This YouTube channel is getting 371,000 views the last 30 days. This YouTube channel is getting 776,000 views the last 30 days. But it's not only on YouTube, this Tik Tok that is doing more of a scary story type vibe, are getting hundreds of thousands of views, and you can do it all with the automation that I'm about to show you. The young boy sat on a cracked sidewalk wearing torn clothes and dusty shoes. His eyes were full of dreams, but his world was filled with struggles. He desperately wanted a better life, more than just survival. One day, he spotted a flyer for a citywide talent contest with a grand prize, a scholarship, and cash. Pretty cool, right? All you need to do to make this is to use the vibe creator. Just select your idea that you put in here. This is a rags to riches story. Then you just click on this button that generates the entire YouTube short script. Then you can click on this button here to generate the voice over. Then we step into turning the script into images that splits the script into 15 different parts that then turns into these beautiful images all in this nice layout for you. If you want to change the image itself, you can easily click on the redo image button. So, say this one for example. In just a couple of seconds, it's going to create a new image for us that may look better, may look worse. I actually really like this image. Perfect for our protagonist. Once you're happy with all your parts and images, we just click on these buttons here to generate the video itself. And as you are going to see, it's an image to video workflow that creates these insane AI videos. Once you have generated all the images and you're happy with them, you can go to the next step, which is to just click on generate video. This takes the audio, the videos back to back to back and creates captions on top of it to make your full video. Once the video is finished and you're happy with it, you can just click here. Then you can publish it to Tik Tok, YouTube, Instagram, and Facebook with just a click of a button. This allows you to get full control over the entire process, combining you as the human with AI for what it is best at because you definitely don't want to do a fully autonomous system from end to end because the quality of the videos are not going to be as high than if you use AI plus human combination. We call this the vibe creator, using AI for what AI is doing best and a human in the loop to check up on AI's work all the way throughout the entire process so we can make the best possible video. So, I'm going to show you how to set this up step by step, even if you never made an automation before. So, by the end, you can have your own vibe creator creating these nice short stories together with AI. Step number one, duplicate the Air Table base. So, as you can see, this is the air table base, which turns into our AI trigger agent. And the real reason that we're using this is because you can create a button that triggers an automation. If you take a look here, we basically only send a web hook to make.com that then runs the automation, grabs all the information back again, and updates this short script. So, let me show you how to set this up right now. to duplicate the base. Go into our Aimate School community in the link in description. Completely free to get started and just go into the classroom here. Click on Vibe Creator 2.0. Then go to the Vibe Creator 2.0 short stories. And here you have a link to duplicate the Air Table. Just click on that and this is the page that you're going to land on. Then just click on this drop-own menu in the top. Click on the three dots. Click on the duplicate base. And it's very important that you change the name of it. So use your name. So it sounds like Vibe Creator 2.0. Bob, if your name is Bob. And just like that, you have duplicated the entire workhub, the AI machine if you will. So we're ready to go to step number two. Download the Make.com blueprints. So if you haven't set up a make.com automation before, I'll show you that we have couple of automations from the script, voice, images, redo image, video, generate, and publish. So we need to set these all up. The first one that we're going to start with is the script generator. This one takes your idea and turns it into a full YouTube script. So, how you get started with that is to go into the school community once again, click on the download blueprints, and that's going to take you to a Google Drive folder here. Just select all of them, right click, click on download, and just like that, you have all the Make.com blueprints that you need to get started with this automation. So, let's go back into make.com. And if you're completely new, we're going to go to the dashboard and then we're going to click on create a new scenario. Inside of this scenario, all we need to do is to go to the bottom, click on the three buttons, click on import blueprint, choose a file, make sure you select the right one. So this is the script JSON open, click on save, and just like that, you have the sack same automation as me. But if you try to run this, nothing is going to happen because we need to go to step number three, set up the button trigger. So as you can see, this web hook right here is what gets triggered if we click on this button right here. Since you have copied this blueprint right here, all you need to do is to choose a hook. Click on add. I'll name it script and vibe creator. Just like that. Let's save it. And here you will get a web hook. Just click on copy address to clipboard. Go into Air Table again. Click on the button in the top here, edit field. And here you have a URL formula. This is way simpler than you might think. All we need to do to change this is to go from the beginning all the way to this question mark and click on paste. Did you see that change? Awesome. And for the other parts of it, you can see that what we're sending is the YouTube short script. This is right here. And we're also sending our idea. This is right here. So, as you can see right now, we're going to send both those data into our make.com automation. If we just save it, and now this button should be connected with your script automation. As you can see, successfully determined. We click on save. And then we go to the next step here of setting up the open AI module. Here, all we need to do is to add our connection. This is really easy to do. And in the next part here, all we're doing inside of this prompt is to have a very high quality prompt that tells stories. This format is from Dan Harmon's story circle that creates incredible short stories. There are a couple of extra things in here as well. And what we do is we take your idea that you have in air tableable and put that at the top of the prompt. Lastly, we have to update the air tableable. So if you haven't set up air tableable connection before, all you need to do is to click on your connection, make a connection with your air tableable account. Then it's very important that you change this to Vibe Creator Shorts. And this should be Vibe Creator Bob cuz your name is Bob, right? So you probably have a name uh that is a little bit different than mine table. You can select the short video. Um in other cases we're not really going to change this if you have made the blueprint record idea is going to be automatically se selected for you if you have downloaded the blueprint and updated it for you. So literally all you need to do in this automation is to select the correct base. Once you've done that the entire automation is finished. So now if we make a new uh thing here and we call it um fairy, we can write an idea in here. If you don't have any ideas of what to get started with, we also have a bunch of really cool storyline prompts in the 10x prompts inside of the community. Here you see the fun story lines like the forbidden shortcut, the unwelcome stowway, the fake identity, the wrong delivery and things like that. So we could easily take something like this. We could post it back in here. And the really important part here is that you add your character. So they is 14year old boy sign a contract without reading the fine print. Now they're magically bound to complete a bizarre and dangerous task. Pretty cool, right? So, all we do then is click on the button. And if you don't want this to pop up all the time, you can click on control and click at the same time. But in just a couple of seconds, you see that we get the entire YouTube script right here, ready to get going. So, what I recommend if you're following this step by step is to set up the web hook for all your buttons. Once again, go and create your new scenario. Make sure you select the web hook, create a new web hook, and for example, connect it to your voice over file as well. And if you don't remember, we just select everything that is before the question mark here and we paste our new web hook right in there. And so if you just keep going programmatically through all the buttons here and just change your web hook, you are finished with step number three, which is setting up all the web hooks. So all your buttons work and we're ready to go to step number four. Set up all the automations. So again, we have voice images, we have redo image, we have video creator, we have the generator that brings it all together, and then we have the publish automation. Now, I'm going to not go step by step through every single module here. I'm just going to go through what you need to change in order for it to work. So, we'll start with the voice over. And I assume that you have the web hook set up. Next, you have to set up the 11 Labs. So, all you need to do here is to add a new connection, sign in with Google, and you're good to go. Next, you should select the voice that you want to hear. This is easier to do if you just go inside of 11 Labs, click on the voices, and just test between many different voices. They're all really, really cool. Some really deep like authoritative. Some are like dapper and deep narrator. And you can even clone your own voice that I have done as well, which is super sick that I'll show you in an automation in another video where we're going to go into avatars. So once you've done that, all we need to do is to get the short script that we get from the web hook. We can select the stability. I like to have it at four and similarity at 08. I'll select a default mode here as well and click on save. Next, we need to set up Google Drive. So in the top here, just add a new connection. And if you're using a gmail.com Google account and you're having any issues with it, we'll happily help you out inside of AIM. However, if you are just going to select a folder, you just select any folder that you want. We have a folder called YouTube shorts voice over that uh we basically store all the voiceovers. For the file name, we map it to the source text. So, it's going to be called the little snippet that we have and the data going out from 11 Labs goes in the bottom here. And then we click on save. Once we have done that, all we need to do is to update the air table record. once again. So, make sure you have your connection set. Make sure you change the base here, but other than that, we're not really going to do much. Make sure you have the download voice over file be web content link. And that is pretty much all you need to do in this module here. Once that is set up, all you need to do is go back here, click on short video, go to this fairy button here with a voice over. I'll controllclick it once again. And in just a couple of seconds, this one goes really fast. We will have the voice over. Let's go. Right on time. And uh this looks a little bit like this. If we click the link, there's a download. 18-year-old boy was sitting in his cluttered bedroom scrolling on his phone. He looked tired but curious. He saw an ad promising instant. Just like that, we have the entire voice over of the video. Pretty cool, right? And if you're setting this up step by step, you would literally just go to the next script to images. You would go into your script to images uh automation here and you just go from step by step. Just select the air table, make a new connection, make sure you have se selected the base to vibe creator Bob. Then make sure you have all the things connected and then we get to the next step which is to make HTTP requests. So these modules here are HTTP requests. And the reason we're using these is because we can connect to any API on the internet with this simple module right here. So let me show you how to set this up. This is a skill that if you are an automation builder and you don't know this, listen to this carefully. So I'll just click on the HTTP module. And here you have my bearer token. I'm going to delete this after the video. So don't try to use this bear token running on my API. But if you have imported this blueprint, this one should be exactly the same. In this automation, we're using the image generator which is called flux Chanel. The next thing is that we need to have a method post. This is basically a send request to this image generator and what it is looking for is an authorization. This is basically how you log in if you're using APIs. So, we need to have the name authorization in the top and the value. We have a bearer, a space, and then our token. I'll show you where to get this token in just a minute. In the item two, we have the name, content type, application/json. These are very important to keep in there. Just keep the body type raw the JSON application/JSON. And here we have the input request. This is the image prompt that we get from this module right here. This is the image style that we get from this style right here. I quite like the 3D animation movie style. So here inside of this uh image creator prompt, we also select the aspect ratio. So, if you wanted to make this a horizontal video, you would do 16x9, but since this is a short video, we do 9x6. Awesome stuff. So, we'll just click on save or basically I'm going to show you where to get this bearer token. All we have to do is to go to replicate.com. And if you want to be on this exact module here, you can search for Blackforest Labs/flux schnell. And again, we're using this because it is so cheap. It's less than 1 cent per image and we get commercial use. However, if you want to upgrade to another version, you can easily do that by clicking on black forest labs and you can use for example flux 1.1 pro ultra fine-tune that is really high quality that gets even better images. But I found that flux does the job quite fine. But easy way to change it over here by just clicking API clicking and finding the URL that we need which should be right here inside of HTTP and this is the link that you would now put in the top of the make.com automation. So I'll just show you here. Just go here and paste that pro and that should be fine. So, that's pretty cool that we can set up all the different image models right inside of here. But what I really want to show you is the barrier token. So, let's go back into replicate.com. Click on your name here and select API tokens here. You just create a new token. So, you can call this vibe creator and just click on create token. Once you've done that, don't share this one to anyone, but I'm going to do it just for the video. Copy the token. Go back into your image generator and paste. Just like that, it's all set up. Ready to send that request to our image generator. And just like that, if we click on save now, it should be working. And we go to the next step, which is to get a request. So here we're sending. And here we are getting. This is the exact same thing that is happening that we just did before. All you need to do is to change the bearer token with a new token. Just like that. We've already set up the URL that we're going to get all the data from. So everything else should stay the same. Just changing the authorization here and click on save. Next, we have to get a file. And this is really we only get the data output from this module right here. And then we upload it to Google Drive. Do you see how easy this is? All we do is make a request with the API. Then we actually wait for around two seconds for the image generator. Then we get a request to see, hey, are you finished? Give me that link. And in this module, we download the actual image itself. And then we upload it directly to Google Drive. I'm not going to explain these three modules because I already did before. They're just going to update the next field here of then we create all of the 15 different images that you want based on the parts here. So part 1 2 3 4 5 6 7 with the image and the image drive file. You get the prompt that we sent to the image generator itself as well as the script text on this side. So you can see what kind of should be in the video when the voice over is talking. So far you've gone through creating the script, creating the voice over, creating all the 15 images that you might need for the video itself. And now we're going to set up the image to video. We also have the redo image button. This one is very simple. Once again, it's the exact same automation that we set up right before this. So, I am confident you're able to set this one up on yourself. However, for the video one, it's a little bit different, but it's literally the exact same except we're not sending a text to image, we're sending an image to video. So, essentially all that happens here is that we make a request once again. This is to the cling API. Here it is on replicate.com. And this allows us to generate 5 to 10 seconds if we want to do that. here as well. You can select 10. We have selected five because I like the video to slide between the different videos. You can also select the new aspect ratio. We upload our image and stuff like that. And the cost, if you're wondering, is 5 cents per second of video. And for us, this was one of the cheapest ways to generate. And it actually ends up being around $3.5 per short video. And some people might say, "Oh, that sounds like a lot of money to generate just a short little video." But for me, if you generate 30 videos, one per day, right? You get 30 videos times 3.5. But let's say it is even a little bit more when it comes to the voice over and chaty BT and stuff like that. Let's even say that it's $5 per video. If you post every single day for a month, that will cost you $150 per month. So, your entire marketing team is essentially $150 per month. Imagine you were to hire a virtual assistant to do this for you. They would cost way more than $150 per month, which is a cheap price to pay for such amazing quality videos in my opinion. If you want to get around this completely, you could generate just image by image by image, but that is not going to animate the video. So, I really wanted to pull all the strings here that we turn it from image to video and then we get the full animated video. That looks really cool. So, to set this up, it looks exactly the same as before. We send to the cling URL inside of Replicate. We paste the authorization barrier token right here. You know how to get your API. It's the exact same. So you can just copy and paste that. And the input that we give to this model is the prompt smooth camera zoom and slight scene animation. This because some of the images are of a character, some of the images are of a laptop or something like that. So we can't be extremely specific when we have all these variables. However, the start image is from the drive link that we created in the previous uh automation. And then we're going to sleep once again because it turning the image into video now. So this is around 3 minutes and a half and that's how long it takes to generate the images. And then we get the exact same request. We just get the data URL. We just change the bearer token and that's all we have to do to get that URL. Now we have a router that if it was successful, it's going to get that file uploaded to drive and update our air table. But if the generation is a little longer than 3.5 uh minutes, we will allow it to sleep once again. And the reason for that is that sometimes if it's not busy, it takes around, as you can see, 3 minutes and 7 seconds, but the total time to get the entire thing is 7 minutes and 30 seconds. So this is kind of like a safe so we don't get an error and we have to run it again. Instead we just wait a little bit longer so we get the generation and the automation works flawlessly once we've slept for a little bit. We're going to make another request after about 8 minutes and that is going to get our image here. And then we have another router that if this was sensitive content, which might happen, we're going to tell you inside of Air Table, hey, there's an error. This was sensitive content, please run it again. But if it wasn't sensitive content and everything was fine, we basically do the exact same thing here. Getting the file, uploading it, and updating Air Table. the exact thing happening here as well. So you can be sure that you get the video every single time besides if this is sensitive content. And if it is sensitive content, all you have to do is change the image and run it again and it should be completely fine. You will get an output video right here that turned your image into a super sick video. And you will also get the drive link because this is important in the next step. But I think here I should demo it and show you exactly how easy it is to do. Once again, we have the fairy story here that is some sort of fairy story. I have the 3D animation movie style. And if I click on this button right here, again, we're going to run through the Flush Chanel model. You'll see it get updated right here with fairy part one. fairy part two number three. And yeah, just let's time-lapse this. Just like that, we got every single image and we are using a prompting technique to take context from the entire script and then saying only make this part. This makes the images match the audio a lot better than some other methods that we tried. And it's looking really cool. There's a sudden contract that pops out of the screen. He signs it and get pulled into this weird universe. Then he has to go and grab an amulet. And immediately when he touches the amulet, he's back in his room. And you might wonder how it looks like if we start turning these images into videos. Well, again, if you have set up everything correctly in this automation that turns images into video, what I recommend is that you save it and then turn this immediately as data arrives on. Then just go back, make sure that it's on in the top right as well. And then you can spam every single button here. I'm just going to show a couple of them. I'll click on this button. I'll click on this button. And which other ones do we really want? I really want this one where he pulls into the uh entire uh new world here. And if we go back to the video creation automation, you can see that four of them are running simultaneously. So this is the time where you can go away from your computer. You can step back again with anywhere from around three and a half to 8 minutes and it's going to get updated directly here in the air table. Just like that, they all ran successfully within 3 minutes and we got the video output here. Let's take a quick look together here. This is the one where he is going into the depths. Very cool. Here is a sudden contract that pops up in front of him. And here is the amolet that he's holding in his hands. Wow. But how do you take all these videos that you created now and smack them together so they're actually a full video? Well, that's why we go to the next step, generate the video. So, as you can see, we have the videos created tab in the top here where we can just click on generate video and it's going to run perfectly. Now, all we need to do here is to click on the title. Do you remember what I called it? It was fairy. And it's important that they have the exact same um name here because we use that in the automation. Next, we have to select all the images that we or the videos that we want to have in the actual video itself. I'm not going to run fairy because I have to generate all the videos once again, but I'll use the rags to rich a story here. All you need to do to include all the parts in your video is just drop down this menu. Click on include in next video. Copy, select all of them just all the way down here. Part number 17. Boom. Click on paste. And uh all of these ones are now going to get included in your final video. If we go back to videos created here, we can call this one riches. Once again, we'll just click on this button right here. We'll hop back into the automation and you can see that is running. It's going to grab a bunch of those videos that we have in Air Table. It's going to pull it all the way into JSON to video that mashes all the voice over as well as all the videos together and it's going to put a caption file on top of that which is incredible. And once it's all completed, we of course get the video URL here in Air Table. The young boy sat on a cracked sidewalk wearing torn clothes and dusty shoes. His eyes were full of dreams, but his world was filled with struggles. He desperately wanted a better life, more than just survival. One day, he spotted a flyer for a citywide talent contest with a grand prize, a scholarship, and cash. Just like that, if you ask me, this is just insane. But this video is no use if you just keep it to yourself. don't share it anywhere. So that's why we go to the next step, publish. So in the last button here you can see publish. We have all the drop-downs from Tik Tok, Instagram, YouTube and Facebook. There is uh possibilities of publishing elsewhere as well. But these are the main vertical platforms that we have integrated. If you check out the automation itself, you can see that it's really simple. And the reason it's so simple because we are using blow potato. This is a made for make API to automate posting on socials. So they make it super easy for us. You might have tried testing out some other posting automations where you use the native make.com automation, but we've found some issues with that. So for us, it was better to use a service that is literally meant to repost everywhere. To set this one up is also very simple. You just have the upload API HTTP request where the URL goes to a backend blootato. You have the API key right here that is blurred and the data is raw. The content type is JSON and the request URL is a drive link. This module gets the transcripts. Then we write a caption that is nice for social media. And then we post that to Instagram with a make request to blot. If I click on it again, you'll see that the URL that we're posting is actually a V2 post and not a V2 media that we used in the previous step. And in the request content, we post to target Instagram with the content text of the result which was the caption. and we select the URL that we actually want to upload and publish. And bada bing, bada boom. Let's post it to all the platforms by just going back, making sure that everything is on. So right here, let's click on publish. It's going to start running our automation that uploads the video to Blowato, gets the transcripts, writes the captions, publishes to Instagram, Tik Tok, and YouTube right here with the native model because didn't work so well with that. But here we go. Published on four platforms. Here it is on Facebook. Here it is on Instagram with one like and comment already. Here it is on Tik Tok as well as YouTube right here in the studio ready for people to enjoy with YouTube shorts. So you now know how to turn your idea into a script into a voice over into a bunch of images that you can redo if you like them or don't like them. Turn them into images and whip it all together and publish on all these four platforms. So I have a question for you. Do you think content will be easier once you have this? And if you're going to start building automations, do you want to do it alone by yourself or would you want a community to do it together with you? Well, if you choose the second option, I have good news because in the AI mate community, you get it all from the automations that we're building right now, including the vibe creator to set this up as fast as possible. You can hop on to calls on Mondays where we share what we are building and working for us. And if you run into any problems, you can just write a quick post and we're going to fix any tech problem within 24 hours. So you can have these automations set up and ready for you right away. Go and check it out in the link in the description. It's completely free to join right now and let's start building amazing automations together. Thank you so much for watching and I'll see you in the next one. base.