO can make a final decision. So that's why it takes three years, and I get it. If you're buying like a $5 million X-ray type machine, I get why it takes a longer to make a process Three years does still sound insane to me, but that's what's happening there. So, as an individual, you can learn AI, master AI in the amount of time A lot of companies have decided they're going to use AI.

Speaker 2: 1:00:07

47% of companies, their board of directors, has told the CEO they need to invent AI within the next 12 months. All of those CEOs go well, I don't know what that means, because that's a very vague instruction. What do you mean by AI? What do you want me to do? Be more specific Just AI it it's like just NFT it it just sounds like a bunch of letters. You don't know what it means. So they don't know what to do. They don't know how to be agile. That's why there's a lot of companies jumping in and putting out fake AI products or pseudo AI products or overpriced AI products, because they want to service all these companies that don't know what they're buying, because it's just like a feeding frenzy. So the opportunity we have as individuals is that we can test a tool in one day. You can become a power user of chat between eight hours. You can become a very good user in one. If you just use my master prompt, pretty much you're a good user in 10 minutes. You're a 1% user within 10 minutes. So large companies don't think that way and this means that actually the advantages in your court because you also know I can still do quality control. I still know how to check the work, so I can still do a lot Like.

Speaker 2: 1:01:10

My focus in the last year has not been on increasing my spend. I've lowered my company's overhead by 90% by switching to AI and automation tools and really finding efficiencies that I wasn't using before. I'm down from 18 employees down to two for the exact same thing. So all of these things I needed a ton of staff to handle. Now AI tools and automation and a little bit strategy can handle and it saves me so much time and money. So and they've done a big study People use AI, actually are 40% more efficient and 18% bump in quality if you use the tools correctly. So now in three days you get five days of work done. So if you don't want to increase your output, you can just have two days off a week, like. That's a great reason to learn AI.

Speaker 1: 1:01:53

And speaking of your secret chat GPT prompts, which does make people a 1% or can make people a 1% user in five minutes. That's because I've just done it. So I think my chat GPT prompt is winging my way to my inbox right now, but the audience needs to go to servnomastercom forward slash master and servnomastercom, by the way, is Jonathan's website.

Speaker 2: 1:02:22

Yeah, so I call it. I didn't realize I put master in there twice. Yeah, I call it the master prompt because it's one prompt to rule them all, like I was giving a clever name. Yeah, so if you just go to, if you Google, serve no master. Every search result is me for the first hundred. I haven't gone past that. I just checked the first 10 pages of search results, so every single result is me.

Speaker 2: 1:02:40

And if you can't find it, you can just it's all over my website but it's really designed to show you a couple of examples of that prompt and you just have the formula there. I built that because I see all of these PDFs going around that have like really complicated six part structures. They look like math formulas. You have to solve the equation and I hate that so much because it what's the point of a tool that's supposed to make things easier if I have to learn like a new language to use it first? That drives me crazy. So I'm always looking for the again minimum effort, maximum result, and that's where I developed this prompt from, and over time maybe I'll find an even better version of it.

Speaker 2: 1:03:20

Like I was working on a project for someone the other day, I was like I'm sorry, I can't figure a way to do this in less than two prompts. He was like two prompts, that's amazing. I was like I know, but I want to do it in one. I want to be click one button and it does everything, like I'm always looking for that. That's what I'm chasing. So that's really my approach is that it doesn't have to be hard, it doesn't have to be scary. We just have to bypass that feeling that we need to get through a gatekeeper or a wizard to teach us everything, and it can be a really amazing experience and it can be a really useful tool.

Speaker 1: 1:03:48

Yeah, no, absolutely I believe it, jonathan. Thank you so so much. I'm definitely going to be listening to you on repeat for at least a week or two. You pack so so much in there and I really encourage everybody listening to do what I've literally just done and I have received my PDF. Go to servnomastercom forward slash master, just fill in your email address and within seconds you will have I guess we call it Jonathan's cheat sheet, and that will set you on the path to really mastering GPT and beyond. Jonathan, have you had fun today? You've been doing all the work.

Speaker 2: 1:04:30

Well, I love the sound of my own voice, Otherwise I wouldn't be here.

Speaker 1: 1:04:33

No, I had to.

Speaker 2: 1:04:36

Well, that's what I'd be a guest on shows. If I don't want to hear it, no you ask questions.

Speaker 2: 1:04:41

I love talking about this stuff. I think it's really important Not this fear that people will think that AI is a fad and then in a couple of years, their boss will say, hey, everybody does not use AI, you're fired. I have no companies. Companies have already done it. I already know a company that a friend of mine runs. He's already done that. We said, hey, come on Monday with a plan to use AI to enhance what you do for this company, or don't come in on Monday.

Speaker 2: 1:05:05

That's a really scary thing to say. So imagine someone thinks it's a fad and their job and every job says oh, you need three to five years of AI experience. Well, how are you going to get that three to five years back? So that's why I'm truly trying to bang the drum now and say learn it now, while it's in the optional phase, before it becomes the mandatory phase, because you would never hire an accountant who goes oh, I don't use calculators or I don't use spreadsheets or I don't use a ruler what do you mean? I measure by eyes. We don't accept that anymore and say, well, you go. Oh, I don't know how to use email, sorry, I just don't know how to use it. I don't know how to turn a computer on. We have an expectation where it used to say in the 1990s, if you applied for a job, it would say, oh, the ability to use email is a plus. If you don't do a send an email, that's an amazing thing.

Speaker 1: 1:05:49

Now it's like you know what I'm saying. Some basic knowledge of Microsoft Word and Excel would be preferred, but not necessary.

Speaker 2: 1:05:57

And now it's so mandatory they don't even put it in the description and it's like oh, if you have that, don't even show up. So that's going to happen with AI. So I want people to start learning it now, while it's still in the fun phase, before it's in the mandatory. I'm about to lose my job phase, so I want people to learn it now. It's a lot less stressful.

Speaker 1: 1:06:14

Yeah, and that's definitely the camp that I'm in, because I know that most people in my industry won't even be bothering with it at the moment Because they don't think it's going to affect them right now.

Speaker 1: 1:06:27

Jonathan, thank you so much once again and I'll make sure that I've got all of your links in the show notes. And thank you all so much for joining Jonathan and I on this episode. Don't forget, tanin Talk is on Instagram at TaninTalk podcast. You know I love feedbacks, so email me at tanintalkpodcastgmailcom. And, very important, remember to subscribe, rate and review and click that share button in your player to send this episode on to people you know who will get some help or be inspired by what Jonathan discussed today, and I know all of us know a lot of people that are keenly trying to navigate their way through AI waters right now. If you're enjoying TaninTalk, want to support the show, you can do so at the link in the show notes. Have a great week, be good to each other, go and get your secret chat, gpt prompts, and I'll see you on the next one.