--- title: 'Video 0XGh3kaj7lc' source: 'https://youtube.com/watch?v=0XGh3kaj7lc' video_id: '0XGh3kaj7lc' date: 2026-07-04 duration_sec: 0 --- # Video 0XGh3kaj7lc > Source: [Video 0XGh3kaj7lc](https://youtube.com/watch?v=0XGh3kaj7lc) ## Summary ## Transcript YouTube using artificial intelligence, this video is the most comprehensive you'll find. In the next few minutes I'm going to set up an automated YouTube channel from scratch [sigh] live using artificial intelligence [music]. I'm going to show you every step of the process, from [music] video, using the same tools I use to generate over channels. And I want to emphasize something, this is not theory. I've been doing this for over 3 years and today I manage more than 15 [music] channels. What you're going to see in this video is exactly what I do, with the good and the bad first point and why YouTube automation with AI this [music] 2026. Before we get into the process, Artificial intelligence has changed the rules of the game in YouTube's [music] automation. Previously, to set up a channel without showing your face, you needed to hire a scriptwriter, a voice actor, a thumbnail artist, and an editor—that is, four people per video. And that [music] added up. Now, with tools like Clou, Eleven Labs, and Gemini, AI covers three of the four pillars of content creation: scripts, voiceovers [music], needs a person and depends on the niche, you don't even need it, is editing. And that means you can start a channel with less than $100. I repeat, [music] yes, less than $100. This was impossible before, but be aware, and this is super important, artificial intelligence lowers costs and saves you time, but it doesn't think for you. The choice of niche, strategy, packaging, that still depends on you. AI is a tool, not a business. You build the business yourself by leveraging AI, not by delegating all of that to AI, which, mind you, is not the same thing. Therefore, the first and most important step is to choose a niche good. So let's get to it. Look, the niche is 90% of the results we're going to get in YouTube [music] automation. I repeat, 90%. It doesn't matter how good your videos are, it doesn't matter how good the software you use is. If the niche has no demand or is saturated, you're not going to make money from this. It's that simple. It's no use having the best videos in the world if the niche doesn't work. And I'm sure you're wondering, then, how do I choose a niche? I look for four very specific [musical] characteristics. The first is that there is a recent channel less than 3 months old that is growing rapidly. That tells you that the format works right now, not 2 or 3 years ago. The second [music] is that nobody is doing it in your market. In other words, if you find an English-language channel that's nobody has done it in Spanish, you have a get rid of all the other people and you take all the [music] characteristic is that there is never any competition. The third is that the production cost is low, because if the niche demands 3D animations or complicated productions, the cost [of music] skyrockets and it ceases to be viable to begin with. And the fourth characteristic is that the Niche has a good RPM, because simply having a good RPM makes the Niche much more scalable, because the difference between a channel with an RPM of 10 with [music] 500,000 visits would be generating €500 per month and with that channel that has an RPM of 10 you go from generating €500 per month or $00 per month to $5,000 per month. I look at two barriers to entry. that the reference channel has many more [music] indicates that YouTube is pushing the content to new people and that there is no competition or very little in your market. If those two things are true, you have a potential niche. And [music] here I want to show you something that is super important and that works very, very well. It's called Nietzsche Bending. Basically, it's taking a format that's already working in one niche and applying it to another market where the audience hasn't seen it yet. In other words, you don't copy the channel, you copy the structure and apply it to another topic. To cat psychology channel that's really popular with a specific video format, you can take that same format and apply it to dogs, health, format works because [music] is catchy, and if you take it to an audience that hasn't seen it, the chances of it working are probably very Now I'm going to show you how I do that process live. The first thing I do is go to this artificial intelligence here, the cloud. And we have to understand that to find potential niches, we must first bring those keywords down to earth niches. Because you don't go to YouTube and say, "Hey YouTube, give me potential niches to succeed on YouTube." No, what you need to do is enter that channels appear and from there identify those potential niches. So, I'm going to give you an example: give me English words to find potential faceless channels [music] of religion. I, for example, put as an example, but you could use sports, spirituality, finance, technology, whatever you want. And here, if you look closely, Claud already tells us that search for potential faeless channels in the religion niche. And here he gives us several examples. And the trick here is to put them on one by one. Something else I recommend, which we'll do here, is to open a new . Because? Because if you actually put these keywords on your own YouTube channel, in the end YouTube knows what you like, the algorithm has you figured out, and no matter how many keywords you sometimes use want to see. What's happening? because it already knows what you like, and that's why we do it in the incognito evaluation. So, I'm going to copy this first keyword here and important that you don't register in the incognito window because otherwise you'll already be logging into your account. Let's stick this here. We're going to filter by filters for [music] over 20 minutes. Because? Because automated channels, especially religious ones—I know this from experience—the videos are generally quite long, or at least about 30 specify here that the videos I want to see are at least 20 minutes long, then [music] the probability of getting automated channels, higher. I want you to see, this is an automated channel. This is a channel, no, this is not an automated channel, this is not a channel. Okay, we see that some older channels are appearing , let's select this year, okay? Because remember, the channels are, the better the music is. We see that, for example, this channel from 9 months ago, this is an automated channel, okay? Full documentary, 800,000 views, 5 months, [music] 369,000 views. Okay, this is a fully automated channel so you can see person appears, and notice that we could have found a perfectly potential channel here subscribers and with videos from 5 months ago with millions of views. Notice that the process I just went through to find potential niches is not people [in music] don't have the knowledge and they complicate everything, when we actually get the best results in this simple. [music] In this case, if we wanted to search for niches and niches, don't but put in all possible keywords to find all the there filter and those that best fit with told you at the beginning, you keep it. Another way to find potential niches is with this Viral [music] tool. Miral is a tool that specializes in finding potential niches and work, I'm going to tell you exactly [music] what I've told Claude to find potential Faceless religion channels. I'm going to [music] press enter here . It automatically generates the keywords for me. I press I put the filters here and put, for example, lengths of more than 20 minutes. and that's the great thing about this tool, isn't it? For example, maximum channel size to 350,000 [music] subscribers. What's happening? That filters, and if we set the maximum number of subscribers we want to be on a channel , it means that Because? Because the channels that are performing very well recently are millions of visits. Let's switch to English here. We're going to include the last 12 months and we're going to include videos that are at least, for example, 12 minutes long. Okay? This remove one of the sources, which, well, we have told it that we don't want them, and in this way we guarantee that more long videos will appear. We apply filters here and automatically, if you notice, all these are automated channels, channels that do not rely on any person. If you look closely, channels with, for example, 280,000 subscribers, this video has 2.8 analyzed, but it's probably a potential channel. Look at this one, okay? It's a religious channel, in this case Islamic religion, with less than 10,000 subscribers, in this case 5,500 [music] subscribers and videos from months ago with more than 400,000 views. Remember that we don't earn money views [music] we get on our videos. It doesn't matter if the channel has 5,000 or 100,000 subscribers. We see this other one from 9 months ago, 15,000 subscribers, videos with 400,000 views. This tool works very well for finding potential niches, and I am personally one of the consultants for this tool. write the word viral to my Instagram, which you'll see here on the screen, I'm going to give you a free trial so you can try the tool potential niches [music] and to boost your YouTube channel. Once we've seen the step of finding niches, let's move on to the next one, which is finding [music] ideas that work. The biggest mistake people make when is thinking [music] has to be original, and it really doesn't. You have to do what works. Always do more of what works, not what you like. What works is already out there. You simply have to know how to find [music]. reference channel and look for the outliers. What is an Outlier? It's a video that has many more views than the channel's average. In other words, if a per video and there is one that has 200,000 [music] , that is an outlier. That video worked for a reason, and what I want is to understand why it worked and replicate that structure [music] with my channel. We're going to do it live. Take a look at this earlier. We see that all of these are outliers. Because? Because the most recent [music] videos don't really have that many thousands of subscribers. But if I go to the most popular ones, we see that this [music] is a brutal outlier, right? It has 2.2 million views with a channel of less than 100,000 subscribers. All the videos, or rather, the most popular videos, the vast majority, exceed 100,000 views. And that tells us that [music] these videos don't come from the existing audience, but from a [music] So, how would I copy the process? [music] has been done exactly in Spanish, as I mentioned at the do things that have already been done before, and if it hasn't been done, I would replicate the exact same pattern, replicate the same idea, package it the same way, but simply take it to a different market. Because? Because generally, 99% of the time, what works in one market ends up working in another. And I'm not talking about English to Spanish [music], but I can talk about English to Polish, French, Italian. And I'm telling you this because I've had many channels and those channels generate income for me from those channels and those markets in the United States . So, that would be exactly the process I would follow. Above all, what I would also do is analyze, okay, if the videos are an hour long, then at least make those videos an hour long. Because? Because those results are coming from all the actions this channel has taken exactly like that, I wouldn't complicate things. I would do the ideas, the duration, and the titles that way too. As I said, we need to keep things simple. Something else you can do if you sometimes want to do more of what works and you don't have a clear idea of ​​how to do it is to copy the titles of each video, go to Clou, the artificial intelligence tool we saw earlier, and ask it to give you a similar idea, but changing small things. Because, as I've said throughout this video, and as I'm works, but you always have to try to add new and differentiating value. download it, and upload it again, you 're not really adding any new value, and people aren't going to want to see something they've already seen. That's why you can also leverage Cloue to make more [music] ideas. Let's move on to the next step, and that will be writing a good [music] script. Once we have a niche, we have ideas, the next step is to start producing content, and the first thing to do is to make a good script. The script is the heart of the video. I repeat, the heart. You can have the best thumbnail in the world, you can have the best voiceover, but if the script doesn't grab you, people will leave. Look, I do n't mean you have to invest especially since artificial intelligence is going to do it for you and always have to do it in the best way possible because the reality is that when someone enters and the people who enter that video leave, YouTube stops recommending your video. And that's as simple as that. Retention is [music] very important. Now, before AI, writing a good script cost you money or took many hours. Hiring a decent screenwriter cost between 5 and [music] per script, and if you wanted a really good one, it was much more, and I'm telling you this from experience. Or there was also the option of doing it yourself and you could spend two or three hours per script when you're starting out. Now with Cloud you can have a solid draft in 10 minutes, but be careful, when I said draft I didn't mean finished script, because this is where many people make mistakes. The AI ​​generates a text that sounds good, that is well personality, it doesn't have your touch, it doesn't audience connect with the channel. And if all channels are using the same scripts and nobody reviews or adapts them, all the scripts sound the same. And when all the scripts sound the same, none of them stand out. That's why I always insist on the same thing: leverage not delegating everything to the... we're going to do it exactly how I'm doing it with Cloue to write a good script. I'm going to give you an example with this channel here, with the channel we want to write the script for this video, okay? So what we have to do is , click here on more and few people know this, but down here where the transcription appears, if we click on show transcription, transcription here. So, all we would have to do is [music] copy it over here. What I've done now is move it to my browser because need to do it in incognito mode. I repeat, I will do the same process. I'm going to go here, show the transcript, I'm going to copy the transcript, then I'm going to go here to Cloud, I'm going to ask you, I want you to help me going to ask you, I want you to help me make a script for my channel. This is the stick the script here and then we tell him, "I want you to get inspired and make one from scratch [music] . Don't just copy and paste, okay? And Claude, before starting with the script, probably not, always asks us best it can be. Notice that it says, "Before key information to make the script truly original [music]." And this is . Look, what angle do you want for your script? a little-known biblical text, okay? [music] Here above all I'm going to choose something, I haven't seen that video, I'm not using it as an example, it's important that you 'm just giving you an example, okay? I'm going to give him, for example, of your channel? Spanish. I'm going to say that it's Spanish. And how long do you want the video to be? In this case, since the one we have is more than an hour, we will always do at least what we have as a reference. So, let's set it to more than 60 minutes. And Claudia might ask you a few more things here. one last thing before I start writing. Tell me which one appeals to you most and I'll start writing the minutes long. Which of these stories appeals to you most? The true story of so-and-so. I do n't have much context for this because I have n't watched much of the video, but I'm going to go with this one. And notice that he says, "Perfect, I'll start writing." Now I repeat, as I have already mentioned before, Cloue is a access. Obviously it has a paid plan, but you can do it perfectly well with the free plan. [music] Notice that it's already starting to generate the script. It's happening over here and it only takes a few seconds. The outline here has given us a little bit of what it would be, okay? If it convinces us, we would say yes. draft of what he wants to go on with. If we look at it closely. Look, one last thing before I release the video. It's camera, you speaking directly, it's not Faceless [music] , okay? That Faceless means without a face. Notice that after a few seconds it has generated the script for us, and if we click here we can see the entire script it has generated, okay? The entire script for that hour was exactly what we wanted. Obviously, depending on the voice, it will go faster or slower, but notice that here we have the entire script and it even indicates it from minute 1 to minute 10 and shows you all the text, okay? From minute 10 to minute 22. So, once we've seen how to make a good script, in this case with artificial intelligence, let's move on to the next step which is also very important [music] and that is how to do the voiceovers with AI. Before we continue, if you'd also automated YouTube channel generating over €2000 per month with the help of in the first link in the description you have a link to schedule a call with us. We continue. Once we have the script, we need the voice, and this is where AI has changed the game the most in YouTube automation, because before, hiring a voice actor cost you between $5 and $7 per script, which wasn't outrageous, but multiplied by the number of videos you need to upload each week, the cost was noticeable, and you also had a problem with timing, I'm telling you from experience. You would send a script to the announcer and it would take between 24 and 48 hours for the announcer to send it back to you. And if you didn't like something, you started all over again. Now with levels you generate the voiceover in minutes and the voices you generate are practically identical to a human voice. We're going to do it live. The first thing you have to do is enter the level and here in this text to speech [music] section is where it will let you start doing the voiceover. Here we would have to this case the one we made earlier, and here on the right is where it gets interesting. Here in the voice section, all the voices that you cannot use with the free trial. This tool is paid, there is a free [music] tool. You'll obviously say, "Eric, it's a paid service, I free, but as I mentioned before, it's much cheaper than what you used to pay someone for music . This could cost you $ a month, for example, 12, 15, 20, whatever, and before you paid seven for each video. [music] What's up? With that amount of money per month you can produce I don't know how many videos, also of the duration [music] of your videos. But hey, there are many voices that use, you see? We can select the voices we want here. In fact, this one sounds pretty good. select it and that's it. If you notice, this is loading and then it's put in. And here we can touch on speed, stability, similarity, and exaggeration. So, here all we would have to do is copy that script that Cloue made for us earlier and click here, for generate [music] to speech. So, all we have to do is click here to generate speech and that's it. Notice that here I can copy the first part of the script that it has made for me, and if I click here on generate speech do it. A phrase that researchers working with NC Hamadi's manuscripts... I repeat, we can copy, we can modify the speed, the stability, the similarity. I also recommend that you do it before generating the voice losing out on credits, but if you listen closely it sounds pretty good. from Berlin. What I also recommend here is that you don't put the whole script in at once. Because? [music] simply because it [music] if you have put the whole script in you would once. That's why I recommend doing it block by block. I'm not saying you should only use five sentences, but you could use, for example, 5,000 characters or 2,500 characters. Let's move on to the next step, and I'm pretty sure you'll be interested: the step of using miniatures with io. [music] The thumbnail is the first thing people see before the title, before the description, before anything. If the thumbnail isn't eye-catching, nobody [music] clicks. And if nobody clicks, in the video is. This is measured with the CTR, which is the click throw rate, that is, the percentage of people who see your thumbnail and decide to click inside the video. [music] A good CTR on YouTube is between 5 and 10, but that needs to be taken with nuance. Because? Because I've had trending channels where, like every day at the same time, we uploaded many videos at once, the one with the worked the most. Because? Because it's monopolizing the entire market. So, when I have Bergreen channels, I've seen that a CTR of 6, 7, or 8 is very good, but then I've seen trending channels where I've reached a CTR of 35 or 40. The packaging is the combination of the title and the thumbnail, and these two things decide whether someone clicks on your video or skips it. That's why I dedicate so much time to this. Before artificial intelligence, you needed a miniature artist, a designer who cost you between 5 and [music] 15 per miniature depending on the quality. And mind you, a good miniature maker made all the difference. Now with tools like Gemini or Chat GBT you can generate thumbnail concepts, test ideas and produce results that previously required a designer. It's not perfect yet, but for getting started and testing it's more opening any tool, the first thing I do is analyze the thumbnails of the reference channel, because that's where the information is. Take a look at the thumbnails this reference channel has . They all have something in common, which is that they don't use a lot of elements; instead, there is [music], an image of a person, some text, and a background that is related to the [music] like to follow the obviously there are thumbnails that elements in the thumbnail it makes it saturated and the message of what it means is not well understood . That's why I always say to follow the three-element tool, which is image, image of a face and the these three elements, let's generate the thumbnails with AI. I want you to see some examples of thumbnails that I've made myself using Chat GPT. If scratch. I simply gave him the prompts and he did it for me. This one, which is similar, which fact [music] with this one I think is the one I kept. And then I also made several variations and I've made miniatures like this one or this one or this one here too. really interesting. We're going to go to the reference channel and pick up the thumbnail we have. Let's take a screenshot [music] or what we can do is literally go here and download it with the PitQ extension and drop it here in the GPT chat and say, I want to make one just like it GPT chat and say, I want to make one just like it in Spanish. Change a few small in Spanish. Change a few small things and just change the text to Spanish. We ask us a few things, he might not, but notice that he didn't. He's already creating the image from scratch. Look, he's already made the miniature for us and it's absolutely we would have to change some small things [music] , we can't do such a blatant theft, but notice that here we would have to say some prompt like change something , some color, but he has taken that thumbnail, changed the text. Here, I repeat, we would have to tell him to change [music], some color, or something like that, but always following what us too. I'm doing it with the free plan. Notice that the free plan appears; I don't have the paid Chat GPT plan, and look how well it made me a thumbnail [music]. This is so you can see that you have absolutely free. You can also do it with Gemini, do it with whatever works best for you. The point is always to give it a very good reference so that it gives you next step, which is the editing and assembly of the video in general. Of the four pillars of content creation, editing is the only one that still requires a person in most cases. There are help you, but these days to have a professionally edited video you need an editor or to do it yourself. And here I want to give you a piece of advice that would have when I started. Make the first videos yourself, don't delegate them yet. Because? Because you need to understand the process, you need to know how long it takes, tools are used, and what result to expect. If you delegate from day one without understanding the process, you won't be able to assess whether your editor is doing it cost you money in the end. To edit you can use Capcut, which is free and super intuitive. You also have Shotcut, which is completely free, or iMovie if you 're using a Mac. You don't need any paid software to get started, so let's get on with it . Okay, I'll explain how to do it with CapCut. Here you click to create a new project. In this case you're going to give it 16 2.9, which is the YouTube format. And importantly, what you'll need to include are stock clips that are related to that video. To do have as a reference, see what types of stock music clips they have posted, and then add them. That way, all you would have to do [music] is put the You could add music, but you can even use YouTube [music] Studio if you want to make sure it's not copyrighted, or from any royalty-free music library . If you're stock clips, you look for stock clips about religion, for example, right? We'll look into it. Stock music clips about religion, okay? And it would be going in , for example, let's see, [music] here if we put religion, you see? Here we already see many, many things, okay? If we log in, some don't have a watermark. Then it would be as easy as downloading these, coming here to CapCut and literally putting them on top, adding the voiceover and setting the subtitles to automatic. So that's how the whole process would be done. I don't consider myself an expert and it's not something I do these days, but I because I know exactly how it's done. Let's move on to the next step, which is publishing and scaling. Once we have the video ready, it's time to publish it and start building the channel. And here are some things that seem The first thing is to configure the channel correctly, the channel name, the channel keywords and verify your channel with a phone number so you can upload custom thumbnails [music] at the channel name level, customization, don't name isn't going to make your videos do better or worse, the name isn't going to make the channel do better or worse. In general, at the channel level, at the at the channel keywords level, it is true that you have to put something that is related to your niche. This is basic, but you'd be surprised how verifying their account and then don't understand why they can't add thumbnails. As for the channel name, I've already told you not to overthink it Cloud, tell it a little bit about what the video is about and Cloud will give you examples of names and you can just pick the one you like best. Once you have it, go to YouTube to make absolutely sure there is n't another big channel with the same could get sued or something, but don't overthink it. [music] In this case, I'm going to show you how to do the process. We say, "I want you to help me process. We say, "I want you to help me create the channel name [music] is about religion. Okay, we'll try to make it a short channel name, not too long, so it's a short channel name, not too long, so it's we've answered these questions, and now it generates 10 options, and of keywords, let's go here to our channel. This is my settings, and here in the channel section, you're going to put your keywords description, if you want, you can paste the script of what your video [music] is about and tell Cloula to generate a custom description for that video. Now, the recommend starting with three or four videos per week, and you don't need to post every day at the beginning. What matters is consistency. It's better to post those two, three, or four videos I mentioned per week [music] for [duration]. Three week and then disappearing. In the end, YouTube, so you understand, rewards consistency, and so does your audience. And now we're also going to talk about monetization. YouTube requires two things to monetize your channel: 1,000 watch time on your videos. With a good niche and consistency, this can be achieved in a period of between 30 and 90 days, which is how long it took us to people who are [music] in our academy. To give you an idea, my personal record was monetizing a channel in 5 days, and our [music] student Raúl did it in 3 days. That is to say, it's possible to go very quickly if the [music] niche is the right one, but again, the average is between 30 and 90 days. This is the stuck for months. Between the end of 2023 and mid-2024 In 2024, I had one or two channels generating around €2,000-€2,500 a month, and I became obsessed same channels to €10,000 a month. I spent months dedicating all my work to the same channels without that figure budging. I was really simpler to have five channels generating €500 a month than a single channel generating €2,500. Because if you diversify, when one channel drops or stops working, you have others to support you, and when you increase the volume, out of every three or four channels you open, one or two will work. And that's how I went from €2,000-€2,500 a month to €10,000, € 15,000, even months of €20,000 a month. And look how AI comes in here now, with AI handling scripts, voiceovers, thumbnails, and opening a A new channel costs you less than €100 and a fraction of the time. [music] The process we just went through from choosing the niche to having the first video published, can be repeated for each new channel. It's the same system applied to different niches, and that's what will allow you to scale without working more hours. Let's recap. In the last 40 minutes, we've built an automated YouTube channel from scratch using artificial intelligence. And the process has seven steps. Step one: choose a good niche, which, remember, accounts for 90% of the results we'll achieve. Look for a recent channel that's your market, and where the production cost is low. And you can use niche bending to take a format that works and apply it to another market. Remember viral ideas. Don't make anything up; go look for outliers, and use Cloud to generate variations adapted to your niche. Step three was writing a script, which you can do with Claudie, and structure of the reference video [music] . It generates a draft in just a few minutes, and you review it in just a few is the heart of the video. Step four was the voiceover, which with Leven Labs generates a natural voice for a fraction of what a voiceover artist used to cost. speed and stability [music], and have the voiceover ready in just a few minutes. The next step was the thumbnails, which you can create using Chat GPT or Gemini, applying the three-element rule we discussed earlier. make the first few videos yourself to understand [music] the process. Then you'll start delegating those videos, paying around $0.70 per minute, and step seven, which we've seen, is publishing and scaling. Remember, we said two, three, or four videos consistently. If you follow that, you can actually start monetizing within 30 minutes. 90 days, and once you have that channel up and running, you repeat that process with more channels. Diversifying is total cost of starting all this with AI is less than €100. The time it takes to have your first video published is really just an afternoon. Before AI, the same process cost you at least €300 or €400 and several weeks of work. And changed. And I'm going to be completely transparent with my figures. I manage between 10 and 15 channels and bill between €10,000 and €15,000 a month, depending on the month, but that's roughly the average with my own channels, and we have more than 600 students in the academy, many of whom are already generating €2,000, €3,000, €5,000, and some are even generating more than €10,000 a month with the same that's the same. Just as important as everything I've taught you today. It's not simple; it consistency, and making good decisions, especially when choosing your niche. Again, AI helps, lowers costs, and saves time, but AI doesn't replace effort. It's the price you pay, and anyone who tells you otherwise is mistaken. If you've made it this far, it's because you're genuinely interested in setting up your own 3,000, or even more than $10,000 a month, as we 've already done for many people who have if you'd like us to contact you to see if we can help you too, we've included a form in the first link of this video's description for you to schedule a call with us and really see if we can help you. [music] I want to emphasize that we only work with people who are committed and truly want to do this right. So, achieve These results also fill out the form in the first line of the description. See you in another video.