# ChatGPT WORK desde CERO: Tutorial COMPLETO + Casos de Uso REALES

Channel: Julio Flores
Video: https://www.youtube.com/watch?v=2fh6uhAlOso
Duration: 44 min
Language: English (United States)
Words: 8617
Transcript page: https://viewrankai.com/tools/youtube-transcript/2fh6uhAlOso

---

[0:00] Hello, hello, how are you? How is everyone doing? Welcome to a new video tutorial. Today, I'm here to show you and explain from scratch what CH GPTW is, how it works, how we can install it , how it operates, what functions it has, and its potential. First of all, I should mention that CH GPT Work is a new tool released by OpenAI that has only been out for a few weeks, and I've been working with it for a while, researching and getting the most out of it, and it's a brutal tool. You have to think of HGPTW as a coworker, like a personal assistant. This new tool is

[0:32] capable of organizing certain folders we have on our computer. It can change, review, or modify certain documents we have here locally on our computer. The good thing is that it also works in the cloud, meaning we can use it as a coworker that can handle certain team projects or research tasks. It can compare flights, even book rooms on Booking for us, it can create certain visual documents like guides or dashboards; we can do so many things with it. And in today's video, that's exactly what I'm here to explain. What is CH GPTW? How does it work? And how can you, starting from scratch, get the most out of it, whether for your

[1:10] day-to-day life or to use for your work ? You already know that I specialize in selling AI services. I have an AI agency, and honestly, since I started using it, it's been helping me a lot, especially in saving time. I use it like a coworker; I give it certain tasks that take me a lot of time, and it does them on its own, with an incredible result and in very little time. So, let's get started with the video. It’s going to be a bit of a long video, I think, because I want to make a super complete tutorial for all of you, so that those watching this

[1:42] video will be practically experts by the end and can use this tool to its full potential. What you are currently seeing on the screen is the desktop application. I recommend you download the desktop application because it works much better, plus you will have —or rather, CHGPT, or CHGPT Work, our agent—access to all the folders and documents we want to grant it access to on our computer. Therefore, I think it is much better and more efficient if we work with it locally, that is, by downloading it directly to our computer . How can we download it? Well, it's very simple. Simply by entering the platform, the official OpenAI tool,

[2:20] searching for "ChatGPT download," the main website will appear. I always recommend this whenever a new tool is released, okay? I always tell my students in the pro community to visit the official page and take a look at everything it offers, what it's for, and what features it has. It’s really where it will be best explained because , obviously, these companies want users like us to use them, and they’ll give

[2:44] you a quite good and complete summary. Simply what I recommend, as I was saying, is to download it for Windows. It is very important that you download it. If you work with Windows, for Windows. If you work with Linux, then Linux. Or if you work with iOS, then iOS. Okay, you download it to your computer. You will also be able to use it from the ChatGPT website, but I think it makes much more sense to work with it directly by installing it on our computer, and during the video, you will understand why I say this. You just click download, and this application will download, okay? You

[3:15] have to enter your account, your credentials, and you will have access. They have done it very well too because OpenAI has modified and redesigned its desktop application, okay? We have this application here and we can move between Chat or Word. That is, we can click here, we have the conventional ChatGPT, the one everyone uses to ask anything or have it write information, but right here in the little window we have Word, and here the new AI agent that OpenAI has developed comes into play, which is exactly what we are going to see today. It is also interesting because from here we can switch directly to Codex to work on or

[3:53] develop any project we want to do. You already know that with Codex we can develop websites, CRMs, or any type of digital product we want to create without needing to know how to code, okay? It is a no-code tool that is also very powerful, and I am currently using it even more than Claude Code, which is the one I used previously, okay? But today we are going to focus on presenting what ChatGPT Word is and explaining everything it has and everything you can do once you download it, okay? This is the main view, this is where you will see and where you will be able to do everything with

[4:31] ChatGPT Work. Directly here we see our chats, okay? They will keep appearing and grouping themselves right here, okay? And here we have the main view. One important thing, okay, is that to use it, you have to have a, okay, paid plan. I mean, I think it only works starting from the Plus plan, okay? If you have, for example, the Go plan, I don't think it will work. The ideal thing is that you have the, the Plus

[4:59] plan, okay? So, let's go back here. Honestly, I can tell you that if you are going to use it professionally or for, uh, daily use, so to speak, I really recommend you have the Plus plan , because I think I pay $ a month, I believe it is, and I have access to all of OpenAI's tools. That is, I have access to ChatGPT, I have access to ChatGPT Work, and I have access to Codex, which you already know, as I said at the beginning, I dedicate myself to developing products and services for companies and I use Codex to do it, okay? In fact, I still pay

[5:33] for Claude, I still pay for the Anthropic tool, but it is true that I use it less and less; but well, at the end of the day, I have it there because in some video I will explain how we can work with both and get the most out of both, but let's focus on this, on this

[5:47] ChatGPT Work tool. Now, let's begin. What we see here is the chat, meaning this works very similarly to, uh, ChatGPT. I can ask it anything here. For example, find me information about such and such and write down the most important things you found. I can start asking it things here and it will respond and do the tasks that I ask of it. Here it might also be very similar to ChatGPT, but let's look at the differences that Work has and why this tool is so, uh, surprising and why it is so useful, okay? Things we should

[6:21] keep in mind before we start working. Here in the permissions section, okay? We have different permissions. If, for example, we set it to request approval, anything we ask, uh, CHGPTW to do, for example: "Go into the invoices folder I have on my computer"—which we'll do now— "and sort them by vendor, by date, or by concept, by price, or whatever you want," okay? Before performing that task, CHGPT Word won't go directly into the folder; it will ask for your permission to let it access that folder so it can carry out the task you ordered. It will keep doing this constantly with everything you instruct it to do. Therefore, you

[6:58] can leave this permission on, but you will always have to be attentive to approve anything our CHGPTW wants to do . The second option, which is the intermediate one, is the best for me. You already know that extremes are never good. Usually, intermediate options tend to be better. Okay? Well, in this case, with the "approve for me" option, what happens is that Chat GPT will work autonomously, okay? But there will be certain things where it will ask for your permission, because this GPTW will notice that it might be some function, some task, some execution it needs to perform, that could put at risk or compromise some information;

[7:34] for example, confidential information, data that no one should know or that shouldn't be shared. So, it will ask for your permission so that you yourself can approve that specific task , okay? Honestly, I recommend that, especially at the beginning, you use one of these two, okay? The "request approval" one might be a lot because, as I say, you'll be constantly paying attention to everything it's doing to, uh, give it permission or not. The " approve for me" one is the one I normally use; when it detects something that, uh, could be potentially insecure , it asks me and I tell it if I want to

[8:05] proceed or if I don't want to proceed. And then we have the option, the last option, okay?,which is full access, okay? Here we give free rein to Chat GPTW so it can access any folder, modify any document, or access any web page I ask it to. These are the three types of permissions. Here I use it; in fact, when you click it, even a warning pops up, okay? And you say, "Hey, are you sure you want to give HGPT full access?" When we are going to, for example, create or do a task that we are really sure won't lead to any kind of information leak, and we won't

[8:40] compromise any important documentation, we can choose this, uh, type of full access permission. But, as I repeat, I honestly advise you to use "approve for me," okay? Then we also have, uh, choosing a project. One thing that is very important and that I always repeat in my community is that when we are going to work on a project, when we are going to do a specific task, for example, if we are talking about developing a website, okay? The ideal is that we create a folder to group all that information, all that agent, so that everything it does is grouped into a single folder and we have it

[9:15] separated by different jobs, different projects. Why? Because if I create a project here and tell it: "Now I start asking it for very different things: hey, look for information for me, compare prices for me to go to Rome tomorrow, uh, then, uh, organize a folder for me, then make me a guide on what type of products sell best in AI

[9:34] agencies, etcetera, etcetera." Everything will be grouped within the same folder and it may happen that when the agent has more information than it should, it starts to ramble a bit, that is, it does not take into account the initial task or the main task that I am asking it for. So, the ideal is that you separate each thing you want to do into different projects. We will also do that now, okay? This is something I'm telling you about on the surface

[9:59] and then we'll go into detail, okay? Then, if we click here on the "plus" section, we can add files or folders, we can even attach a Google Chrome link , we can work on a project, we can set a goal, okay?,which I will also talk about later, what the goal is. Uh, we can also change it or activate plan mode. We can even draw, make a sketch so if we have a graphic idea that we don't know how to explain, well, we can draw a sketch, okay? If you are good at drawing, I'd better not do it because I don't think NichaGPT would be able to

[10:32] understand it. And then we have the topic of add-ons. Basically, the subject of plugins are platforms that we can connect our ChatGPT or ChatGPT-4 to, so that information can go from one side to the other. For example, I can connect my Google Calendar and ask ChatGPT, "Hey, please tell me what tasks I have tomorrow, okay?" or during the week, and give me a brief summary because I don't know exactly what I have. Or, for example, I can connect it to HeyGen and ask it to create a video from an image I've sent to ChatGPT or from an idea I've come up with. I can ask it to draft a new email, send it to

[11:07] someone, summarize my latest emails, or upload a document to Drive. I mean, there are so many, many, many plugins, okay, that we can connect ChatGPT to. And this is a tool, a differential capability that makes this tool even more powerful. Okay? We already know that at first, ChatGPT was just a text tool where I sent text and it responded with another text format answering what I asked, but it didn't have that execution capability. Okay? Now, the difference is that it can do certain

[11:39] things for you that you order it to do. It no longer just works as a text agent ; now it is capable of executing certain tasks, and for that, it's very important that it can connect to different platforms, which are basically the plugins you have here, okay? It doesn't mean that the ones here are the only ones that exist. We will see later that we can add many more, okay? We can actually add any website we want to connect our ChatGPT to, and it doesn't have to be here, okay? We will see all of that later in more detail. Then we have here, like, a history of the conversations we are

[12:12] having, and for now, we'll leave it at that. Lastly, before we start putting test examples here, okay? We select the reasoning model that we are going to choose, okay? Here, for example, we will be able to differentiate ourselves , choosing from GPT-3.5, which is the lowest model that OpenAI currently has in orbit, meaning functioning, up to GPT-4o, which is the most powerful model currently available. I think, I would say, and I am sure of what I am saying, that it is the most powerful artificial intelligence model, okay? We also have to be aware of what we are going to do here, okay? If, for example , we are going to use GPT6 Astra to

[12:53] organize a folder of documents, obviously that’s absurd because any model can do that, okay? And we aren't going to use a model that requires more computing power—meaning, one that consumes more credits—to do a task that any model can really do. So, I really advise you here to think about what you actually need to do before starting any project, what your goal is , and what kind of result you want to achieve to choose a more or less powerful model. Obviously, choosing a more powerful model, as I was saying, means higher credit usage, more computing costs, more effort, and therefore, we could consume more and spend more on a project, okay? So,

[13:33] it’s very important that we work efficiently and choose the right model for the project we want to carry out, okay? For example, I personally tend to use GPT6 Sol or GPT6 Luna, which are like those all-terrain models that work for everything because they have, honestly, quite good performance, consume much less than Astra, and have better performance than 5.5, for example, okay? So, as I say, in the end , it’s about keeping in mind and being clear about what we want to do, what we want to achieve, and which model—which is basically the vehicle —will take us there, okay? We could use a Ferrari, okay? But you know that

[14:10] a Ferrari consumes much more than any other car. Or we could choose a model, an intermediate vehicle that will also take us to the point we want, but perhaps it takes a little longer and the result might not be as spectacular. Okay? I think it’s a good example so you can really understand the difference between models. Then you also have this little bar you see here, which is like the model—or rather, the type of effort it’s going to make . We can choose, for example, GPT 5.6 Sol, and it will have a medium effort level, or I can set it to very high, or I can set it to super high, okay? I

[14:45] think it’s ultra, okay? Here, too, you have to keep in mind what you want to achieve and what you want to do to basically reach the final goal. If we are going to do a job for a specific company where we know we will have a profit or need to do the best job possible, then obviously we can choose the best model, okay? With maximum effort, okay? But you also have to keep in mind that if you have a Plus account , for example, you don't have the Max account—which is the one I have on this account, or maybe I do, actually —you’ll likely run out of credits

[15:19] very quickly, okay? And you'll have to wait a few hours, okay? So, as I said, you need to be aware of what you want to do, what you want to achieve, and which model best suits your goal. Okay? I think we’ve more or less covered everything here. We also have the plugins section here, okay? And as I mentioned, in the chat we had a summary of the plugins that are predefined or come pre-designed within GPT, but we

[15:47] can click here and add many more, okay? All these you see here are models that already have an established connection with ChatGPT that I can use. There are a million of them, honestly there are so many, and some are very interesting, like Stripe, for example. We can create a website and use a payment gateway like Stripe to link it directly in a very simple way. We can use Canva to— imagine asking GPT directly to go into Canva and create a design, a thumbnail for my YouTube channel, or create any kind of presentation for work, for school, anything you want or anything you do with Canva. GPT can do it

[16:28] directly. Okay? Then there are many others, like Zapier, as I mentioned at the beginning; Shopify, if you have an e-commerce store; Slack, if you work with teams; Zoom, for example, if you have a work meeting recorded with Zoom, you can connect it directly to GPT so it can give you a summary of everything that was discussed, okay? As I said, these are connectors, and these are the ones that are pre-selected, so to speak , okay? But we can also add any other one. You know that I can click here, for example, and either create a plugin , go to the marketplace, or upload one, okay? Like a plugin. If I click on

[17:07] create plugin, this skill window appears, and we’ll talk about skills in a moment, to connect any type of platform I want. Okay? One other interesting thing before we start, if I type a forward slash, okay? Or rather, a slash, okay? In Spanish, these appear as commands we can execute directly. Basically, this is like a shortcut that GPT has. The moment I type a slash, all these commands open up, all these

[17:34] functions for me to work with directly. In other words, it's a way to invoke these commands much faster, without having to type it all out, without having to ask for it directly. Right? How can we, for example, use this, or what’s most interesting? Well, for instance, the MCP—basically, MCP stands for Model Context Protocol. Think of it as a way, a solution to connect GPT or GPT Work with any other tool, allowing both platforms to be linked so information flows from one to the other. I can connect "plan mode," which is for when we're working on a project, okay? I ask it to activate " plan mode" so it works much more

[18:18] intensively, drafting all the information in a much deeper way; even if I ask for research work, it performs a "deep research," which is basically researching the documentation I've requested in a much more thorough way. Right? Then we also have "opinion," and later we’ll see how we can use the opinion feature to modify certain things in a very visual way. And basically, it's just shortcuts, okay, to commands we can use or to the built-in skills. Skills are abilities that we are going to give to our ChatGPTW agent. Imagine I want to create—I have an Excel file here with a lot of information and I want to create a visual that summarizes

[19:00] everything my company has captured in the last month, okay? Well, the ideal thing would be to give it a skill that allows it, okay, to create that dashboard visually. It’s like a skill , an ability that will train the agent so it can accomplish certain tasks in a much more defined and reasonable way. Right? I won't go into detail here either because you'll see that with ChatGPT Work, this is usually used more for coding. With ChatGPTW, we really won't need many skills to use it. So, this would be the introduction to ChatGPTW. You've already seen everything you need to know to get started, and in the next part of the

[19:39] video, we'll start using examples that can help you in your day-to-day life. Keep in mind the most important thing, which is to download it; my advice is to save it to your computer's desktop and work directly from there. Okay? We also have the plus icon here, where we can attach files, work on different projects, objectives, plan mode, and draw. Then we have the plugins section, everything we’ve shown, and the type of permissions is also very important, okay? The model type we choose is also very important, and you have to keep in mind that it will depend on the project we want to carry out. It’s something

[20:14] logical, just common sense, but it’s very important because, in the end, you might consume your credits much faster or not be working efficiently. Okay? So , having said that, we’re now going to look at several examples so you understand how we can ask for certain things from GPT, okay? For example, to get a folder, to tell us the tasks we have in the calendar, to create a new task, or to create a file, a guide, or a webpage. We’ll see everything that can be done, and you’ll have that coming up in the next part of the video . So, I think everything has been clear up to this point. It’s been a fairly

[20:51] extensive introduction, but I think it’s worth it so you really understand how absolutely everything works. So now, yes, let’s continue with the video. Alright, after this extensive introduction to the video, let’s put our CHGPTW agent to work. We’re going to ask it to enter a folder where I’ve saved different invoices from different vendors with different prices, different names, and so on. And I’m going to tell it to access my folder directly and sort them

[21:18] by month, year, etcetera. Okay? You’ll see the folder I’ve created, which is this "invoices example," and there are different files, okay? Of, as I said, different types of payments I’ve been making. Well, if I put here directly, hey, check the invoices folder, for example, the content of each file to identify the date and the provider, then rename all files with the format year-month-provider and group them by moving them into their folders organized by month. If I ask for this, okay, it will directly access and organize those folders. How do we do it, okay? To start working on

[21:51] everything within the same project? Well, if I want to group or rather set all this work within the new folder, I am going to click on choose project, okay? And here appear the folders that I have been working on previously. I want to select a new one that I just created and we are going to see how we can add a folder directly. We can create a folder ourselves from here that will then be archived or created on our desktop, but we can also add one that we have previously created. So, in this case, I am going to add the folder that you just saw, where I have the

[22:23] different invoices grouped that I have shown you. We are going to click create project, we trust the file and we are going to give it full access, okay? Why ? Because it is something that doesn't need much information, nor does it need to or will it compromise any type of documentation, okay? They are invoices. So, I am going to give it full access so that it doesn't ask me anything at all. Then, in the model, here I can also change the effort to medium because it is not something that needs to be done with much effort. And let's

[22:52] leave it with, for example, sol, okay? sol medium is more than enough. Okay, in reality we can use anyone for this, for this case, but this would already be a first use case, a first example with which we can use GPTW. You will see that this folder, okay, when it does its work will appear organized in a different way and I am not going to do absolutely anything. So, I am going to send it, okay? Let's see that I am going to send it. It will start to think first to understand what it is that we are asking for. It will see what is the path that it has to choose

[23:22] to be able to do that task that we have ordered and it will start working directly. Here it tells me, "Julio, I will review all the files in the folder , extract each date and provider, and then I will rename them and organize them into their monthly folders." Before returning anything, check the final names to avoid duplicates, okay? so that's very good. It tells me here that it found 11 invoices, all in PDF format, and it's going to use a specific workflow for PDF documents, meaning it's already using its own integrated tools to do the job in the best way possible, okay? In this case, it's going to take, I think, a few

[23:54] minutes to do it, if it even takes a few minutes. And here we also see the notification that we have full access enabled. Okay? In this case, I'm going to dismiss it, because I know we have full access; I chose it myself. But it's good, because often if I create a new project later and I might have this permission set, it will warn me so I really know that ChatGPT has access to everything and doesn't have the option to approve on my behalf. Okay, here it tells me it has identified three

[24:24] vendors: Digi, Booking, and Entropic. There are two Booking invoices issued on September 26th and to overwrite one blah blah blah blah blah. Here it tells me that the files are already renamed; that is, it has changed the names of the files that before...well, it has already organized the folder. But you know that when you download an invoice, often they appear with a strange concept, with a strange name, and here it has organized absolutely everything for me. Okay, here it says the task is ready and tells me, "Hey, there are 11

[24:51] invoices read, identified, and renamed. It has organized them into eight subfolders with a year and month format , meaning it sorted them by month within the year, okay? All PDFs are still readable, no files were left loose, and the two duplicate invoices from September include their invoice numbers to avoid overwriting, meaning there is the same Booking invoice from the same month, okay? But what it did was change the name so there's no confusion. Basically, here it tells me the folder has been organized, which is the name we gave it, and here it knows me and states who I am. Okay? Well then , if we go into the folder now, you'll

[25:26] see it's the same one, okay? because I haven't changed absolutely anything. The documents aren't just lying there anymore; he already organized them for me and now I go straight in here, he even changed the file names. It shows me that this one is from August 2025, from Booking. If I go into another one, for example, this one is from September 2025, also from Booking. Here, this one is from Anthropic in April. In May, well, we also have another from Anthropic, June from Booking; I mean, it has been organizing them, okay? By concept, by provider, and then also by the date of issue. For example, here in

[25:55] August we have three that fall in that same month, meaning they were issued that same month, and it just organized them into the same folder, but naming them after the different providers. Okay? This is a simple case, okay? It wasn't a super organized folder either, because I had just created it. I already have my invoice folders organized, but imagine, I think everyone today has a folder with hundreds and hundreds of unorganized documents, images, invoices, or something that we leave in that folder because we know we aren't going to waste time organizing it. Well, here is a clear example we can use, or a clear use case for ChatGPT Work to do all

[26:31] that work for us. Honestly, it would have taken me—I won't say an hour, but a good while—and that was just for a few invoices, but as you've seen, it did it for me in a minute and a half , so that's a minute and a half we save . Okay? So, that is the first use case of how we can give it access to different folders that we have locally on our computer, and it will be able to enter, read files, organize them, change them, modify them, everything we ask of it, okay? So now let's move on to the second use case, and you'll see that little by little we will increase

[27:04] the work we ask of it, and you'll see that the potential is incredible. Before continuing with the video, I encourage you, as I always do in every video, to join my AI training community , Builders Club Pro, because here we really teach how to get the most out of artificial intelligence; not just for using it in your day-to-day or your work, but also for monetizing it. Here you will see success stories from students who close sales of digital products or services to companies almost daily. For example, just yesterday, one day ago, Jose and Alberto closed a sale together. This is very important and it makes me very

[27:41] excited because as a community, members also collaborate, help each other, resolve doubts, support one another, there's an incredible atmosphere, and as you can see, they even collaborate to close projects worth up to € 4,000 , okay? Alberto, I think, basically had an idea and a client to close, but it was too big for him and he needed someone to help him develop the project . He contacted Jose and they both closed that project for basically € 4,000, okay, a CRM for a company, I believe in construction. That is a

[28:12] success story we have within the group. Then we have, for example, LCER who has closed two websites in just a few days. Jose, who has already made several sales, also closed a CRM a few days ago for a logistics company. Fabi joined and closed a chatbot for almost € 2,000, Hector made one of the largest sales in the community for € 7,200, and these are basically real people who are managing to monetize artificial intelligence. Without a doubt, it is the most scalable business model with the greatest potential currently in the digital world, in the online world. We have the Classroom section here, which is where all the super complete modules

[28:46] are, where you will learn from scratch to develop skills to monetize AI. You have the fundamentals of artificial intelligence, which is the B coding; very complete courses, like for example Cloud Code, so you can go from zero to expert; AI agents, where you have full courses like this one you're seeing on CHGPTW, Codex, Antigravity...I mean, many courses within these different modules. Go High Level, voice agents, which we also have a module for creating voice agents that you can sell to a company to answer calls at any time, and these types of agents are sold for € 2,000 or € 3,000 very easily. And then we also have community

[29:23] resources, a template warehouse, even mindset modules, modules where we teach you how to sell all these services, and even how to deliver AI services. Truly, it is a very, very complete community and we are already almost 80 people from all over the world. Here you can see how they are distributed, people from all over the world who have decided to get trained and have decided to make a change in their life to have more freedom and earn money with AI. So if you also want to learn how to make money with this business model selling services or digital products to companies, I'll leave the link in the

[29:52] description so you can join the waiting list, because it’s usually closed. And on that waiting list, I will notify you of new openings, new free live classes, and all the news I share on my social media. So without further ado, I'll leave the link in the description. If you're interested, click and join, and I'll see you inside. Let's continue with the video now. Well then, we've already seen the first use case. We've seen how it's able to access a folder, modify certain files, organize them for us, and so on. Let’s step it up a notch. In this example, I also want to show you the voice model that ChatGPT

[30:23] has and how we can ask it for things by speaking to it directly. It's no longer necessary to write; you'll see that it understands, comprehends, and it's absolutely wild how much OpenAI has advanced the voice model to communicate with these people, okay? Ultimately, it's a no-code tool, and you know that no-code is basically a tool where we don't have to write code, we don't have to program, we don't have to do absolutely anything. It takes care of everything. We simply give it the instruction in our own words. Before, it was with our own words written in text, and now it also has the function to do it by voice, okay? And honestly,

[31:00] I use it a lot too, and it's incredible how it understands, how it communicates , and how you can interrupt it while it's speaking, okay? And you'll see that it can do many other things too. So, let's go ahead and ask it to access my Google Calendar and give me a brief summary of the upcoming meetings I have , okay? For next week, for example, or the jobs or tasks I have to do or have noted in Google Calendar. The first thing we have to do is go to extensions , obviously, and here in popular ones, it already appears so we can, for

[31:30] example, connect our email directly. Here, I can also ask it to draft an email or give me a summary of the latest received emails. I also have access to Google Drive. I can give it access to my documents so it can download, modify, or upload a new one to Google Drive, whatever I want. And here we have, for example, the one for Google Calendar. If I click here, okay, the option to connect an account already appears, okay? I have one of my accounts connected, and that's where I have all my tasks within the my day-to-day. I have it organized in the

[32:00] Google Calendar of this account. Okay? So, let's go back to the chat, and now we're going to try, as I said, the voice chat. You're going to see that it's absolutely insane, and we're going to ask it directly, hey, to enter my Google Calendar and give me a brief summary of the upcoming tasks for the next few days, okay? Hello, how are you ? Good morning, chat. Look, uh, please, I'm going to ask you for a task. I want you to access my Google Calendar and give me a brief summary of the tasks or appointments or any event I have until next Tuesday, since I don't remember

[32:34] exactly what I had. And please give me that summary

[32:41] while we can keep talking, okay? Because, uh, chat, sorry to interrupt you, but anyway, keep going with your task as I'm making a YouTube video, okay? I don't know if you want to say hi to them or not, just keep doing the task I sent you, and anyway, uh, while you can even talk, you can ask for other tasks that you want, and that's the capability this voice chat has, which honestly, when you start trying it, it's brutal. Okay, here it is already accessing my calendar list and now you're going to see that it's going to give me that full list, okay, of all the tasks I have. Let's wait a few

[33:16] seconds. You see that it's also being transcribed because normally, uh, voice agents or voice chats are, uh, that. In the end, we emit our voice, a voice text, okay? And it takes care of transcribing it, translates it directly to voice, and executes the task. Okay? Here it is already making the summary for me. Okay. Uh, tell me, please explain to me more or less in your own words what I

[33:39] have. Sure, look, between now and Tuesday you have seven things. Friday is the busiest day with training, a lawyer appointment, and onboarding. On Monday you have the gym and another onboarding , and on Tuesday the gym, the meeting with Vicente to finalize the partnership, and on the weekend you have to be witnesses. All right. Could you please add a task for me? I have to go visit a company on

[34:05] Saturday, they asked me to go. Basically, well, it's a real estate agency. Put down" real estate visit "on Saturday, this coming Saturday the 26th . Please set it for 11 in the morning. Let me check it.

[34:29] Okay, it's already creating the event, okay? Not only does it tell you and give you a summary, but you can also modify calendar events, create new tasks or events; we can do so many things, and as you can see, it's done directly through voice. Right? I think this is also an advantage this tool has because we no longer have to type, The real estate visit has been added for Saturday the 26th from 11 to 12 in

[34:50] the calendar. Okay, thank you very much. You're welcome, we can adjust it whenever you want. Well, you've seen another function it has, besides connecting to any tool—I've shown how we can connect to Google Calendar, and the process is basically the same for any other platform—is that we can communicate directly by voice, and have it access that platform, take a look at all the tasks or events we have, in this case, using Google Calendar, and give me a summary. It doesn't just give me the summary in text here, but you've also seen how it reads it out, and I can even interact with it to ask questions or add a new task. So, this

[35:33] is a great example, and you will see that, honestly, I don't think it has any limits; we can do hundreds and hundreds and thousands of things. And honestly, for a person's day-to-day, whether you want to apply it to your work or your personal life (you're seeing this is something personal, as it's my schedule), I think it has incredible potential. So let's keep looking at use cases because I think this is where your imagination comes into play and the limits you want to set for this ChatGPTW model, which is incredible. Another very interesting use case for ChatGPTW is having it browse the internet and using it as a

[36:10] prospecting or research agent. Imagine I want to find different platforms to generate videos or images with artificial intelligence, and I don't want to go in one by one, I don't want to have to register for each of them; I want to know which ones are free or which ones I can use that are more

[36:26] cost-effective and have more utility. Well, here we can also use ChatGPT to give it that internet access so it can search, investigate, obtain all the information we want, and summarize which platform is the best fit. Likewise, this can be adapted to anything. We can search, or we can ask it to look for sectors where we can sell AI services, sectors where we can sell products from our own company. We can ask it to compare flights to a destination, or to compare rooms or hotels on Booking. We can do so many, many things, and you'll see that it really has that ability to browse the internet and basically do what a person

[37:04] can do in front of a screen. So, we're going to do something interesting, and we're going to invoke it, we're going to say," Hey, I'm giving you permission to enter the browser, "okay? If we click here, we are asking it, we are telling it to enter that browser. So, now we are going to ask it, we are going to dictate—this is no longer the voice chat we had before, this is dictation—which will directly transcribe so you don't have to write the entire prompt or the entire instruction I want to give it. So I'm going to tell it the following. Hello, chat. I want you to go on the internet

[37:34] and make a comparison of all the tools currently available to generate images or videos with artificial intelligence. I want you to look for all these platforms, for example, HeyGen or, I don't know, all the ones that exist, okay? I want you to summarize how much each of them costs and if any of them also have free plans, okay? And make a list for me of which ones are the best to use and which are, of course, the cheapest. Okay? I want you to find me a place where I can generate images or videos without spending much money, and if I can spend absolutely nothing, then even better. And please, can you give

[38:10] me a list when you're finished of the tools you've found that best fit what I'm asking for, okay? It's going to start transcribing; we already have the instruction, and now it's going to get to work. In other words, it tells us," Hey, I'm going to research current prices and plans on the official pages. "Since the market is huge, I will make a broad comparison of the main platforms and separate image, video, and all-in-one options, highlighting

[38:33] the free ones and the lowest-cost ones. Here I could, as I say, go into each of these websites; I could search manually , register for each of them, see what it costs me, what the plans are, if any have free credit, if any don't, but here GPT is really going to do it all for me, okay? This would also save me and will save me a lot of time. And as I say, don't just stick to this example . These are simple examples of use that I want you to then adapt to your own needs. If you have, for example, to look for, uh, what can I tell you? Do

[39:04] you want to run an email marketing campaign and do you also want to know which tools allow you to send more mail at a lower cost, you can also ask it. And you're going to see now that I have my little hands here, okay? I'm not doing anything and it has directly opened here on the right side, this that you are seeing here, it has opened the browser, okay? And it is going to start working and investigating and

[39:26] searching for information for us, okay? As I say, GPT has incredible potential and this is just another example of everything we can do. I think it's the last, or second to last one we're going to see. Here you are already seeing that it now enters Midjourney, it's going to see what the plans are, it's going to draft all that information, then it will enter Hisfield, it will go little by little into each of these platforms until it gives us that

[39:47] complete summary of what we want. Okay? Now it enters Runway and you see that I am doing absolutely nothing, that I have my hands here and it is doing all that work for us, okay? So let's wait a few minutes to see how long it takes to do this in-depth search and this comparison we want of platforms, and you're going to see that the result is basically what we would do but in much less time and working automatically. So we wait a few minutes and see if it is really capable of giving us this list and if it is capable of investigating the internet all by itself, which you

[40:19] are already seeing here that it is doing on its own. It says, well, I was going to cut, but it says," Important data has already been confirmed. Midjourney now includes image and video from € 10 a month, and Runway keeps a free trial of 125 credits, but they are one-time, not monthly credits. I'm still comparing the most affordable alternatives. Well, uh, let's wait now and see if it's able to, uh, make this comparison for us and see which one we are more interested in using. It’s honestly flying because it’s going from one website to another quite fast, but well, I’ll cut for a second until it finishes and we’ll see how the

[40:51] result turned out. Well, it has already done the comparison we asked for. It went in, it, uh, searched the internet, it accessed, I think, over 20 tools, okay, for image and video generation with artificial intelligence, and it made an extensive table on which one best adapts to what I asked for. Here it tells me that we have image and design tools, it tells me if it has a free plan, if it has monthly credits or not, approximate price, and for each one, it highlights what the best feature is. For example, here it says, "Hey, Microsoft Designer is the simplest option to start for free." Uh, with Gemini it says, "Image and video,

[41:24] another tool from Google." Uh, it makes absolutely like an outline, a table with all the information that you asked for, okay? Then it also tells you, hey, platforms that generate image and video . It tells you all these generators specialized in video. Well, it mentions Kling AI, HeyGen, Synthesia, Pixverse and says, "It has recurring free credits." It has three videos a month up to one minute. It tells us the prices, also the best of each one, and finally it makes like a summary more or less of, uh, what would be the best apps or the best tools to work with according to what I asked. Okay, here

[41:57] it says, "Hey, well, my ranking for you would be Microsoft Designer, Leonardo AI, and Ideogram." The cheapest paid ones, well, would be this, this, and this. My personal recommendation for you would be to start with this, this, and this. So, it has done the job perfectly. This would have taken me, honestly, quite a lot of time, surely a few hours to enter each one of them, write down how much it costs, if it has credits or not, see the different plans , the subscriptions, which ones also have, uh, what each thing is good or not. And here I haven't even really done it, I don't know how long it took,

[42:28] but I think it took a minute or maybe eight minutes to do, which is practically nothing. And that concludes the tutorial, so to speak, or the use cases. I'll surely make another video because there are so many more use cases, but I don't want the video to be super long. You now have the complete tutorial on everything you can do with ChatGPT Work from scratch, and honestly , I think it's a very, very, very useful, efficient tool, and you'll surely be blown away once you start using it because it truly has no limits and has incredible potential. So that covers the different use cases, and I

[43:03] hope you liked the tutorial, which I think turned out quite well so you can start testing it and using it, whether for your work or your day-to-day life. And that's it for today's video. I hope you liked it and that it brought you value. It was a tutorial that was honestly quite complete and quite extensive. I think we explained from scratch how to get the most out of it, the maximum potential of ChatGPT Work, which is an amazing tool. And if you really want to learn how to use these types of tools to earn money with AI, to create high-value digital products, to be able to sell the business for, uh

[43:34] , 2, 3, 4, 5, up to 7,000 euros, as you've seen in the Pro community video, from those real cases, those real people who are getting results and are managing to monetize...I'll leave the link for you to join the waiting list for my Pro community, as we will soon open the doors again so that many of you can access it and start training properly in artificial intelligence. So , with that, I'll say goodbye. If you liked the video, I would really appreciate a like, a subscription, a comment telling me what you thought, and also any use cases you can think of , because as I've repeated several

[44:07] times in the video, I believe ChatGPT Work has incredible potential. So, without further ado, thank you so much for watching until the end, and we'll see you in the next video.

---

About this transcript
Read from YouTube's own caption track and laid out by ViewRank AI (https://viewrankai.com).
ViewRank AI finds the videos already beating a creator's own average on Instagram, TikTok and YouTube Shorts, transcribes them from the audio itself in more than 60 languages, and turns what worked into new ideas and scripts.
Free transcript tools, no account needed: https://viewrankai.com/tools
How to read any video this way: https://viewrankai.com/llms.txt