Usa ChatGPT como un PRO (3 plantillas + método)

Ninety percent of people who use Chat GPT get very, very bad answers, and at best they are mediocre and generic. The thing is that you ask Chat GPT for things and it returns boring texts, soulless ideas that seem to have been written by robot from the 90s. And the result is that you get frustrated and think that this AI thing is not that great. Today I am going to show you the exact method that I use to turn ChatGPT or any artificial intelligence into what I call the infinite intern. A 247 assistant that will help you with everything you need and can save you up to 20 hours a week. This method is based on the official guides of Those who really OpenAI, Gemini and Tropic. It is a reliable, solid method that has helped me go from getting mediocre and really bad answers to getting answers from a true professional. Stay until the end because we are going to see the biggest mistakes and how to fix them. We are going to debunk a very widespread myth. Let's look at three tricks that you need to know to get the best answers. And at the end of everything, we're going to look at three complete prompts that you'll be able to copy and paste and put into practice right now. So I won't go on any longer. And let's get started. The main problem is simple. You treat ChatGPT or any artificial intelligence as if it were Google. You ask it a question and expect a magic answer. And that doesn't work that way. Using Chat GPT in this way is like having a Thermomix at home and only using it to blend food. The Thermomix can do many other things. So what we need to do is change the way we think. We're going to stop using it as a search engine and start using it as an Infinite Intern. Think about it. This intern is an absolute genius. He's read all the content on the Internet and has unlimited work capacity. But he does have some limitations and it's very important that we take them into account. And it is that it has no initiative. It knows nothing about you or your business. And if you do not give it clear instructions, it will use the law of last effort and give you mediocre answers. Your job is not to ask it, it is to direct it and to properly direct your Infinite Intern. We are going to start by dismantling the first mistake. Let's see what that mistake is and how we can solve it. The first mistake is to think that the AI knows from what perspective it should speak. And no, it does not. The mistake is not giving it a roll, not telling it what expert it is, how it should behave. And here I want to clarify a debate, and that is that you have heard out there. There are many people who say that with these new models of artificial intelligence, that reason, that think it is no longer necessary to give it that role. And that is half true, half false. In other words, if we ask it a very simple question, for example, what is sweeter, a banana or a lemon, for example. In this case, we do not need to give the artificial intelligence a role, but if we want artificial intelligence to give us quality and detailed answers on a specific topic, it is necessary to give it this role. And it is very easy to understand when we are faced with an artificial intelligence model with so much information, giving it this role lets it know where to look among all that information. This is as if we went to a hospital because we have heart problems and thought that any doctor could operate on us. After all, they are doctors and they are all smart. And we all know that this is not the case, but rather that we need a surgeon who specializes in cardiovascular problems. The exact same thing happens with artificial intelligence. And now we are going to see it with an example. Look, we are going to go to a new window, to a new GPT chat or any artificial intelligence you want. And we are going to tell it this. Write a post for Instagram about sustainable sneakers. We are going to send it to it, and its response, as you will see, is going to be a fairly generic post. Perfect. Look, your steps also count. Every day you travel miles to work, to the gym, to that get together you almost canceled. Can you imagine that each step will also help the planet? Sustainable sneakers aren't just fashion. They reduce the environmental impact. Okay, so here we have a fairly generic post. It must be said that it's also quite correct. But now we're going to do the same thing, but giving it a specific role. So we're going to open a new chat and we're going to say you're a copywriter that is an expert in sales or persuasive writing, specializing in sustainable fashion. You're addressing conscious millennials. You speak clearly without exaggerated promises and with a touch of urgency. So we're going to send them this. And look here, they're already starting to ask us questions. Perfect, I got it. Give me the specific product or approach. So we're going to say sustainable sneakers. Perfect. So you don't need 20 pairs of sneakers. You need areas that are durable. That respect the planet and that won't go out of style in two months. These are made with recycled materials and local production. Less footprint, more life. Change is not preached, it is stepped on and stepped on. Are you going to keep looking at them or are you going to wear them? Notice that the quality of this post has improved considerably and we had also asked him to give it that touch of urgency and he has fulfilled it perfectly. Therefore, if I want you to take away something from this first mistake, it is that whenever you want a more detailed, better quality, more specific answer, use the role to improve your proms. And now we move on to mistake number two. The second mistake is asking for things and not giving them the full playing field. Your intern knows absolutely nothing about you or your project and needs you to give them context. Lets see this with an example. Let's go back to a GPT chat window and in this case we are going to use as an example a brand of organic cotton T shirts. Okay, so in the context we are going to tell it we are a brand of organic cotton T shirts made in Spain. Our ideal client values quality and sustainability over price. We are competing against fast fashion and our goal is to launch a social media campaign to increase online sales by 15% next month, educating people about the value of local production. In this way artificial intelligence knows what our context is, that is who we are and what we want to achieve. We are going to send this to artificial intelligence and you will see that it will give us a much more complete answer. Note that it tells us the central idea of the campaign. The common thread should be cheap as expensive, local lasts longer, fast fashion wears out, pollutes and exploits. Your thing is quality, sustainability and local pride. So it gives us some key messages. Quality and durability. A T shirt that lasts five years versus one that lasts five washes real sustainability, less carbon footprint because it is produced in Spain with organic cotton. Then the local purchases that generate employment in your area, not in factories without labor rights and price versus value. Investing in a T shirt is saving in the medium term. Then we have the axes of content for networks. He is already talking about educational content, another more emotional one, a more visually striking one, social proof, and then the calls to action that must be direct. In addition, he gives us specific actions in four weeks and closing strategies to lead to the purchase of those T shirts. And he also tells us what to measure. Note that by giving him our context and the objective, the quality of the response we receive is no longer something generic and mediocre, but rather he has taken the time to be able to assess our situation and in that way give us answers that perhaps we cannot take full advantage of, but where most of it is very useful. In addition, by doing this we avoid the blank page syndrome, that of putting ourselves there and not knowing where to start. The third mistake is giving him wishes and not tasks. And be careful, because here we are not talking about objectives, as we did in the previous mistake where we told him that our goal was to increase the percentage of sales. But here we are telling him, let's give a specific task. Look, let's see it with an example. We're going to go to the same chat that we had open for the previous mistake using the example of selling T shirts. So here we're going to give a specific task, which is to create a content calendar for Instagram for the next seven days that drives traffic to the new collection on the website. Each post must have a visual hook that works in the first three seconds in a single clear call to action. A call to action is a call to action. And look here, I'm putting this here as a way of thinking, but I want you to notice that this task is specific, measurable and unambiguous. In this way, we've given them a work order and not just a wish. The next mistake is not asking for a format. Almost no one does this and it really has an incredible impact because instead of having a huge text where we don't know exactly where the information is, we're going to have an organized table with what we want. Look here. Using exactly what the same chat we did did, and we didn't give them a wish. Previously, we had our task up here and now we're going to add the template or format. It doesn't matter. We're going to tell it. Give me the calendar in a table with these columns. One column will be the day, another will be the main idea. Another column will be the full text of the post. Another will be the idea for the visual and finally the call to action. So we're going to send it these two parts of our prom again, which would be the task in this format. Note that now it's going to give us that content calendar in this table format. On one side we have Monday the main idea. Note fast fashion versus local. The full text of the hook post. What would happen if your shirt could talk? A fast fashion T shirt would say, I made it in two hours in a factory without royalties. I'll last five washes. Ours were sewn in Spain with Organic cotton and I'm ready to accompany you for years. When you buy local, you're not just buying clothes, you're buying a cleaner future. Not bad. It could improve some of the parts, but not bad. Note that in the next column we have the idea for the visual. It tells us recommends a horrible video comparing two T shirts on a table, one worn and the other impeccable. After several washes, text on the screen with the phrases and at the end it gives us that idea of a call to action. Discover the new collection made in Spain. It has done the same as we asked it to do on Tuesday, Wednesday, Thursday, Friday, Saturday and Sunday. In this way, we have not only told it what kind of expert it should be, we have given it a role. We have also told it what our context is, that is who we are. What our objective is to increase sales. As well, we have told it what task it should perform and we have also told it how we want it to deliver that task. And now let's move on to the last part. The last mistake. But first I want to remind you that if you like this type of content, don't forget to subscribe to the channel. Also, if you want to stay informed and have much more detailed information, I'll leave the QR code here so you can subscribe to my newsletter. And now let's move on to that last mistake and those three tips that you absolutely need to know. Okay, the last mistake is what I call the checkup, the quality control. And that is that we must tell the artificial intelligence to perform a self audit. So before sending our prom, I want you to make sure from now on that you add these parts that will help the artificial intelligence control its response on its own. Look here we would tell it to do this checkup, that each post has a powerful hook, that there is only one call to action per post, that the tone is consistent with the brand without exaggerated promises, and that the language is appropriate for conscious millennials. So let's do it. We would send this from here and here it will go point by point. Each post has a powerful hook. Yes. Then it tells us that we could reinforce the hook from Friday a little and suggests an improvement. Then it will verify if there is only one call to action. And therefore here it tells us that yes, each one closes with a single and clear call to action. Then it will verify if the tone is consistent with the brand and says yes, that there are no phrases like you will change the world with a T shirt or inflated promises. This is wonderful. And then the Language is appropriate for conscious millennials and says yes, use a clear, direct tone and with references that appeal to social and environmental awareness, local pride, minimalism and durability and smart consumption price versus value. So in this way we know how to solve the five main mistakes that the vast majority of people make when speaking, communicating and using not only chat, GPT but the rest of artificial intelligences. But we are going to go a step further and we are going to see those tricks that will further increase the quality of your proms. The first really professional trick is to use meta prompting. This is very simple and it involves using artificial intelligence to improve your proms. Look, what we are going to do is the following. We are going to open a new window of a new ChatGPT, chat or any artificial intelligence. Remember that you can use what we are seeing in any of them and we are going to do the following. We are going to take all these instructions that we have given it for our T shirt brand and we are going to tell it the following. Analyze this prompt that I have given you and tell me how you would improve it. To obtain an even better result. We send this to it and it will tell us itself what points we could improve. Look, your prompt has context, objective, task, format, verification criteria, but if what you are looking for is an even more refined result, I will point out the points for improvement. So it asks us for more precision in the audience to add style restrictions. If we want to force a variability of formats because it can return seven posts that are all the same, then we can specify if we want at least two reels, two carousels, two stories, then a direct connection to the sales objective. Note that it tells us that the PROM asks for traffic, but it does not ensure that there is progressive storytelling towards conversion. So we could add, add and it gives us two points that we could add to that prompt. Then clarity in the length of the delivery and then indications for visual creativity. Note that in this way you do not have to worry too much about your prompt not being very good, but rather we are going to follow a very solid base that we have just seen. And then we are going to use this meta prompting. We are going to use the artificial intelligence itself to help us improve our instructions. And now let's move on to the second professional tip. The second professional tip is that from now on I want you to stop making Frankenstein prompts. That is opening a GPT chat window or any artificial intelligence and asking for everything in the same one because you are confusing it. If you have a complex task Divide it into parts and ask for it in different windows. This will allow it to focus on doing one well. The only thing is that later you can have that project build that project much better without making so many mistakes. And the third and final pro tip is that I don't want you to get tired of iterating. In other words, if you send it an instruction and it doesn't give you the result you're looking for the first time, don't despair, don't get frustrated. It's okay. The crux of the matter, the master magic trick here is as simple as iterating. Ask it, tell it what it's doing wrong and how you expect the answer, what it can improve. That way it will be able to improve that answer and get closer and closer to what you want. And now we're going to finish everything off with three final prompts where I'm going to show you very quickly what most people usually do and what we're going to do from now on. I'll also leave those prompts on the screen so you can see them. Take screenshots if you want, and that way you can copy them and use them today. The first prompt is the following. This is what most people would do. Give me ideas for posts for a nutritionist and we would get a fairly generic response. What we are going to do is this. You are an expert in digital marketing for health. This will be our role. My client is a nutritionist who helps women over 40 lose weight without dieting. This will be our context. The goal is to capture leads for her new master class. The task is to create five ideas for reels. The format, give them to me in a table with columns for hook, suggested music and call to action. And in the check in, make sure the tone is empowering and does not talk about miracle diets. This would be the first prompt and I guarantee, as you can see, that we are going to get a much higher response. So, you know, stick with this prompt and let's move on to the next one. The next example we are going to do is to give us a summary of a report. Okay, here I have uploaded a report about AI in the world in the Hispanic market. And here most people are, I would say give me a summary of this document. And here they would stay. Okay, so we will have a summary of a document that will probably be quite good, but it will have a lot of room for improvement. And there are other ways that we can avoid mistakes using the method that we have seen. So instead of doing that, what we are going to do is upload our document and do the following. Imagine that we upload a report. In this case I had an example with a financial report. But we are going to tell it. You are an artificial intelligence analyst specializing in the Spanish speaking market. We are going to remove this part here and this would be our role. You are an AI analyst specializing in the Spanish speaking market. This is our report and I will present it to them. This can be real that we present it to investors next week. This would be our context. The objective is to convince them to invest and see we can add to it and see the importance of starting to use and incorporate AI in our company. Then their task we are going to tell them is to identify three key strengths, two hidden risks and one growth opportunity that we do not let see. The format is to present it to us in five clear and direct bullet points with simple language. For non experts, the bullet point would be a list with the main points and then a check where the use of jargon is avoided. In this case, complex technical or technological jargon. So this would be an example of a prompt to make a good analysis and summary of any type of report or document. And finally, let's go with our last example. In this example, we are going to prepare a romantic weekend or four romantic days in Paris. And we are going to tell it prepare a four day itinerary around Paris. This is what the vast majority of people would do. So here it will probably give us the typical things. Okay, look, the lube can't go wrong. Probably the Eiffel Tower over here, it's very good. Here the Orsai Museum. That is to say, it will give us the most typical things. But now we are going to do it as we have really learned that it has to be done. And we are going to tell it the following. We are going to tell it act as a local Parisian tour guide, an expert in experiences authentic. This is going to be our role. We are a couple in our 30s traveling for four days in May. We hate tourist traps. We love art, amazing food and aimless wandering. Medium budget. This will be our context. The task is to create an itinerary and format it in a table day by day, separating morning, afternoon and evening. And suggest a non touristy restaurant for each dinner. And in the check in we are going to tell it to prioritize plans that avoid the most extreme crowds. In this way we are now going to receive look a complete table where it is going to give us these destinations, these slightly more elaborate activities directly for example, it suggests Monk Mir in Bohemia, Rifkosh in Secret Art, Contemporary Art and Maje a Belleville. So where before it was talking about typical things like the Eiffel Tower or the Musee du Louf? What it does here is give us slightly more alternative options and away from all the tourism. Okay, we have just seen a very simple method that you can use to improve the instructions you give to artificial intelligence and therefore you will improve the quality of the answers it gives you. We have also just seen three tricks that you can implement that you can and should implement right now to also improve those answers. And we have also seen three practical examples. I hope that all this information is very useful to you. Remember that in the description I leave you the link so that you can subscribe to my newsletter. Let me know in the comments if you put any of these things into practice. If you are already using any of them. I will not ramble on any longer and I will see you in the next video.