From Fired in a Snowstorm to AI Entrepreneur: Jonathan Green | S3 E36

Jonathan: 1:09

It's like when you hire an employee if you can give them an example of a good task and then the raw materials for that task, it's much easier for them. Most of the time we hire someone to just do social media and they go. Why don't that mean you go, just run into my social media and make it grow? So because we can take a task, we're doing a lot. That's repetitive. Those are the first things I like to push down. And it's real.

Joeri: 1:29

Hello everyone and welcome to the Web3 CMO Stories Podcast. My name is Joeri Billast and I'm your podcast host, and today I'm really honored to be joined by Jonathan Green. Hi Jonathan, how are you?

Jonathan: 1:44

I'm doing good. Thank you so much for having me. I'm glad to spend some time together today.

Joeri: 1:48

Yeah, me too. I always like to get to know people on my podcast and then for people that don't know, you, Jonathan, guys, Jonathan Green is the best- selling author of 300- plus books. He's a celebrity ghost writer and a high- ticket affiliate marketer who now lives on a tropical island in the South Pacific. He has turned being fired during a blizzard, into a thriving online business. Yes, so those things already are. If I read this, it seems fascinating, like more than 300 books, best selling author, your story about the blizzard turning into, then being fired and turning into a thriving online business. So can you share a bit more, Jonathan about that story, so that it seems kind of dramatic? But how did that go? How did that work out? And, yeah, I would have other lessons for you.

Jonathan: 2:47

Yeah. So It was Monday morning of my new job. I'd been there just under two weeks, I just signed a lease on a new apartment for six months, and bought a new car the day before I was driving the new car and it was a snowstorm, and normally right, it's a blizzard. You're like, oh, I can take the day off work, it's too dangerous to drive in. But I go no, I wanna show I'm a good worker.

Jonathan: 3:06

And I come in and it's a meeting with my boss and my boss's boss and they basically were like we wanna get rid of you, it's not working out. And I'm driving home just thinking please don't crash. Like, don't make it worse, it's already bad. And the lesson I learned was that when you work for someone else, they have so much power over you that they can devastate your life because you don't just lose your job, you lose your ability to pay for the doctor, to pay for your rent, to pay for your car, to pay for your kids to go to school, to pay for food. And the lesson I learned, the change was I never wanted someone to have that much power over me, ever again.

Joeri: 3:47

Yeah, that's one of the reasons also, you know why I've been an entrepreneur already for so long? Because I indeed don't want that. I don't want to be my own boss, but of course, there are risks involved and people of course need to be prepared to take those risks. Now, today, we have something like AI and ChatGPT is there, which can help a lot of entrepreneurs, a lot of businesses, to grow, and you have a new book called ChatGPT Profits and you will discuss everything about leveraging chatGPT for businesses. So, yes, can you give maybe one or two examples of how ChatGPT can be a game changer for online businesses?

Jonathan: 4:31

So when I started out way back in 2010 and even my first online stuff in 2007, every task took a really long time. So for me to make a single webpage that just had a picture and like three lines of text and a button for people to give me their email address would take me an entire day to put together and it would look not very good. I had to make it an HTML. I'd use Dreamweaver, which doesn't even seem to exist anymore, so I had to learn. Really, my level of coding is copy and paste, so I don't really know copy and paste. I probably know 10 things like how to center something, how to bold something, but I had to learn these things. It was the only way to do it and over the last 10 years now there's so many tools that I can build the same type of page now in five minutes or 10 minutes, so the amount of time it takes to complete a task has really gotten quicker. You don't need to be as technical, so the barrier to entry for people into online business has really gone down and with AI, a lot of those tasks that were content creation whether it's writing social media posts or copywriting or writing an email things that a lot of other people get really hung up on or that's where they get stuck. The barrier to entry there has really lowered. So the main kind of places I teach people I talk about in chat to be profits is to start with, what are things you do every week that are repetitive Is you can really push that repetitive task to chat GPT and because you've done it so many times, you can quickly error correct. So you go from doing a task to managing the AI doing the task. So when I, for example, edited that book, I edited it with chat GPT and it took one day. I did the entire book while I was watching a movie on the other screen and I'm just error- correcting instead of manually editing, which is how I've had to do it in the past. So a two- week task took one day and not 100% of my attention.

Jonathan: 6:15

The second thing I look at are tasks that are outside your area of excellence. So everyone's really good at certain things. Like I work with a lot of coaches and the last thing they want to do is sales. Every time you talk to someone like I just want to coach or I just want to do, I don't want to sell. I don't want to do phone sales. Everyone dreads phone sales like it's a nightmare and I go well, okay, well, what if you use the chatGPT and just write a script and you just read what it says when you're talking to someone, and removes that entire pressure of I have to come up with something to say and suddenly the thing that's the hardest for you becomes easy. So for a lot of people, that's copywriting. For a lot of people, it can be something really simple, like for a podcaster.

Jonathan: 6:53

Head of task is every week you have to make show notes, you have to make podcast description, you have to make. A transcript, you have to write a blog post around it. These tasks used to take me days. Now with ChatGPT, I can do all of those in 15 or 20 minutes because it's the same time of every single week, the same structure. So, especially when you have an example of what you want it to look like, you know this is a perfect show notes. Here's a transcript of an episode. Use this to make show notes in the same format, because you have an example of what the final product should look like. It's really easy for ChatGPT to give you a great result. So it's just like when you hire an employee if you can give them an example of a good task and then the raw materials for that task, it's much easier for them.

Jonathan: 7:34

Most of the time we hire someone, they just do social media and they go. I don't know what that means. You go just run into my social media and make it grow. So, because we can take a task, we're doing a lot that's repetitive. Those are the first things I like to push down, and it's really about finding areas of efficiency. So anything that you do that takes up a lot of time.

Jonathan: 7:54

Is that something else you want to look at? Right, and people spend huge amounts of time answering email. As you found out right before our call, I don't check my email that much, so I mostly either I run it through. I use so many tools to sort and sift email because I get a lot of emails. Right, you get so many emails and like 10% of them are actually to you and the rest are newsletters or mass emails from everything. So it's like all of this sifting just to find out is this even an email actually that I'm meant to read? It's really hard. So I use all of these tools and using AI tools, and then you can have AI help respond.

Jonathan: 8:25

One of the things that I do is I have an automation. If someone leaves a comment on a YouTube video, it sends a trigger, it goes to an automation, it sends that comment to chatGPT. ChatGPT writes a positive reply and posts the reply. So the person feels heard, even though because in the past I would have never noticed. I'm really bad at responding to comments on social media because I don't get enough that I'm checking it every day. So today I was checking the comments on my blog and I have like a thousand comments in foreign languages that are obviously spam and then one really, really detailed comment from a person that was very heartfelt. But I had to dig for it. I had to dig through, I think, 81 comments that weren't real to find the one real one, so I didn't notice it in the comments like a month old. So a lot of these things that can slip to the cracks. Ai allows us an automation to handle them in a little bit of a better way.

Joeri: 9:18

Yeah, this sounds really interesting. As you may have noticed, the questions that I prepared for you for the podcast, I also used AI to come up with, like questions based on the problems that I give them, because that's, of course, really important how the type of questions that I wanted to chatGPT to generate. Also, the fact about I like to create audio content, so it helps me to create, as you say, this blog content, repurposing Not the automation part that I guess that you use a third- party tool for that integrates with chatGPT to send. Maybe, if can you name the tool that you're using?

Jonathan: 9:56

Sure, I use a tool called Pabli. It's similar to Zinniar, it's just the value of Pabli is that you pay once and you have it for life. So I get 15,000 events per month. I don't have to pay a monthly fee like Zepiro, spending $300 a month. Basically, I paid $500 or $600 once four or five years ago and it just works every month.

Jonathan: 10:16

They have a really responsive team that have added a lot of features and I'm someone who I'll send in feature requests and they tend to get made a lot because I have pretty good ideas or really good feature requests. So they've added a few things I've wanted. Other tools have added access, anything that has an API, and I know the loosest. I'm not a very technical person. I'm very good at AI, but I'm not very technical. I'm not a programmer, but I know enough to if there's an API. That's what Zepiro uses and that's basically like a connection. So as long as there's an API with a software, then I say, okay, let's find a way to connect your API with Pabli. I just actually my new email provider. I got them to create a Pabli connection just for me and I was like listen, I don't want to change all of my forms on my website. I just want to change where they point so I can just change one thing in Pabli and point to you, and they built the thing for me. It was really great. So that's one of my favorite tools is an automation and we often skip over automation. So I have.

Jonathan: 11:19

One of the cool things is that there's a lot of local automations, like on my Mac. I have an automation that when I download an image from mid-journey, it automatically compresses the image. It runs it through one of those compressors which normally like a three- step manual process, and it saves me so much time. So little automations that we sometimes ignore. Like one of my favorite things to do with chatGPT is if I have a list and the list is in quotes or the list starts with the number, I'll say give me this list back, but with no quotes and no number. It saves me a bunch of time. I just have to do that manually.

Jonathan: 11:49

Or if you've ever done that thing where you have a list of first and last names in one column in a spreadsheet and you have to split them into two columns. That is always a nightmare for me. Now chatGPT will do that for me. So those go back and forth between some. I have an automation in Pabli that will do that if it's in a spreadsheet as well, so I kind of go back and forth. But I think that automation can save you a huge amount of time and it's something really worth learning. I think that I can always get better at it and combine with AI just really can save you a huge amount of time and decrease really overhead. I look at how much money do I have to spend on staff and software with AI and Pabli? It's gone way down.

Joeri: 12:29

Yeah, yeah, for me the same thing with my content generation, even with my podcast editing. Now, as I said, we are testing out a new tool now podcast to see how it works. It will also save me time and costs. Actually, it's interesting talking about automation because, yeah, sometimes people don't think so far to do that, like you say, because you use you, you gain a lot of time. Already do this like correcting, like first name, last name, those kinds of things, but adding the automation part really is interesting. I know you also have written a lot of books, so the content creation part of AI. Of course, people can also make mistakes by generating content that's, you know, not really interesting. Maybe give some tips or approaches to how you can use AI for books or something else for content creation.

Jonathan: 13:27

You have to put a limit on whatever AI tool you're using because if it just using general knowledge, the breadth is so wide. Right, because you and I would not write in the same way, even someone who grew up the same age. We can have a different style or a different tone or different way of doing metaphors. So as soon as you start to think about narrowing down, most people think oh, the AI needs more data. No, it needs less data. So there's a couple of cool ways you can do this. One of the ways is to say write in the style of. And then you can be specific, like oh, write in the style of someone who writes children's books. That narrows it down a little bit.

Jonathan: 14:04

But most children's books are bad. If you go to a bookstore, 90% of the books never sell. Right, we want to go. No, narrow it down to the good ones. So you can say, if you want to do like children's poetry, write in the style of Shell Silverstein, right, it immediately narrows it down to an author and all of their books are good. But you don't just have to use people. You can actually use brands. You can say we're going to write children's books. One of my favorite examples is all right in the style of Harley-Davidson, which is a motorcycle brand, and suddenly everything is a metaphor about sunsets, freedom and the open road, like everything is about the open road. And you then have to say but don't name motorcycles or motorcycle parts.

Jonathan: 14:43

And so you can actually write and create styles this way that are not even copying a person and you can merge a couple of people together. You know, now they've added this functionality with GPTs where you can create and save a character. I do that a lot and once you do this you can then have kind of a data set. So what I'll do is say, okay, who are? Give me a hundred authors who've written science fiction books and I'll write a chapter in the style of each of these authors and whichever one I like, I'll start to use that as the character it's. I'm in there. You can write in the style of Jonathan Green.

Jonathan: 15:13

I was working on it, I was doing live training and I said to write a story in the style of Jonathan Green, and I thought it would choose one of the more famous Jonathan Greens. There's a famous one who's written movies like A Fault in Our Stars. He always writes movies about teens falling in love, but one of them is dying. They have a fatal disease. There's another Jonathan Green who writes science fiction, is more well known than me, and it started writing this chapter and I go which Jonathan Green are you using? And it started to describe, and I was like this sounds like me and it described me and I was like, oh my gosh, I'm in there, so it's read at least some of my books and created a data set around me and it sounded enough like me that I noticed it. And you can. There are two ways you can take that. You can go oh my gosh, anyone can copy me. Or you can go how can I use this? So that's how I added to my book. I said, oh my God, I'd switched into that mode. I created a character that's my personality type from that and it was very accurate because I kept getting into fights. Because it's why I think if I had a twin, we would always fight with each other, because I have very strong opinions and I have this issue.

Jonathan: 16:15

When I'm editing, I'll do a lot of rewrites and that's what it was doing, and I go stop doing what I already do. That's the problem. I don't want to repeat. So it was actually doing something that I am guilty of. I've done many a rewrite where the book ends up twice as long during the edit when it's supposed to be getting shorter. So it's a really interesting experience.

Jonathan: 16:32

So you just have to decide how you're gonna approach it. You're gonna try and put the genie back in the bottle and say open, AI, take me out of there. But you can't, because once it's out, it's out. Or you can think, how can I leverage this? And if a bunch of people choose me and want to write books that sound like me, I guess that's okay, right, like, just accept it, because you can't undo. So it's not just me saying, oh, let's do this to other people, you can do it to me. So I'm in the machine and it's kind of like it's a weird thing. You just have to, I guess, be flattered by it rather than scared by it, because it's just more and more people are gonna be in it, because it's just the data sets always gonna grow.

Joeri: 17:07

Yeah, it's because you put out a lot of content on blogs and perhaps that you know that ChatGPT you found you.

Jonathan: 17:15

So that's a great question. It won't tell me. I was trying to figure out if it had read one of my books and which book it had read, or if it just came from my blog post. From the writing style, I think it's from my books. So I tend to write on my blog in a more casual style. Right, it's a little bit more familiar. My books, my earlier books, were more serious and then it became a little bit more casual as I switched to dictation. But I felt like and it's not, and I don't. I think that because it wouldn't tell me it's not supposed to read books, right? So if it's read my book, that's naughty, right, because that means it's stolen, so that's. But you know what am I gonna do? And that's the same thing, because I know it's done. Other books, like I know some other authors are trying to sue them because it has their data sets and it definitely read their books because they're authors who don't blog. So that's probably what happened.

Jonathan: 18:08

And it's very interesting to kind of be, because I don't think of myself as that famous, right? Like I always tell people I'm a little bit famous. To 10,000 people or to 50,000 people, I'm somebody, but to everyone else. They've never heard of me, right? But I guess I crossed the line enough to where it pulled me into its data set and it allows me to create content faster. It gives me a great brainstorming partner to come up with different ideas and to write things that are in my style, because it does tend to sound like me and my oftentimes this is how I write is I love to do a rewrite. So whether it's a sales page or an email, I'll have it write the first version. That won't inspire me to write my own version. So even if it's bad, other people hate that. I've worked with people who they hate that They'd rather just write from beginning their own way, but because of how I write, it kind of works for me.

Joeri: 19:01

Yeah, I did some interesting experiments some time ago. I did my own sales page and then I said, OK, now ChatGPT may make it better, but not just make it better. But you are in a competition with Gary V, with Ryan Dice, with all of the big names. And then in a few iterations, I came with a better text than in the beginning because I had it in a competition with others. So, yeah, you can play and try to make it better. And what's also interesting is that you can create your own GPTs. I'm not sure if you're referring to that too when you said creating your own different characters. What I did was I created my own GPT and I just connected to my blog to have it create stuff based on already my writing style. But I don't think that there is already an option to just limit it, the chat GPT, to just one source, because if it doesn't find the answer, it will go broader. Do you feel like it's always your style or it's going still broader?

Jonathan: 20:07

Yeah, you have to make a really strict kind of prompt that says you're writing in this style. There's some tricks. I'll give you those tricks. Here's what I do is I say, when you're writing as Jonathan, start and end every answer with this emoji, and you just pick whatever emoji you want for your character and what will happen is there's an element of drift. Right, there's different names for it. People call it dreaming. It's where it starts to stop, like it starts to lose your initial prompt. That's a warning sign for me. That's my carry in the coal mine. When it stops doing the emoji, shortly thereafter the quality plummets, right, it stops staying in there. Now they've recently increased the amount of memory to 128,000 tokens. It's so frustrating to use tokens because it's when you go to the arcade or you want to go to the casino. The reason they give you chips is so it doesn't feel like a real amount of money. You can't tell how much money you're playing around with.

Jonathan: 21:01

That's when they get kids with tokens in an arcade. It doesn't feel real Like tokens. Here and I take my kids to the arcade a token is worth like six and three quarters of a penny or like who's going to remember that? Right, it's a really weird fraction of money to make it. So it doesn't feel like you're spending and I'm like I think that's why they did tokens to just make it frustrating, because no one knows what a token is, but it has a longer memory. But you can.

Jonathan: 21:25

When I was editing my book I think I had to restart five or six times where it just really drifted and it couldn't pull back. So that does happen. And again, part of it is how large of a data set it has to pull from. So if you choose a copywriter that has written like 10 sales letters and one is written to 1,000, with the second person as a smaller data set, so it's either going to get really repetitive or it's going to start to drift sooner. You're absolutely right. So what I do when I'm looking to create, I'll say and there's something I talk about in the book specifically I say, hey, who are the top 50 brands? That's how I came with Harley-Davidson, hoover and what I'm looking for is not who's their favorite, but who do they have the largest amount of data on. So if they have a lot of data, then it gives me enough to pull from to create a character, or I call it a cyber staffer or create a GPT it's called now, so that's where that can come from. But you want exactly a large enough data set. So a newer company is much harder to work with, right. So if you say, oh, just right, in the style of SpaceX has a very limited data set, we say, right, in the style of Apple, apple's been running commercials since the early 80s. They say, right, in Harley-Davidson, I think they've been doing commercials since the 40s or 50s. So the longer brands been around, the more commercials they run, the larger the data set. So that's definitely important.

Jonathan: 22:39

When you're doing and you have to manage, you never switch from me doing the job to AI doing the job. But I don't watch, you're just moving from I'm. You're basically moving yourself into management. So you still have to read everything that AI writes. You still have to error correct it. You still have to check, because AIs will do things that are not appropriate. They don't have any morality, they don't have a sense of right and wrong. We've all seen when they put an AI on Twitter. Every time Microsoft made a Twitter AI bot, people got it to say horrible things within one day, because that's the game. How fast can we turn the AI evil? And it's very fast. So they're very easy to trick and you want to be aware of that. So it's going to make mistakes sometimes.

Jonathan: 23:24

So even as a power user, as an elite chatGPT user, sometimes I just go. You know what, I'll do it myself. Sometimes I get frustrated and then sometimes I come back and go OK, I found a way around this, I can get the result I want. Again, sometimes it just doesn't write the email I want or it's not quite getting what I'm saying and it's like sometimes it's a smaller problem, it's faster for me to just do it. So today I was writing a LinkedIn post. It didn't.

Jonathan: 23:50

The response from chatGPT wasn't quite right. It kind of took the wrong angle and I was like I'll just write it myself because this is a five minute task, so the training would take longer than the task. But that's part of the process is to really see it as a cooperative venture, like I see chatGPT as a car. Car doesn't work unless I'm driving it, and if I'm in a car I'm faster. So it's cooperative. So that's really the process I take with it, but it's not 100% correct all the time. It's always getting better, but the human element is still important. Like I don't know if you saw the news this week, sports Illustrator just got caught being very naughty, creating fake people with fake biographies and writing fake articles. And when you read the articles you can tell, because it had some things about volleyball that weren't correct it had a wrong rule there and that's what can happen is if you over trust, you make that mistake.

Jonathan: 24:42

That's why you have to trust, but you still have to check it, just like any employee.

Joeri: 24:47

Yeah, what I sometimes do, or what I did, was, for instance, for my blog, so that I challenged an alternative from chatGPT, like, for instance, kulot or Bart or whatever, and then see what is the best that comes out of it. Are you focusing, because, of course, with your book it's chatGPT, but are you also using alternative and to see how it performs, comparing to chatGPT?

Jonathan: 25:10

So I like a lot of what Claude is doing. You know, every time chatGPT increases their memory, Claude goes. Because chatGPT goes, we'll do 128,000,. Claude goes fine, we'll do 200,000. So it's interesting to see them have some different features. I have a limited amount of bandwidth. I haven't even tested all the things I want to test with chatGPT. There are so many AI tools, there's so little time and I can only spend so much time kind of balancing testing things with teaching and creating content. I think that Claaude is doing some very good things.

Jonathan: 25:39

I'm a really actually a big fan of Perplexity. That's one I played a lot with. I love Perplexity for research, so whenever I'm writing an article or a sales letter or a webinar and I need evidence, so if I need to find a piece of information with a citation and a source, perplexity is so much better than any other tool I've ever used before. It will give you a summary and have like a note with a number you can click on to go to the article, the source. It gives you the sources, which is the opposite, because chatGPT won't. ChatGPT is like notorious for not knowing or lying to if it doesn't have the information as table found out to their chagrin. So Perplexity is very good for that. I really like Perplexity for research. They just released their own AI model, which I'm very excited by, because before they were relying on the chatGPT and Claude APIs and in fact with the paid version of Perplexity you can choose which version you want. But now that they've done their own, which just came out recently, I'm really excited by that to see that they're kind of making some development. So that's very interesting to me. I think Claude is really good for longer content. So chatGPT finally caught up and then Claude went longer, so you can put in like a 400 page book into Claude. So I think most people are doing books in longer form content. That's where they're operating.

Jonathan: 27:00

It's just a matter of. I haven't played around with it enough to give a really strong opinion. I also really love what's happening in the open source world. There are some really really interesting things happening that are getting really good to where completely free AI models are competing with the paid ones. It's just a matter of for most people. If you say, here's instructions to install an AI in your computer, it's really complicated. You have to start by saying, okay, first you have to install, like this, visual code executor. Then you have to install this version of Python. As soon as you say that, people go wait, it's intimidating. So it hasn't quite gotten. It's getting closer with some cool downloadables. I think it's going to get much better next year. That's the area I'm most excited by. I think open source AIs are going to be the great equalizer, which means people who are from countries where they don't have a lot of money, because one of the things I don't like about Claude is that it's very, it's really hard to access outside of America and.

Jonathan: 27:57

I have a real problem with that. It's like, oh, we don't want and maybe this isn't how it is, but this is how it feels oh, we don't want people from poor countries to have access to this tool. It's kind of like remember when DVDs they started making regions because they didn't want people from like America to be able to watch a Japanese DVD, and they even still do it. With digital stuff like Netflix, it's like you have to use a VPN to watch movies. I really don't like that. I think that's a form. I don't much talk about cultural stuff, but I'm like just if it's online, can we stop pretending that these borders exist? For digital stuff? There's no border, so it's artificial, right? It's artificially giving you access to different movies. It's artificially like preventing people from access to certain AIs because I believe the biggest value of AI is not for the West, I believe it's for the East.

Jonathan: 28:45

I live in a country where the average income is $100 a month. Now, what's happened with chat GPT? Because it's free, every single person here can make more money. Because now the biggest challenge for someone whose second language is English is writing in perfect English. Right, there are certain things mistakes you can make as someone who taught linguistics for 10 years that's my master's is in applied linguistics I can usually tell what country a person's from based on the mistakes they make in their writing. Now with chatGP T that's gone. It just equalized everyone whose second language is English, and to me that's very, very cool.

Jonathan: 29:21

So it's not. Yeah, it can help people that are like me a little bit, but it can help people who just got access to a market. They've always been locked out of right, because chatGP T never misspells anything. It ever makes radical mistakes. So that gets me really excited. And with the open source stuff there's a very good chance that the open source AIs will win, which means that everyone will have equal access. There's no cost to an AI. Your only limitation is your computer, and that's going to go away because they're inventing some things where computers can kind of daisy chain like a bit torrent style. That gets me really excited. That's the thing I'm most personally excited by, but that's kind of a little bit overwhelming for most people here. So that's why I mostly talk about chatGP T.

Joeri: 30:01

Yeah, it makes sense, but it's a good message at the end of a podcast episode that is coming, because for me also sometimes it's frustrating. I mentioned Claude, but I use Spoe. com to access Claude because I'm in Belgium. I cannot access it otherwise. So, Jonathan, you gave a lot of value. Thank you for that. I guess people that are now listening and they are interested in knowing more about you, about your book, about your books. Where would you like to send them?

Jonathan: 30:31

You can find everything about me by going to servenomaster. com. If you just Google it because you can't remember the domain name, every single search result is me, servenomaster. And if you go to my website, servemaster. com, forward slash master. I have a free gift, which is the one prompt that I teach people to be able to get much better results from ChatGPT takes five minutes and will make a huge difference, especially for people that just starting to get in. It kind of switches the mode and it switches ChatGP T to smart mode, and so you don't have to be a good prompter anymore because it will ask you for the information to make your life so much easier for you.

Joeri: 31:05

Meaning. I will put all those links in the show notes, as always, if you are listening. There is a blog article there, show notes made with the help of ChatGP T, but with all the right links in there, Jonathan. So again, thank you so much for all your knowledge bombs.

Jonathan: 31:23

Thank you for having me.

Joeri: 31:24

So, guys, again an amazing episode around AI this time. If you think what Jonathan said, and I think a lot of you can benefit from AI. So if you think this useful for people around you, be sure to share this episode with them. If you are not yet subscribed to the podcast, this is a really good moment to do this and, of course, I would like to see you back next time, take care.