15 Unique GPT-4o Prompts and GPTs

Sa. It's jesus. It's. Okay. Hello? Hello? Hello. Where am I? There am I? Okay, we're in focus, ladies and gentlemen. I believe this. Oh, there's still a timer. Let's get rid of the timer, shall we? I believe this and I'm not exaggerating as I am in some YouTube titles. Fair enough, fair enough. YouTubers doing their things. You know, I'm not exaggerating if I'm saying I think this is going to be the most value packed live stream we've had so far. And it's not because of me, it's because of you guys. Because last week I challenged you to submit your GPTs that you created with the GPT builder that I shared with you. And I gotta say, I was floored by the quality of the submissions. You guys are incredible. Matter of fact, you know, I want to say there can only be one winner in this competition, but it's not entirely true because we're gonna have three winners, then we have a cut extra category and there's going to be like 10 honorable mentions here. I'm going to go through like all the most relevant ones. So there's going to be a total of 15 learning opportunities for each and every person watching this live stream. Whether you know, either live right now, hello, welcome in chat, or as a recording later on. 15 learning opportunities. If you wanted to learn more about ChatGPT or prompt engineering, which is essentially using ChatGPT well. Or building GPTs, which is essentially using ChatGpt well, to the second. This is the right place. This is the right place. Because today we'll be looking at a total, a grand total of 15 submissions, one by one. This might take a minute. Okay. The stream is scheduled to take an hour. If it takes more, you know, sue me. Too bad. Also, some of the comments. What's up with this? Akira the dog. What's up, bro? You need a GPT? A girlfriend, Trans person? What? What? What is this? What is this Vibe, man? Why? Why? Please explain. Why would I need a GPT girlfriend, trans person, man? I don't know. You got me confused. Anyway, much love to you, man. Thanks for taking care of me and thinking of me. And also thanks to everybody else sharing where they're from all across the world. This is fantastic. I always love it when you share the countries you're from. Again, I tricked you one time. You remember two weeks ago, three weeks ago, where you shared the countries and then ratted through the code interpreter and created an interactive world map? I guess it wasn't interactive. It just sounds nice if I say interactive, but where I created a world map from all the countries you were at. We're not doing that again. I'm not stealing your data from chat, you can just share where you're from. But I'm going to be doing a different kind of fun thing to get warmed up here today. You know, we like to use the first minutes to get acquainted a little bit. Akira the dog sharing, you know all his thoughts already. Please share 0 to share yours too. Before we get into the education here, and trust me, there's a lot I looked at it all, the team looked at it all, we combined our thoughts and now you're about to get the extract of all that presented to you one by one again, 15 various GPTs and use cases and building blocks you can be using in your prompting, all served to you on a silver platter today. Matter of fact, I'm going to be going for them one by one and we'll be going in and we'll be turning this into a little resource that I'll share in the link below. Anybody who's on our mailing list will also receive this. As you know, we only. We basically send out one. You get four emails when you sign up, I think four or five. Then you get one email a week with kind of the extract of the AI space. And then we barely ever email. Like if we build the product once every four months, we launch it on their sure. Or once every six months, depends. And we're going to send this out though, because this is absolutely amazing. 15 use cases, 15 various building blocks or GPTs or variations. I believe that no matter who you are, no matter what level you are from today's session, you're going to be able to take something away that is interesting, that is valuable, that is applicable. Isabel, hi, from Portugal. Well, hi back from Portugal. Also, I'm sitting in Lisbon right now, very close to Marques de Pombal. You might know either way, Euromaster saying, good thing I didn't submit this week. Why is that, my man? Because you won last time, that's why. Right. What about Apple AI announcements. Interesting. We can talk about all these in the end. I want to focus here though. Okay. I'm going to focus on first of all, some fun images of Igor in the countries you are sitting in right now. That's the first thing that we're going to be doing in. Okay, doing here. Okay. So we're going to take some countries here that you shared in chat and run it through the kind of One step in our workflow to create thumbnails. We've been getting this question way too much. I'm addressing it here. Maybe we'll create a dedicated video. I don't know. I'm addressing it here. We're using a fine tuned version of stable diffusion, but we're using it in a completely primitive pay to win very simple to use application called ArtFlow. It is a paid subscription. This is not sponsored at all. We just love using it. It's so simple. And I'm going to just quickly show you my workflow based on one of the countries you guys are from. How does that sound? So I'm going to switch to my desktop here. I think this should do it. There we go. Let's add a camera. Let's add a camera. There you go. Hello everybody. And what I'm talking about here is I'm just going to take, hmm, we're going to take a random country here. We're going to take Canada. I see a lot of Canada in here. And here's basically my workflow. I have a little GPT that I call Igor around the world and then in here I can type in anything. And it's set up to be a custom, you know, prompt generator for this ArtFlow model. And in ArtFlow here I have one custom trained model based on my images where basically, you know, I think there's a subscription to this. I'm not exactly sure how much this cost, to be honest. We added like, yeah, we added like $30 and it's. I don't know, then you're good for like a long while. It's not even a subscription. But basically we trained this custom model. See, this is where we got our, you know, Igor, Steve Jobs thumbnail from yesterday's video. Basically we just iterate a lot and then we pick some of these. Right? Isn't this great? Like these are really good. These are really good. And the point is, you know, this is way better than Midjourney and all the others. So we're just doing a fun little thing here where we're taking a country from chat, which was Canada, and then this little GPT of mine generates me the prompt. There you go. By the way, quick little side note, as we generate this, I just wrapped up a 1 hour, 20 long session in the community teaching people the last step of the entire. What am I doing? Just clicking random buttons. Here you go, right here. We just wrapped up a, you know, 70, 80 minute long session where I fleshed out this GPT. I added actions and Knowledge base. I added keyboard shortcuts, I added the hacking protection. We did it all in one session and that kind of rounds out our learning path for GPTs. But again, that's kind of the advantage of the community if we, you know, like it builds on stuff. Like we did a separate lecture on actions and now I could reference it and show you like, yeah, this is, this is how you save stuff, the notion really easily. Anyway, here we're going to be focusing on self contained knowledge stuff that you can implement today. So, you know, this is really simple to use. I have a little GPT that generates me the prompts that I set up for myself and then I just hit generate and it's going to give me four variations of Igor in Canada. Let's look at what we're working with here. Dynamic portrait of Igor in the breathtaking wilderness of Canada. The scene is set in a dense, vibrant forest of Banff national park during autumn. Ooh. Igor, with his athletic build, of course, dark brown, medium length hair and deep brown eyes is captured mid hike on a rocky trail. So this is going to be hiking Igor. But as you can see, this is the workflow that we use. This is where it starts. It starts in ChatGPT. We generate prompts, we bring them in here, we iterate a lot, then we take the best ones. And as you can see, this is pretty good. Yeah, Igor just like hiking. But then we can also remix this stuff right in the GPT that I built in the community session right now, I basically created this and also here I can say Canada, but I'll show you some of the capabilities here. Okay, then it generates me, it generates me the prompt like so. But if I go ahead and actually zero here just to show you like what's possible, then it gives me a little menu of keyboard shortcuts. Nothing revolutionary, but like one of them links to my notion database. So I can just store things in a notion database or I attached a file with all US states in there and then I can just generate a variation remixed in the context of free random US states. Or I can just create a bizarre analogy for a busy New York professional type target audience. So that's what I'm going to do. So I'm going to create a, I'm going to generate a prompt of Igor in Canada targeted at somebody in New York living, you know, hustling away. Let's see, Igor is a high powered executive in the corporate jungle of New York City, but he's in the secluded lake in Yukon territory. Okay, I don't know, let's see how this performs. But basically, custom GPT like this, if I wanted to store this to my Notion database, no problem. I just type 2, hit enter, allow the action, and it's in there. It's using files. It's all built on this custom GPT generator that we talked about. Okay, Yeah, there you go. We're going to confirm this, it's going to save it to my Notion database, these prompts. Then my team can keep working with it. This is how we work. We use a combination of ChatGPT, we use a combination of GPTs and ArtFlow and then Photoshop to kind of do the final touches, to do the text, to do edits. But basically here you have that New York analogy. Not the best here. I would need to do some refinement. I didn't even read the prompt. But fair enough. I think you get the point. So this is how we integrate it. We combine keyboard shortcuts, actions, knowledge basis, custom instructions, and other applications like ArtFlow and others in our workflows to create results that. Otherwise, you know, how long would this take? As somebody, let me tell you, as somebody who's learned Photoshop, when I was 15, now I'm 30. So that's 15 years ago. Creating something like this in Photoshop, first of all, you need a reference image of the person in the right position, right? If you want to create this, you can't just have a headshot like this, like you're seeing me right now. So you kind of need this position. So you need to create those images. That takes a while. Then you need to cut out the person. Then you need to find. Then you need to, like, transform the heads to fit. Then you need to blend them in. Then you need to do color correction both for Luma and Chroma, so it fits the original image. And then you're set. And then your boss is like, oh, yeah, that sounds so good. Can you give me, you know, 10 more variations? And then you're screwed. That's why stable diffusion is great, because you just use a GPT to create, you know, variations. I don't know, I just hit one. It's going to use my knowledge base and code interpreter to give me three more variations based on different states in the us and then you can just generate them here for each power of AI. Anyway, just wanted to quickly demo this to get things going. Let me catch up on chat. And with that being said, it's time to dive into GPT building and it's time to dive into submissions Agenda for today, very simple. We start with welcome music that already happened. Then we go to GPT. Then we go to the artflow process and the thumbnail creation process, showing you kind of the GPT and the app that we use. And then we're moving on to the GPT building challenge. I'm going to recap what the challenge is about and then we're going to look at each submission that has some learning in it for you. So if you ever enjoyed learning from this channel, then I think you'll absolutely love this because now you're going to be learning from the, from some of the best people viewing this channel through my voice. Crazy stuff, very informative stuff. Then I'm going to give you some updates. Yeah. Then we're going to announce the winners. That's going to be great. Those are the best ones. And then I'm going to give you some quick community updates. What we're doing there a lot. We're building unreal demos with multiple AI agents. We're building. We're having guest speakers like Javier Lopez from Magnific coming in in two days. Matter of fact, in the community, where we're just doing a lot where no code stuff, it's incredible. I just want to update you on that briefly. Then I'm going to update you on kind of the changing pricing as announced a week ago. And then I don't know why I just did that. I'm just in a good mood today. I went kiting. Usually I'm in a good mood after I go kiting. It was, it was wild, by the way. Holy moly. 32 knots, wind waves, taller than me. It was wild. I'm going again tomorrow, but either way, then in the end, then in the end, I have a surprise for you. I'm going to show you something that we developed with the community and with a very talented prompt engineer that I've been collaborating with since quite a while. And it's a free GPT that we're releasing into the wilderness today. This has been in development since, I don't know, three, four months. Internally, this is version 12, I think 11, something like that. Externally, it's version 1.0. Because it's clean, it's new, you're gonna love it. If you, if you're into GPT building, if you're into, you know, using ChatGPT, you're gonna love this one. I would bet you're gonna love this one. So, yeah, gonna give that away for free. In the end, if you stick around for all this Madness, all this learning. If you're gonna stick around and listen to Igor blabber about large waves in Portugal. Let's see, let me quickly catch up on chat and let's get into the challenge and all the learnings from that, because there's a bunch. So let me see. Let me full screen myself and let's see what chat has to say. Some great discussions between community members, YouTube community members. Oh, everybody talking about WWDC. In the end, we'll do Q and A. Okay. Once I'm through with this, should take about an hour. I guess we'll do Q and A. I just slowed down a little bit. This is. I'm like a bit. I'm high energy right now. Feeling stoked, excited about all this value that I'm about to, you know, present to you. Should be good. But yeah, wwdc, very interesting. I don't know. There's so much hate. Where does the hate come from? I mean, some. The privacy aspect. We'll talk about it in the end, but, like, I think the privacy. They handle the privacy as aspect better than people would have expected. And it seems to line up. Some deeply technical people are, you know, confirming that it's everything they're promising. So we'll talk about that more later on. Okay. Oh, no. Pigeon saying, bro, I still have the new ChatGPT UI. Mine still has the old one. Not the new bubbly, more conversational one. That's what you mean. You don't have that? Nobody has that, bro. Pigeon, my man, you're not alone. You're in good company. Matter of fact, you have 125 other AI enthusiasts that are bummed out on a daily basis. Probably checking the ChatGPT app like, I am still not there. Still not there. It's a real shame. I hope they ship it soon. Okay, Okay, good, good. This is a good intro question. This is a good intro question to transition us into the challenge and all the GPTs you guys have built. Onor is asking. Might be a dumb question. No dumb questions here, Onor. Let me establish that. Except maybe the one in the intro. Like, why. Why don't I have AI girlfriend still? We'll still handle them, you know, so no dub questions here. What is different GPTs? One more time. What is different GPTs? Why would you create different GPTs? Well, owner, my friend, let me pull up my whiteboard. And yes, I brought a whiteboard. See, this is. This is how serious I'm about to. This education here. We have a whiteboard. Just got to make sure I Don't use the permanent marker again. What happens way too often. All right, my man Onor, check it out. This beautiful whiteboard. It's not so wide up here. Perfectionist and me needs to get rid of that. This, my friend, is all of GPT4. And then you know, the little. Oh, this is all of GPT4O knowledge. All of the Internet, right? Sort of including YouTube, probably including comment sections, including blog posts, including everything you can find on Google. It's referred to as the wider Internet. What exactly is in there? Nobody really knows except some OpenAI employees, but it's basically everything humans came up with on the Internet. Now what is a GPT? What is GPT? Is this, is this. It's just, it's just one thing. So. And by that I don't mean it's just one like piece of the data, but like if you're, if you're interacting with GPT4O, you can assume that it has knowledge of everything. And that is, it's an assistant based on the context of everything. It's an, it's an universal assistant. Okay. It's a very highly qualified and extremely, not just well read, but like it's like a godly entity that had tens of thousands of years to read everything humans came up with on the Internet and now it's there to like answer questions. And with godly entity, I just mean like it had like a long lifespan. It's not, it's not, you know, super high iq, but it just had a lot of time to read all the stuff on the Internet. People came up with GPTs. Not a universal assistant like that. GPTs is more like result. I'm actually coming up with this stuff as I go, but I think this is really great. It's getting you from, you know, the universal spot to a specific result. Specific, okay, that's the key word here, my man owner. GPTs are there for a specific purpose. It's not a universal assistant. It's an assistant for one damn task. Like I showed you. It's a prompt generator for images of Igor around the world for a custom model that uses the wording in it with these characteristics. It's a super specific use case specific chatbot. That's what GPTs are and that's why we've been building these since a while now. There's a lot of criticism I. And it's all well founded, you know, like other models may be better for some like niche use cases. Opus, probably better for prompt engineering. Gemini probably better for. I'm kidding. It's actually, it is better for certain things. It is better for, like, writing in your own style, to be fair. You know, it's just. It's just kind of a meme at this point. Gemini. So, you know, there's some advantages to other models, but at the end of the day, GPT is a specific use case, and that's what we'll be looking at. We have 15 different examples of specific use cases for you. How community members handle them, how they implement the specificity, how they implement the behavior that they want to alter off the GPT. Okay. And what's cool about GPTs is that you can set them up in a way where it becomes really relatively intuitive for somebody who has never used ChatGPT to actually get great results. You can hide things behind hotkeys. You can implement any prompting techniques behind under the hood without them having to interact with it. That's why building GPTs is so great, but it's a bit trickier. And that's why last week. Well, last week, matter of fact, when the GPT Store came out, was that like a few months, five months? I don't know. Was it like January? I'm not sure. Yeah, I think it was late January. Anyway, when the GPT Store came out, I released a video. It was like, hey, GPT Store is here, by the way. I have no life. And I spent the past few days engineering a prompt that helps you build GPTs from scratch that works better than what OpenAI has. And it does. I'll stand behind that always. This works so much better than what OpenAI has. It just does. Please, you know, I'll fight you on that. Oh, Gregory, hitting us with the super chat. What's up? What's up, Gregory? Appreciate that. Thank you. Good to have you here. If you have a question, hit me up with it. But that's what a GPT is. Okay, so basically, I showed you this GPT builder. Now we're recapping. Now we're transitioning into the knowledge part. Are you ready? You know, all the people watching the recording. Look, just get to it, man. Patience is the key to life, my friend. Patience. And now I kind of just a part of me just wants to sit here and, like, piss all the people off or, like, just get to it. Just want to sit here for, like, a minute. Not gonna do that. We're gonna. We're gonna get to it. Okay, what are we getting to? We're getting to GPT building. So last week. Well, in January and last week and before now in multiple locations. I showed you this GPT builder. It's very simple, it's freely accessible. I share the link in the chat right now. By the way, if you enjoy it, you know, please leave an honest rating. If you, if you really like it, you know, express your opinion. We would appreciate that feedback always welcome. But basically you just start by building a GPT. You say, what's the goal? I don't know. Mental health Coach for busy YouTuber overwhelmed for all this AI craziness. Okay, that's fine. And then it creates it for me with five clarifying questions in between. You can always just say auto here and then, yeah, autop, doesn't matter. And then it automatically, I built a feature in there, it automatically just answers all the clarifying questions. And now we get two sections now basically speed running what we covered last time and in the videos. And then we get a communication sequence which is pretty basically a manual on how it interacts with the user. And then we get instructions down here, which are basically guidelines for how it should behave. This is where its behavior is programmed. This is where the communication sequence or the interaction with the user is programmed. It's one way to look at it. Then you get a little image for the GPT. You get a little title for it. It's the AI Mental Health Coach. And you get a little description. You can copy this, this, and then here are all the instructions into a new GPT where you just say, explore GPTs, create. And you just copy this over and you're ready to go. So these are the instructions, right? Boom. Paste. This is the name. Copy paste. This is the description. And again, I'm just speed running everything we covered last time. You put it in here so everybody's on the same page here. And then we have a beautiful little image that it also generates for me and a shout out to the best AI learning community on planet earth. The AI Advantage community. Not sponsored. It is sponsored by us. By me. Great. So we have this upload photo. By the way, I'm a little hyper today. I gotta gotta admit, like the beach kite surfing always gets the best of me. Just upload this. What a great image. And I have my little AI mental health coach. Wait, I want a little applause for him. Great. So now that we have this, we could publish this. You can use this, you could refine it, you can add a conversational stutter like, hey, I was at to add something. Or let's start and you're off to the races. This is what we covered last time I showed you this. Okay, I Showed all of you this and I was basically like, guys, now it's time for you to use it. And I issued a public challenge. I went ahead and I said, guys and ladies, so many ladies here. Actually ladies, it's great to have you around. Always bringing in some of the best feedback. To be honest. Everybody, I challenged Everybody to create GPTs with this. And you guys did. You guys did. Some of them submitted, some of them slid into the emails and were like, hey, no one, you know, publish it. Some people send me DMs on the community, some people shared it publicly in the public Challenge. Matter of fact, you can find a link is in the description I would need to double check, but I'll share the link in chat. You can find all of the public challenge submissions in the public area of our beloved community. What does it look like? It looks like this. Right here are all the submissions we'll be spending the next probably 45 minutes talking about. Okay, copy paste, there's a link for you. There's a link for you. Excellent stuff. Okay, Gregory saying the meaning of life is 42. Yeah, that's a movie quote, right? No, actually that's from I didn't read the book. But that's a, that's a literate, a literary reference to point this. Okay, deep stuff here in the chat, but let's focus on what is happening in here. And what's happening here is all the submissions you guys sent in. Okay, so much good stuff here. I looked at it all, the team looked at it all. There's some fantastic discussions happening in some of these comment sections here. Look at this. You can talk to the creators, you can talk to the team, you can talk to other people in here. Free account can just check this out. Everything I'll be talking about today you can find in here. And if it's not here anymore, if you're, you know, late to the party and you're watching this like two or three weeks from now, it's not going to be in here anymore. We probably moved it into the challenge archive and we'll provide you with a separate file that includes all the learnings, all the building blocks, all the GPT links, everything. I'll be showing you here. Mark Smith, what have you missed? Nothing really. Just eager talking nonsense for 15 minutes, sort of. I showed you a technique how do how we create our thumbnails step by step. I speed ran the GPT creation process. I introduced the challenge, what it is about. Few more details about the challenge. Matter of fact, you came just in time it's going to be amazing. Speedy. Yes, it's free. You can join this area of the community for free, this public challenge area, chat, few guides and event recordings and stuff. And basically what I challenge you to do is use the GPT creator. But there's some constraints. Everything is fleshed out here, but the main part is that there's going to be free prizes and a bonus prize, a yearly membership, two quarterly memberships, and one more quarterly membership for a fully custom GPT. Now, on the main competition, there's one constraint. You know, where is it written out? Add only one building block to instructions. It should roughly match the length of the other building blocks. Now, I realize this is tricky. Some people added two or three. That's okay, but I'll be only judging one. Okay, so I'll be only judging one. Some people did multiple submissions, which is fine. I think in the stream, I said up to three submissions. And we're going to be judging the creativity and the usefulness, but not the usefulness to everybody here. The usefulness to a specific person that might have that problem. Okay, so with that being said, we can get to looking at these. What do you think? Because there's many of them, people use the builder and they shared so many GPTs and so many instructions for GPTs. Like, the variety here is just incredible. I think this is actually, you know, I kite often, not always am I this hyped. I'm kind of excited because literally there has been so much knowledge that has been, like, openly shared by you. And I just gotta say, I'm kind of like moved by the fact that you guys are willing to, you know, publicly share what you guys have discovered and your own experiments and your own thoughts. You know, just like, being vulnerable like that in public is something you don't see every day, especially in everyday life. This is great. You guys are literally sharing your thoughts and your words in form of custom instructions for GPTs here. Now let me point. Let's get to this. Let me point out some of the ones that have been built here. Now, quick overview and then we'll get into them one by one. Again, there's 15 of them. Okay, really great stuff. I'll just go in a random order here. Okay, I'll just pick a few. There's a journey map creator that maps the customer journey step by step for whatever you're doing. Incredible concept. There's an escape room with random rules that are generated every single time you use the GPT. Okay, I'm going to full screen this because, like, I We're going to pull these up one by one and we're going to talk about them. We're going to have a discussion about chat with chat in them. Super excited for that. But basically there's an escape room that you can play where the rules are always random generated. It's wild. Then we have GPTs that help you manage specific illnesses in your life via diet. Then we have random infusion of emojis with the wildest building blocks. They're just like randomly add emojis here and there's. We have another one that has a job matching algorithm now. Doesn't work perfectly, but I think the idea there was incredible. We'll talk about this. Basically the GPT is meant to find you your dream job and you gotta just start by uploading your cv. And then there's a building block that links everything in your CV to other info you provided with. And then it kind of matches that and gives you a recommendation. Great stuff, right? Super interesting. Then we have another one that, where it splits the character of the GPT into three parts and then it basically. I think it was Albert Einstein. Marie Curie. Is that how you pronounce it? Curry, I think. And then one more, I forgot. But basically everything you type in, everything you brainstorm with it, it brainstorms from the three perspectives. But not just that, it brings it back together and it gives you a resulting list of results from all those three perspectives. Debating it. Basically a very simple agentic workflow for you right there. That's very interesting stuff and much more. We have community members building very interesting stuff for their children. Or we also have had one very interesting one in here where I kind of like hopped in on the discussion. I believe it, this one. Christian here built a universal Translator ahead of OpenAI so they can't even ship. They're so slow at shipping that the AI Advantage community is rebuilding the tooling that they announced right here. Amazing. So guys, this, this should be very interesting. We'll be looking at everything I talked about and a few more. Actually there's gonna be like a 15 in total we'll be looking at. So let's just get into it, shall we? I need to take a sip of water because this is about to get interesting. And by the way, I'm very much looking forward to your thoughts here too. Okay, this should like. Sure, it's a live stream. I. It's sort of a monologue, but it's not really a monologue because there's a chat that I'm watching every Single second. Every time you see me, look here, I'm looking at chat. Okay? This is my main screen where we'll be doing the work. This is. I don't know, the camera, but here's the chat here, you guys. So anytime you have thoughts, anytime you have feedback, anytime you have questions, please share them. Also, what's up with this, like, philosophical discussion here in chat? That's refreshing. I like it. So. So, Yeah, okay, I'm just doing a final review of everything here. Everything in place. Looks good. We have this one. Shall we start here? We can just start here. Okay, so these are going to be in complete random order. Okay. So as I'm done with this, I'm going to announce the winners, but now it's about learning. Let's get into this, shall we? So Franco, that also joined the community recently, and this is going to be a mix of, you know, community members and external members. We open this up to everybody. Basically, Franco created something super interesting here and had some. He had many custom blocks here, but. Ah, yes, yes, this is. I remember. So I have. There's so many. There's so many of these. This one is. Yeah, okay, so this was the learning assistant. Okay, okay, okay, okay. I was just getting a bit lost in my organization here. This is a learning assistant. You can try it out in here. Okay. Again, you can find all of these in here. You just have to look at the fame and Feynman mastery. There you go. It's right in here. So Franco shared this with us, and there's so many interesting building blocks that are in here, kind of in between. Now, the only kind of, like, thing is that there's so many custom building blocks that it doesn't really qualify for the main challenge of, like, add one building block. Right. Because most of the stuff is custom. I just really wanted to highlight this because there's things like gap identification, there's things like reiteration and refinement where Franco puts the GPT in sort of a loop where it should recheck what it has been doing. Okay, Interesting. Okay. Just the chat is a bit distracting. Moderators, please, you know, keep things. Keep things in order. But basically, there's some building blocks that we know and love that we talked about last time, like the interactive learning block, but also things like this. The chatbot encourages the user to refine their explanation. Try explaining the concept again with the new information. It's all about practice. You can add little phrases and little behavioral instructions like this here and there. But what I really like this is, is this building block, gap identification. Okay. And I'll be highlighting specific building blocks and then in the end, in the document where we'll compile it all together for you, you'll have every kind of like exceptional building block that I'll be talking about in the database for you. So right now you can just listen and then you can check in, we'll send it out by email. We'll have it in this live streams description, all that good stuff. But basically, I really enjoyed this particular, this particular building block from Franco. And you know what? Let me actually take a breather here. Let me take a breather here and let's get into teaching mode. This was all fun and games so far, but we're about to get serious. So, you know, if you want, you can join me. One deep breath. Let's learn. Okay, let's go. It's got to have that guy in there always, always. Okay, so gap identification, what is this about? It's about looking at the inputs from the user and then identifying gaps in it. Now this is something I haven't used before myself, but you could define what a gap is, right? If you have a tutor GPT, right? And sometimes I'll just be going off and sharing my thoughts with you here and there. This is what I'm talking about. Okay? Gap identification. You can define what a gap is, and then you can include a building block in your GPT where it is looking for gaps. So you can look in gaps in knowledge in a specific domain, for example. So let's say you're teaching a specific skill. Maybe the skills, the skill of, I don't know, playing the drums. Just because I've seen whiplash recently and there's certain techniques and there's certain progressions that you kind of need to master, right? So one thing that you could set this up to do is, is, for example, like when you start playing the drums, you. You start with just. Just playing the basic notes like 1, 2, 3, 4, 1, 2, 3, 4. And then, you know, there's double time. 1, 2, 3, 4. 1, 2, 3, 4. 1, 2, 3, 4 dot and this translates to all the entire drum kit, right? So you could say, look for gaps in the basics of drumming, and then you can use the particular vocabulary that I don't have right now because I haven't been playing the drum I've been playing for four years, but it's been forever. But you could define the gap as mistakes in the very basics, right? And then as something like GPT4O with the voice assistant, that is yet coming listens to your session and tutors you through the session. It can always be in this little loop of checking for gaps where you define the gap as to be the very basics. So even if you play a more complex set, if you mess up on the basics, it would remind you with a building block like this. So this is what I mean by building blocks. It's like a lot of these are inspirations that you can take and then run with and implement into your solutions. And we're gonna have 15 of these today. Okay, so I thought this was very interesting. You can define what the gaps are. This is a great little building block for educational GPTs. And a few more things here like this additional engagement with a variable in it. Let's test your understanding with few questions. Can you answer this question? You could then add them in here. This is great stuff. Simple GPT expanded, meaningfully expanded. Okay, there's a few more things here. Yes, here. Adaptive learning paths create personalized learning paths based on user progress, providing foundational resources when necessary. So this is fantastic for a tutor, for a beginner level tutor. And this is where I would encourage you, if you're building tutors, make them even more specific. Make them specific to the level of the user as we talked about. Like if you know ChatGPT is a universal assistant and then the GPT is a super specific assistant, then if you're building a GPT tutor, make him customized to the level. Because you don't want to be defining learning paths if you're at an intermediate level. Right. So I don't know. This is what I wanted to share with you so far. I'm just kind of going randomly through my list here. Thank you so much, Franco for submitting this. You can try this GPT down below. You can leave your thoughts in the comment section. Yeah, this is just a very interactively designed GPT that is going to go ahead and it's going to be more as a learning partner rather than just somebody standing over your shoulder or watching you. A lot of interactive, like all the interactive building blocks we have in the builder and many more integrated here. Gap detection and adaptive learning paths and user guided decision making and a few more. Everything set up. So this is as interactive as possible. If you're building something educational, you will want to have a look at what Franco did here. Fantastic stuff. Move in on to the next one. And again I need to. Okay, perfect. We'll do it like this. So this time we'll be looking at a very interesting one. Okay, so this one Is from Faya. Monkey records, I suppose. Right here. Thank you so much for submitting Faya. And guys, this thought and the implementation of this thought is a very interesting one. What Faya did here is she integrated three Personas into one GPT and she host a debate between them. But usually when that happens, you know, how often do you guys use it? There's like the whole process of like generating the debate and then like reading the debate and then like drawing conclusions from the debate. Let me show you what happens if you just try fiasco here. Okay, Chat. Random question. Oh, wait, this is perfect. A random question just popped up. This is as random as it gets. Do you know anything about AI tables? Perfect tone mind. You're coming in clutch here. Thank you so much for the random question. So, fiasc. Let's have a look what it knows about. Do you know anything about AI tables? I don't. I'm not sure what exactly you mean, but what do you mean by AI tables? Could you maybe define that tone? Mind would appreciate that. Anyway, back to this GPT. So it uses three different Personas, right? You have Albert Einstein, Alan Turing and Ada Lovelace. And then what it does is it generates kind of viewpoints on all of them. Now I'm not sure GPT knows what AI tables are. Hey, Gerhard, welcome back. But basically, it takes all these three perspectives and then it summarizes them for you. I think. I think it's just confused by this question because I'm also confused by this question. I'm not exactly. Not exactly sure. So like, how should I navigate life as a young adult that spends way too much time on my phone. Right? A common problem these days. Let's just throw at it, throw it at it and see what it comes up with. So again, it's Bond, Albert Einstein, Aristotle and Marie Curie with three different approaches. So usually if you do this, it would take a while, right? You would need to generate the various answers. You would need to like, either create different GPTs and then mention them and have them talk to each other. You could create a prompt where you get all the viewpoints. I really liked what happens here because it's for the lazy ones like me. And it goes ahead are specific questions. It gives you the conclusions here at the bottom. Okay, so you don't need to look at the whole discussion. It doesn't even like reveal. Like usually these discussions get super long. But I really like this because with this GPT, you can just go ahead, throw a problem at it, and you get three completely different perspectives. Right here. So what does Albert think? My man Albert, what does he think about this problem? This matter of fact quite serious problem? He thinks you need to identify patterns, identify phone usage patterns. Use apps that track your screen and highlight the most time consuming activities. Okay. Then you have to set goals. Establish clear achievable goals for reducing your screen time. For example, aim to cut down social media use by 30 minutes per day. Great and very rational analytical advice that uses frameworks. Fantastic. So let's see what we get from Aristotle, AKA the philosophical perspective. Embrace the principle of moderation. Reflect on how much time on your phone is necessary for productivity and social connections versus how much is excessive self discipline. Meaningful activities. Great stuff. This is not an analytical approach. Right? It's more grounded, I would say. And then we have the scientific approach down here. Let's see what we get here. Tech free zones, digital detoxes, healthy habits. Understand the impact. So like if you understand it more, you might like change your behavior. I don't think this is how humans work, but fair enough. This is the scientific approach, right? So fantastic stuff. These are workflows and then here are the implementations. All of this is kind of like very actionable. This is stuff you can bring to a discussion. This is not them. Just pondering the thought and sitting there and thinking about it for two minutes and then you have to read all of it. I think this is really great. I think this is a real time saver. And then you can go ahead and you could say something like okay, that is interesting, but what about what do we want? What kind of perspective would be interesting? Give me now do it with Schopenhauer and then do it with, I don't know, Freud. And then do it with, you know what, let's do Adlerada, more of a fan. And then let's do it with Jordan Peterson, I suppose. I don't know, just three random names that came to mind. And it's going to take Albert Einstein, he's going to split himself into three instances. And then you get the various perspectives right here. What a great implementation. This workflow is usually quite cumbersome. Quite cumbersome. I know that I'm pinning this GPT. This is great for whatever you're doing. Brand new approach to things. Let's move on to the next one. But yeah, I mean seriously, good job. Faya. This is very interesting. Here you have the different perspectives. So you have Adler's perspective. You need to engage in activities that enhance your social connection. So any problem that you have here, you can throw it at the historical grades. And they will give you actionable, you know, either workflows to navigate through it or implementations and reflections. So it's very action or it's very like results oriented. That's what we like to do on this channel. Yeah, exactly. Gehat you could like throw the problems that your 10 year old is facing at it. And it's going to come up with, you know, Albert is going to split himself into three personalities. Very interesting indeed. Moving on to the next one. So this was the teamwork GPT, right? We have more. This was like two out of 15. Okay, so stick around for more of it. Let's have a look at the next one. I'll just kind of randomly select one here. Let's go with Pedro. So Pedro just got into GPT building. By the way, this is what I love to see the most because Pedro has a lot of domain knowledge when it comes to health, nutrition and workout. Matter of fact, he's an author of a book about how to like, work out properly from what I gathered here. And he's building one of his very first, if not the very first GPT. Okay. With the GPT builder. This is what the GPT is made for. And Pedro did something very cool, in my opinion. He combined what he's passionate about with what he cares about in his, like, private life. So he combined like what he's passionate about in his career with what he cares about. He's passionate about both health, but also Dragon Ball. So that's why he created the personal trainer Goku. And there's actually. And I have to admit, like, I watched so much, but like Dragon Ball kind of just missed me as a kid. I'm sorry. I was more of a One Piece guy. I don't know, just kind of missed me. I learned German from spongebob. I watched One Piece a lot. I watched many other shows, played many games. But Dragon Ball, I don't know. I think the character is amazing, but I. I just never binged it. Anyway, I thought it was always cool. Like, always. You know, the big boss fights and then it gets stuck on one frame for like 20 seconds. It's like, I'm gonna destroy you. And then he sits there. And then the boss is like, no, you're not. Anyway, so I've watched some, but what he did here is he combined these two. Okay. And there's two interesting building blocks that I want to point your attention towards. First of all, he included an Easter egg. Where are the gamers here in chat. Anybody? Are you guys familiar with what an Easter egg is. I know all the gamers know it. I don't know if other people actually, if this is like common knowledge. Easter egg is a hidden goody, A hidden functionality that does something unexpected. Usually I don't know. That's the. That's my definition. What does Urban Dictionary say about Easter eggs? A hidden item placed in a movie, TV show or otherwise. Visual media for close watchers. Yeah, okay, fair enough. So he included a little Easter egg. When the user asks, want to know how a Saiyan? Do I say that correctly? How a Saiyan. Saiyan. Super Saiyan. I think how saying trains responds. A saying always prioritize form overload. A saying is always humble and if required asks for help. And for the most important part, a sane is consistent with the workout. So it's a little Easter egg. And I'm not showing this off. So you improve your GPTs with this sane training advice. I'm showing it to you as a framework that you can use. You can include Easter eggs, you can trigger hidden interactions when the user uses a specific phrase. Right. I think this is something I haven't thought of before. Heck, I know very well about Easter eggs. I didn't think about including them in the in GPTs. Matter of fact, I might start doing that from here on out. Just super kind of very simple but interesting idea behind this building block. I wanted to share this and there's one more thing I wanted to share with you. And it's the motivational messages. And this one is also motivational messages. There you go. This one is very universal. On the other hand, I think you might want to include this in Most. In many GPTs. This is just to show that you can infuse certain phrases in the GPT's answers at a random pace. Okay, so you can give it a list of. Here there's two messages, but you could give it a list of 15. Very simply. You would just format it like. So if I create a new chat, we have this, such as. And then the colon. Then do this. And then you could structure it something like this. Very simple. We don't even need the quotation marks here. If we do a list like this and then you can include multiple messages here and it will just kind of, you know, you know the salt bay meme where he just sprinkles the salt over his. Of course you know it. You could do that with various messages on your GPT interactions. Okay. You sprinkle them all over it. And I think this, this might be an interesting concept to Give it a little bit of flair, a little bit of randomness. There's only so much you can achieve by defining the voice and the tone of the GPT. But you can kind of go the next step by adding a building block that adds random phrases here and there. I just thought that was interesting and worth sharing. Yeah. Great job, Pedro. Way to go, man. Like, just like build your first workout assistant GPT. Fantastic stuff. Thank you so much for sharing and thank you for like sharing your ideas for like Easter eggs and like random phrases in there. These are things you guys can build into your GPTs. Okay, then we have another one and this one is from Pranjal, I believe. Guys, so the formatting here, you know, it's all one wall. But there is one thing in here. I was playing with every single one of these GPTs that you guys submitted to here. I took a close look and also the team took a close look at all of your submissions. Okay, this is. There's actually something very interesting in here. Although the formatting is, you know, not perfect, Pranjal built a very specific tutorial inside of a GPT. I thought that is a thought worth sharing. That's why I'm doing that right now. So if you share the put the instructions in here or into a GPT, you will see what this does for you. It's a step by step tutorial on how to build your own WordPress site. And it's all built in here. Isn't that kind of cool? I don't know. I haven't seen this done before. I mean, maybe I have, but it's good to. It's good to like. I love this idea, you know, imagine like, think about, think about like the context of a company and a new employee coming in and him getting like all these sops. Why isn't there a little GPT for some of the SOPs where the step by step process is done in here, especially if it's linked to something with AI, it should absolutely be inside of ChatGPT. The new employee will get used to this interface as they progress through the training. I can totally see this being implemented in very smart ways. But as you can see here, it guides you through the whole thing step by step. You can any process you have. DaveyB asking how do I create thumbnails? You want to go back to the beginning of the stream. It's recorded, by the way. You can check it out later on. I started the stream with kind of showing my process with a custom GPT and the app that, you know Generates custom images. But basically there you go, you know, a step by step guide, a detailed step by step guide on how to get all of this done. In other words, I just wanted to show you this capability of you can include any step by step tutorial inside of a GPT and then use all the other capabilities you have in here along the way. Some people so lazy they still won't learn. Absolutely. That's just true. Don't wait until GPT can do these steps for them. That's true, but that's not the people that are here. Right. We all want to like learn and see what this can do and then use it to our very own advantage. Fair enough. So I just thought this was interesting. Wanted to briefly share with you that you know, also this can be done. You can have specific tutorials inside of ChatGPT. Great, next one. This is turning into sort of a use case streamer stuff you could do. I like it. You guys seem to like those. So we're doing more of those. Okay, let's pick out this one. Ah, Gerhard. Gerhard is also in chat right here. Gerhard Kristan right there. And Gerhard builds very useful GPT to him. Can you make it sound like Yoda? Yeah, absolutely. Yeah. That's a great comment. Gyad. That was a missed opportunity. Matter of fact. Done. But from here on out, answer in the style of Yoda. Sorry for jumping around. We'll get to Gerhardt's story time magic right here. Done. You have good. Yeah, There you go, Yoda. But Gerhardt basically built a GPT that helps him create stories for his 11 year old, I believe. Yes, that's what it, that's what it is. And here's the thing, here's also thing like that I haven't considered just because like I don't have kids or I don't create GPTs for kids. And there's one really interesting thing in here that I wanted to highlight. With Gerhard's GPTs, he included content restrictions. This is the building block he added. Okay, avoid political topics, discriminatory content and any subjects that may not be age appropriate. So he's extending the red tape that OpenAI has put around some of the data, some of the topics. Okay. Now you know, usually people are in favor of kind of like reducing the censorship. But if you're reading and creating stories for your 11 year old or whatever other context where, you know, restriction might be a good idea, you can always extend the censorship like this. Like ChatGPT is not just going to avoid political topics. Overall it's going to avoid generating deepfakes, it's going to avoid giving spicy takes on popular politicians. But look at this. Like content restrictions, you can restrict stuff more. So very simple, you know, some of these points, they're not like these elaborate like one hour long lectures on like what to add. But just consider that you can add more restrictions to your GPTs. And if you're creating a storyteller for your kids, you might just want to do that. And there's even like a process to like generating the story. I thought this was interesting to get. Also, if you have anything else to share here, he's saying it's a bit hit and miss. 5 out of 10 instances, it generates the image. See if we get one of those. Okay, that's a chat GPT type of mistake. Let's see you choose the lazy option. Yes. How long should the story be? Yes, yes, perfect. Okay, we're just creating a random story. Let's see if we get the image. Gerhard said 5 out of 10. And then we're moving on to the next one. But interesting, right? You can restrict, you can, you can restrict things even further if you want to. Where is my image? I'm not even sure if it's time for the image yet. But there we go, creating the image. Let's get a little story time image and then let's move on to the next one. Let me pick the next one in between. Oh, Bella. The bunny in an enchanted forest. Excellent stuff. Okay. Oh, okay. So moving on to the next one here. Gerhard, thank you so much for submitting. Great stuff. I hope your kids are enjoying the GPT. Next one, Ruben. He actually shared two versions of this. One where just one building block is modified and one where he modified multiple things. Some interesting things in here. Okay. Some things I haven't considered before. Let's have a look. First of all, we talked about this. I talked about this in the video. You know, five GPT prompts that 99% don't know. I think one of the problems was just this very basic but very essential, you know, ask me clarifying questions that are relevant to what you're doing. Something along the lines of that. And basically Ruben went ahead and he built this into the GPT. This is just one of few things I want to highlight here. I think Ruben did three interesting things here. Yes, yes, yes. Let's do this. Clarifying questions here. This building block right at the top of the instructions. Before starting, ask me clarifying questions that will help you complete your task. Very simple. But just keep in mind you can always keep this in here. Now, this might be better suited in the communication sequence. I think I also left some feedback there, but I think this is a great step. This is a step that I have in my GPT builder. As you have seen, it first asks for the purpose and then it gives you five clarifying questions. I picked the categories for those clarifying questions. They're not random. There are categories that matter the most to GPT building, as in my experience and my research. But like this is, this is a good one that you can always add to your communication sequence. So as you can see, we're creating a big list of building blocks that you could be adding to your GPTs right here. Next one. Story and Folklore database. Interesting look at this. Always use stories that are linked to Japan's history or folklore to express popular Japanese morals and virtues. I love it. It turns it into more of a storyteller. By the way, I just realized I forgot to highlight what this does. It basically is a personalized Japanese cocktail maestro. So it helps you pick Japanese cocktails and create Japanese cocktails. It gives you recipes and helps you along the journey. Now, I think this is really sweet because I can see so many situations where this is useful. We're infusing a little bit of history, we're infusing a little bit of Japanese morals and virtues. In between can be a great asset. You could, heck, you could add this not just to GPT, you can add these building blocks to your prompts, right? Maybe you have a blog writing prompt and you want it to be more grounded in history, more. You want more of the Japanese morals and virtues to be present in the writing, right? So when you're defining the style, you could go beyond defining the tone, the voice, the sentence structure. You could go beyond giving it examples of how you write. You could add something like this too. You should tell it and here and there insert or like every two paragraphs, insert Japanese virtue somehow, right? And it will just randomly insert that. Now, some might work, some might not. You know, it's prompt engineering. There's going to be trial. It's chatgpt essentially. There's going to be trial and error involved in implementing that. But I thought this was an interesting one that you should be aware of. Let's move on to the next one, which is improved talking points. Ah yes, Improve talking points in here. Also interesting, improve a did you know section at the end of your update. This section would give five interesting trivia information related to the ingredients or brands of the spirit liquor. So again, this is sort of a building block that just enhances the user experience, but it doesn't do it in an explicit way. It doesn't say, hey, add this for these five, you know, pieces of. What is it? Interesting trivia, informations. It doesn't say that explicitly. It taps into the intelligence, into the knowledge that you have at your disposal. Remember, where's my whiteboard? There's a lot of information in ChatGPT. All of the Internet is in here, okay? And you can access it and you could randomly infuse it into separate messages into. At the end of the conversation, you have all the facts that you can access for ChatGPT there at all times. You need to become cognizant of that. And then, and only then can you start utilizing it to your advantage. Then you can start doing and crafting little building blocks like this where you say, yeah, just do a did you know? Section. And you don't even need to provide examples because you know that ChatGPT has this knowledge and you just tap into it. Great stuff. Thank you so much, Ruben. Great Submissions right here. Now moving on to the next one. There's a whole lot more. How many more? We're not even halfway through here. Oh boy, oh boy, oh boy. We have 1, 2, 3. We have 4, 5, 6, 7, 8, 9, 10. We did one third. Okay, so let's speed up a little bit, shall we? Okay, let's look at Marcus's GPT. So Markus, one of the OGs of the community, been here from day one. Marcus didn't even bother including the instructions. He's just like, I, you know, I'm not even. I just want to share what I did. Oh, wait, did he or did he add them now? Oh, yeah, yeah, yeah. So. So he didn't use the GPT builder. He just wanted to show what he built himself. And that is fantastic too. It's always welcome. But basically he shared a solution here to a very tricky problem. Right? And the tricky problem, and maybe Markus, is even here, if you hear, you can leave comments in chat for sure. But basically Minier's disease, something I wasn't aware of, but it's a chronic condition affecting the inner ear, leading to episodes of vertigo, tinnitus, hearing loss and feeling of fullness in the ear. Named after a French physician and so on. Okay, so this is the disease. Okay? Now there's solutions to this that are not pharmaceutical, right? There's solutions to this like diet. Diet. And this GPT is basically set up to help people that have many years disease. I Hope I'm pronouncing that correctly. Either way, people who have this disease with with diet rather than pharmaceutical solutions and drugs. Right? Now, look, I'm not recommending you rely upon ChatGPT as your medical professional. Always consult a medical professional, right? But I just think it's interesting that you can do things like this that are a bit more esoteric and a bit more like, unexpected. You could tap into this, like, wide knowledge that GPT has and get recommendations, right? These are recommendations, they're not prescriptions. But look at that. Chef Manure has proven to be a valuable tool in managing my Meniere's disease. By leveraging the power of GPT technology, I've been able to maintain a strict low sodium diet which has significantly reduced my vertigo attacks and improved my Overall well being. ChatGPT. Thank you. The added capabilities such as daily sodium tracking, barcode scanning, and the potential of building a community sharing platform can make it an even more comprehensive and supportive tool for others managing the same condition. I believe this case study highlighting the potential of custom GPTs to make a meaningful difference in health management. Fantastic. He even presented it to some specialists at a UK hospital who are guiding him. They found the GPT to be innovative and have asked me to monitor and record its effectiveness. Great stuff. Look at that. You can do super innovative stuff like this. Again, recommendations, right? Know the limitations of these tools. Don't, you know, like, just don't be stupid. Don't take this verbatim. It's a recommendation. But it knows so much about nutrition. It knows more about nutrition than I do and probably most people in here, right? That's just assuming, like, you know, like, who hangs out in a stream looking at chat GPT building blocks on a Tuesday afternoon or Tuesday evening. Like, probably people who are into tech. Probably people who care about like, GPTs. I don't know, maybe I'm just drawing conclusions based on myself, but I'm very much, you know, order everything type of guy. So I could use something like this. I could use a GPT that helps me kind of like manage my diet a little better. I thought this was very interesting. Wanted to share this. A few more interesting ones right here. Okay, let me just mark this one as covered. Perfect. This one. How about this one? Let's reset the room. Okay, everybody? Didn't you ever, like, think to yourself, you're probably not. That's kind of a freaky thought. But like, have you ever considered how interesting it would be to talk to yourself in like 40 years? Well, you can't really do that, but you can sort of do it. Hey, what's up, doc? You can do this. Okay. It's based, it's inspired. So Christian Graham shares that it's inspired by this article and a research paper that is all about future you. A conversation with an AI generated future self reduces anxiety, negative emotions and increases future self continuity. And he attempted to rebuild this experience inside of a GPT. Now, this is one of those where we don't have the instructions and stuff. Sure, you could like extract them and stuff. But. But anyway, the point is this. The point is there's a GPT that Christian shared with us that he built where you can talk to your future self. Isn't this great? Isn't this super interesting? Now, you know all the links are available here in the community. You can just go to his post. This is all publicly available and you can go and try it out yourself. Shall we try it? It's gonna take a minute, but please answer a few questions to create a realistic. I need to tell about my life, say, Igor Ferdy, Mr. I don't know, where do you live, lisbon? Creating generative AI content on occasional content on YouTube newsletter and the rest. I'm just going to let it, you know, auto assume so we don't spend too much time here. Okay. It's going to make assumptions and I'm just going to do this a few more times. Obviously, if you're doing this, you want to answer these questions to have them or. Matter of fact, you know what I could do? I could just use my personal custom instructions. Aha. These are coming in clutch yet again. Yeah, let me do that. Let me just do this. You know what? I'm just gonna post it in here. Boom. This is way more details. These are way more details. Perfect. Yeah, I'm not gonna answer all that right now, but that's interesting, right, Speedy? That's one way to gather data. Yeah, fair enough. But the GPD creator doesn't get it right. Just open AI. So, same old worries. Teams accounts are opted out of data sharing, I think. But, you know, still. That's still perfect, actually. We can talk about that. But like, the Apple stuff looks really promising. I'm serious. Some. Some people who are way smarter than me when it comes to all this technical stuff seem to like, confirm that this is like a first in terms of cloud computing and that it is actually secure and anonymous. But we can talk about that later on. So there you go. So here it makes the assumptions. Now, would you like the GPT to act as a supportive Listener, provider of coping strategy or something. Someone else, I don't know, just a kind of wise oracle. I don't know. And then you can talk to yourself, essentially. Cool little use case, isn't it? I mean, it's just an idea that flows so well, right? If I like. I like these ideas where, you know, if I leaned outside of the window. By the way, this is kind of my secret to YouTube. If you want to hear it, the ideas have to. The ideas for the videos have to be so good that if you just lean out of the window and you just shout it at a random person on the street, there's got to be a good chance that they won't just like run away and think you're a weirdo, right? Because if you're like. Because look at the channel, like the most popular video, right? Do you want to learn how to use ChatGPT in two minutes? Right. If you take that as an example, that resonates with a lot of people, right? You're going to find some people who are like, Wait, what? Like, okay, interesting, right? If you shot a bad idea at them, 10 out of 10 people are just going to run and you can kind of like. I don't know if you have. I guess, yeah, that. That is a good way to evaluate, like, good ideas. And this idea is like, would you want to talk to your future self in 40 years? Right? Like, you can talk to your future self in 40 years right now. Are you interested? Some people would say yes. Right? It's kind of a good razor, I suppose. Yeah. Okay, so now it embodies this. And. And now you can talk to it. It gives you questions, it's conversational and you can do this. So again, I just wanted to share this. Links are in the community. Fantastic stuff. Or in the description below. If. If. I don't know if you're watching this in like, July or later. Okay. Interesting future you. Thank you so much, Christian, for sharing this. Thank you for sharing your inspirations too, and the GPT. You build great stuff. If you guys, by the way, if you have comments or you want to like, just say thank you. You can always do this in the comment section here, right? You can just create a free account in three seconds and boom, can interact here. Great. So this has been that one. I have one more. This one is a bit random. Okay. This was like a very philosophically interesting one. This next one from Levi or Levy. Man, this one is kind of fun. It's quite simple. Look at that. Instructions, communication sequence. What this one does is it puts you in an escape room and it randomly generates all the rules every time. So again, I thought this is one worth sharing and one worth talking about because it just randomly generates all the ideas. How cool is that? Okay, you can just copy this stuff. Let's go into ChatGPT, create a new chat, Put it in here, I think. Was there a GPT link? Oh yeah, there's a GPT link. We can just play it. My bad, my bad. Let's just do this. There we go. If you could redo your life and want and not want to change a thing, that is the ultimate benchmark. Jonathan. I could talk about that for hours. That's, that's a great point. And I think I actually really disagree. I think I matter of fact, and this is something that I arrived at like a few months ago. I personally, just my opinion, I think it's okay to have regrets. I think the only way to not have regrets is if you never take action and then you're going to have the worst type of regrets. I think it's okay to have regrets. I think it's okay. I think it's okay to sacrifice things that you actually care about and you know you're going to regret it later on. But I think there's reasons that are like significant enough for you to live your life in a way where you're making decisions that you will regret later. I don't know. But it depends. It depends what you consider like a higher good. I don't know, that's a whole different discussion. But like, yeah, good points. Always good points in this chat. Love this stuff. Okay, back to the mini escape room. I want to play a random mini escape room. Yep. So this is great. Look at that. Levy, thanks for sharing this. This is a really fun one. So there you go. She even said herself that, you know, she hasn't won it a single time yet. So this is really hard. Look at that. For everyone who loves escape rooms, to make it more engaging, it's trained to kill you off in funny ways. If you act stupidly, you can, you can of course do stupid stuff on purpose. I guess there even may be a way to solve one of the rooms. I just never made it. I always die. Chat. Shall we play? Shall we play? This could be our next like warm up game. I love this stuff. So now I spot the philosophical discussion chat. Guys, let's play escape room. Let's play escape room. We can do this in the end. Which ones of these do you want? 1, 2 or 3? First? First number in chat decides. Okay. The air is cold and you can hear the faint sound of dripping. Okay. You wake up in a dimly lit room with stone walls. The air is cold and you can hear the faint sound of dripping water. As your eyes adjust to the darkness, you notice three objects in the room. A rusty old key on a wooden table. A cracked mirror in the wall. A heavy ancient book on a pedestal. What do you do first? 1, 2, or 3? Inspect the key. Do you look into the mirror on the wall? Or do you open the ancient book on a pedestal? Speedy saying, free. Okay, we do free. Let's play this. This shouldn't take too long, right? Like Levi said, she dies every time. So let's do this. So you approach the heavy ancient book on the pedestal. The COVID is made of worn leather and it feels cold to the touch. You open the book and it creaks ominously. The pages are filled with strange symbols and eerie illustrations. Suddenly you hear, oh, wait, can I. Do I have a. Oh, yes, I do. Suddenly, you hear a low growl behind you to turn around to see a shadowy figure emerging from the darkness. It's a spectral guardian, summoned by the book's opening. I will get you first time. This filter is coming in handy. Nice. What do you do? Try to communicate with the Guardian, close the book and run to the key. Or smash the mirror to use a shard as a weapon. What do you guys think? What do you guys think? First message here gets it. So do you try to communicate with the guardian that has a voice like. Like this? Do you close the book and run to the key? What do you do? Let me know. What would I do? I think I would close the book and run to the key. It seems like the. The book opening is kind of. The book opening kind of caused this. Kafikom inva invacao. Try to pronounce. Says smash. Night class also says free. Okay, smash the mirror. We're going aggro. Okay, so we smash the mirror to use the shard as a weapon. Let's see how this goes. You quickly grab an air by object and smash the mirror, shattering it into pieces. You pick up a sharp shard, ready to defend yourself as you turn to face the spectral guardian. By the way, anybody tuning in, we're looking at various GPTs and building blocks that you can include in your GPT. This is a unique one that randomly generates rules, and we're just quickly playing it. So let's see how. How does. How does this go? You pick up a sharp shard, ready to defend Yourself. As you turn to face the Spectral Guardian, you notice that the broken mirror has revealed a hidden passageway behind it. Nice. Okay. However, the Guardian lunges towards you. It's ghostly form flickering in the dim light. You have only moments to react. What do you do, Nyklas? What do you do? I want to hear from you in chat, man. You made a great decision last time. We rely upon Nyclus here in the community, man. What does your intuition tell you? Do you fight him off with the shards? Do you stick to the plan? Or do you run into the hidden passageway? I think if life presents you with a hidden passageway, you take the hidden passageway. That would be my take. But try to distract the Guardian with the key. I don't know if that. Creating dall e free images for each. Yeah, unfortunately, can't create free images. Yeah, that'd be so cool. Right? Where it's like. Like interactive thing. Yeah. He says one fight. Ah. Nyclus is set up. Okay, okay, okay. He's, you know, he's a man of principle. He's sticking to his guns. He took the shard to fight. He's sticking with it. Okay. With the mirror shard in hand, you bravely face the Spectral Guardian. You swing the shard at the Guardian, but it passes through its ghostly form without causing any harm. That's not good. Night, class. The Guardian's eyes glow with a menacing light and it raises its hand. Suddenly, a burst of icy wind envelopes you, freezing you in place. You try to move, but your limbs are locked in a sheet of ice. The Guardian lets out a chilling laugh as the ice encases you completely. Gotcha, bitch. Soon we are nothing more than a frozen statue. A warning to any future adventurers who dare to challenge the room. Secrets, man. Rest in peace, night class, man. We should default him, man. There was a secret passageway. Now, if we were cheating, we could edit this, you know, but like, nope, nope, not happening. So we lost. That was kind of fun though, right, guys? Levy, thank you so much for sharing this amazing stuff. Some other people here loving it too. This is really fun. This was really fun. Yeah. Okay. Great stuff. Let's move on to the next one. Thank you for submitting this one. Also, just, you know, before we move on, first of all, I'm gonna miss that monster voice. Heck, how do I integrate this, integrate that meaningfully? That was great. But there's some interesting stuff in here, so. Like these consequence guidelines. Look, there's random rule generation here. Escape room rules. There are no rules. You make the rules. You are the master of Escape rooms. You've built escape rooms for 40 years. You know everything about escape rooms. The user will have the best escape room experience of his life with you, you know, handing it over to ChatGPT. Interesting. And then consequence guidelines also well written here. Very inspirational stuff in this one too. Lovely. Love you. Thank you so much. Here's another GPT. So now we're looking at kind of like the GPTs that have been shared. This is the journey map creator that has been shared by Christian. Wait. Ah yes. Did Christian submit 2 GPTs? I think he did, yeah. You can. Yeah, he did. Two interesting ones, the future you and also this journey mapping. By the way, you know, he took the opportunity to also share the Medium article here. But you know, I thought it was a great GPT. Shout out to Christian's article here. Feel free to check that out. I believe the GPT link was somewhere in here. Wait, I have to enter GPTs here at the bottom. The journey map creator. You can access it like so now this is great for anybody who's as business minded as me, enjoys creating things we create like for any creators out there, basically, whether you're an entrepreneur, you're a content creator, or you just enjoy woodworking, whatever it might be. We create for others, right? That's the whole point. If I were to create for myself, where's the fun in that? We create for others. But the problem is there's only so much empathy you can have for somebody else's situations. That's why, you know, common wisdom and startup land is like, it all begins with the customer's journey. What are their needs, how, what problems are they experiencing and how are we going to solve these? Right? So Misha, sharing. I'm pretty self absorbed. Talking to myself is awesome. Thank you for this fantastic comment, Misha. Well, appreciate your honesty. I mean, fair enough. Yeah, I always like that comment of like I like talking to myself. At least I'm in good company. It's also funny. I think both can be true. It's fair. But anyway, I think like, especially like creation, like Creation really like gathers meaning when you do it for other people. And this journey map creator, I thought this was particularly great because let's say, you know, you are, you have a AI educational YouTube channel now. It might be a little, kind of like it's not intuitive to figure out how people go through discovery, how they feel along the way, how they feel when they discover more videos. This thing does it for you. So let's just give it some info on me again, my personal custom Instructions. Love having those. It's just like some info on me and then now it creates a journey map for. In becoming a pivotal resource for generative AI education. What we're attempting here. Right, so here it begins and I think this is great. Look, there's many ways to do it, but I think doing this with a GPT actually really makes sense because it's so simple and if you have your personal context in here, just throw it in there. Boom. It creates a journey map for what you're trying to create. If you have startup ideas, this should probably be your first step. What is the customer's journey going to look like and what emotions are they going to go through as they do it? So let's have a brief look at this. This is useful, I'm telling you. I wouldn't be highlighting it otherwise. It's great. So it starts with awareness. Alex, a YouTube content creator. So this is the journey we'll be following, Alex. Right. Focused on generative AI, especially ChatGPT and related technologies. Alex learns about the growing demand for up to date AI tutorials and the potential for building a dedicated online community through industry conferences, social media and feedback from the subscribers. Okay, so this is the other perspective. See, what do we do? We just follow up. I would like the perspective of a YouTube viewer that discovers the. What did it call it? The resources for education. So just tell it, you know, like this is not what I was expecting. I wanted the other perspective of like a YouTube viewer, not the one of the creator. I'm familiar with that. Let's see. Oh no, I'm just looking at the graph. We lost some people as we were doing this cape room or before that, but now they're coming back. Fair enough, Fair enough. Okay, noted, noted. Actually, as we were progressing the escape room, it was good. We lost some people before that that didn't like the idea of doing escape room. Interesting. You know, always keep an eye on the data, you guys, when you create. Anyway, so here we are. Here we have a user journey of somebody discovering Jamie, a tech enthusiast and aspiring AI developer with a background in software engineering. Nice. So here we have a technical person that discovers kind of our ecosystem learning ecosystem that is non technical. Right. We focus on no code solutions when it comes to technical stuff. We keep it super basic. We keep everything super friendly towards people who don't know code. Just because I'm creating what I would want to consume and I know the basics of Python. I always say I know enough to be dangerous, I know enough to kind of make things happen that I want to happen. Right. I can create Selenium scripts, I can code basic scripts from scratch even without ChatGPT, but I have never created a production ready application. I've never worked for a company, I've never worked in the team of developers. I'm just the self coder that knows enough to, you know, do stuff, do basic stuff. And that's why I also teach in that way. Right? I teach for people. I don't even assume, actually I don't even assume that Python knowledge. But this is what I'm trying to say is this is very interesting because this is not an avatar I'm trying to appeal to with all of my content. I think the prompt engineering stuff definitely does because you need that as you build pipelines. But nevertheless. So this is going to be, I'm just long winded way to say this is going to be interesting to me. I want to see how Jamie perceives me. Jamie becomes aware of the need for generative AI education. While browsing for socials and online tech forums. A friend shares a link to a YouTube video about the latest advancements in AI technologies. Okay. Jamie clicks on the link which leads to one of Alex's YouTube tutorials on ChatGPT. Alex would be me in this case, right? Okay. It has an engaging thumbnail. Promising title catches Jamie. Jamie's attention. Good. What happens then? Then he starts considering. He watches the video and is impressed by Igor's clear explanation and practical demos. Jamie reads through the comment sections noting the positive feedback and recommendations for other videos on the channel. Hey, that's you guys. You guys make that happen. I appreciate the comments by the way, like the video if you like the stream or video if you're enjoying this, leave a. Like I always forget to ask that and then every time I'm surprised by how much it actually helps. It really makes a difference if you leave a like so you know, if you're enjoying this, please leave one. Now let's move on. Intrigued by the content, he watches the video. Jimmy's motivated by desire to learn about the latest AI tech to enhance his skills and advance his career. The practical approach and hands on demos in Igor's videos seem particularly useful. Okay, so he's intrigued here, right? He's like considering he's seen something. Now this is the part where it becomes interesting because like what you want on YouTube is you want session time. You would multiple people to you. Oh, it's the ads that throw you out. Oh, thanks for the feedback Misha. I didn't realize really no, no, no. This dip is way too large for ads. Like, lost like 15% of people. Now they're back. Anyway, so back on track. Jamie subscribes to Igor's YouTube channel and watches several more videos, starting with beginner tutorials and progressing to more advanced topics. The first video watches a tutorial on prompt engineering of ChatGPT. Fair enough. Yes. Okay. How does it feel, though? This is the engagement part. It's the interesting one. Jamie feels engaged and excited by the way Igor breaks down complex topics into understandable segments. Yes, I feel that I get that feedback a lot. Interesting. The practical examples and real world apps make the learning process enjoyable and effective. Cool. So he's, like, excited about, like, the learning. That's what we want here. Excited about becoming a better, better version of yourself. That's. Yes, that's the goal. Jamie begins to integrate watching Igor's videos into his daily routine. Prompt engineering daily. I don't think. Yeah, God damn it. I don't. I don't upload enough for this. You know, my attention is too scattered, dedicating time each evening to learn from the tutorials. But then, you know, we have a library. There's. By the way, there's a playlist called ChatGPT. Something. I don't know, we don't have many places. One of them is like ChatGPT beginner course or something like that. All of my favorite prompt engineering videos. So if you feel like Jamie and you want to, like, brush up on, like, some free videos, those are the best ones. They're in a playlist on the channel, so I guess this could be done. Although we don't upload daily, Jamie also downloads some of the ebooks and templates Igor offers. Perfect. Okay. Yes. Signs up to the free newsletter and gets that template that everyone should have. Awesome, consistent, high quality and relevance of our content. Okay. Sorry if this is getting a little too long winded. I'm like, super engaged and interested in this because this stuff is useful, you know, if you set this up to account for your personal context and the user journey that maybe you're looking for a person to go through. This is great stuff. This is for generative AI YouTuber. And you know what? I'll keep this around. I'll finish reading this later on. I don't think this is, like, too interesting for all of you. Hey, Reda. What's up, man? Good to have you here. Good to have you here. But yeah, I just wanted to point out, like, how relevant is it? I wanted to demonstrate it, like, how well this works and I got so sucked into it that I forgot that I was like kind of streaming just like learning about my users journeys. Amazing. Okay. Oh, Rene is just catching up on the images with the, with the, with the image generator in the beginning kind of depict weird timing or. I fell for the Twilight Zone. Yeah, you definitely in the Twilight Zone. Rene, just wait until you. We go through the escape room later on. Okay. But yeah, there you go. So user journeys. Now you can do this for everything. That's my point. You can, you can do this or you can be like, hey, you know, I am building SaaS that helps people, you know, write better. I don't know, like brownie recipes with AI. Yeah, that's like, that's. This is what the world needs. Please. If. If this doesn't exist, someone build it and then it creates the user journey for somebody. Right, Emma. And then a home baker who loves experimenting. And then like you can this really. Even if you're high in empathy, empathy, this really helps with kind of getting a feeling for what the user is going through and what he's feeling at every stage of your user journey. Great stuff. Good. Moving on to the next one. This. We did spend a bit of time on this, but I thought this was particularly interesting, you know, and particularly useful to all you entrepreneurially and entrepreneurially minded individuals here. Yeah. Okay, What else? Justin Sullivan with his garage sale GPT. Heck yeah. Look at this one. So this is the last kind of like GPT that I want to show off here. Then we have a few more. One, two. We have.