How To Create A History Channel With ChatGPT & AI Video Generator

12.01k views3434 WordsCopy TextShare
AI Andy
Generate your original, high-quality long-form content now! Use the coupon code: AIANDY and get a 25...
Video Transcript:
take a look at this history and mythology channel that in the last 30 days got 5. 6 million views and made anywhere from1 to $222,000 before this used to be very hard writing a script voicing it over with a microphone and finding footage for hours but this made me think what if we write the script with AI we use AI voice over we make a script that creates 90 AI images that matches the script and we bring it all together to make that video and by the end we answered the question is the video actually any good so if you want to come along with me let's try to make a video like this together step number one make the script so first we're going to go to SEO writing. a and unlike Chachi BT it actually creates longer articles around 2,000 plus words so let's just click on get started now that we're on the inside I'm going to click on oneclick blog post and since I saw that this sad stories of zoos lovers was one of the most populars I want to piggyback off of that idea we'll put Zeus lovers into the main keyword and generate a title exploring Greek gods romantic Tales that's pretty cool then let's pick the core settings let's make the to of voice quite witty the point of view third person he she it Target country United States article size let's keep it at medium at around the 3,000 words and in the AI model let's use GPD 40 last Model I've already included my API keys so that brings our total cost down and they've also included this humanized text that I want to keep at a fairly low reading level so anybody can watch it now let's scroll down and if you were to make a Blog with this you could add AI images inside of here but I have a very interesting tool that I created for us that I want to show you a little later so for this we'll click on none YouTube videos also a none for we really want to click on this NLP keywords generation so we can rank better this will include those words in the actual article itself and then we'll land at the structure one of the best parts of SEO writing is you can include a good hook and that's really important in a YouTube video because you basically out a 5 Seconds to grab someone's attention and with this exact video I definitely believe the anecdotal or story is going to hit the hardest now since we're using this for YouTube generation we're not going to include a conclusion table h3s lists and everything like this this looks really cool for a Blog but for us let's take them away and let's continue going down here we really want to connect to the web I'm going to do Deep Web that analyzes hundreds of the most relevant websites we can also include the source link that will be amazing to have in the link in description so people can actually see where we got the information from and then we also get syndication so this is really cool if we have a Twitter LinkedIn Facebook for our business around history and Greek mythology we can basically just get the entire post written for us I'll skip the WhatsApp message and Pinterest and then we reach the outline editor this is where we can edit the outline before it actually goes out but I trust in SEO writing to do this for me and lastly we reach publish to Wordpress this would be awesome if we had a Blog but that's about it let's scroll all the way to the top and click on run and just like that the entire script is written for us with multiple short stories of Zeus and his lovers and at the end of it it includes all the source links that the script is based upon it ended up around 2,300 Words which is around a 10minute long video so now all we need to do is Step number two create a voice over let's copy the entire script and let's go to 11lbs doio because this has the most realistic AI speech obviously if it sounds way too staccato and robotic people aren't really going to watch so I hope we can find some good voices in here let's go to the voice library and listen to some of them I'm thinking a manly American voiceover so let's change the language English accent American and gender male let's take a listen after all the most important quality of speech is that it is yeah maybe not John do I've seen things you people wouldn't believe wo that sounds good for history and mythology but let's take a look at something else later that same evening detective Carlson received an anonymous that's actually not bad let's keep going it is a profound truth that the Deep things in science I'm feeling maybe John do I've seen things you people wouldn't believe yeah that would be perfect so we just did all the three parts now let's just download them and head to step number three generate the video so for us to create this video we need a bunch of AI images but it would be pretty annoying to prompt again and again and again with hundreds of different images so what I did was I created a script for us using the replicate Doom API that splits our text into sections then feeds replicate that exact prompt and in turn we can automatically generate hundreds of images automatically which saves us a bunch of time all we need to do is copy the entire text from SEO writing AI open the script I made and paste it in here then we Define the style that we want and I would love to have cinematic and then we choose how many parts or images that we want since the video is going to be around 15 minutes after all the audio is done and I want to have a change around every 10 seconds that means we need around 90 images so in the parts here I'm going to write 90 but for demonstration purposes let's just do to 9 save it write python script to image.
piy enter and as you can see we get image part 1 2 3 4 and it just keeps going and creates all of these amazing images for us all neatly sorted in this folder so as you can see all of the images perfectly created so I want to try the mythological highly detailed digital art let's see how that looks like and here are the results okay this one is really cool this one too seems very mythological and very creative yeah I love this one and especially this one if I saw this I would be really engaged so let's keep going with this style and I also want to show you how to use this and if you've never coded before and you think that you need a lot of technical experience to use this it's actually super simple I also don't know how to code so if you know how to copy and paste then you can do this but before I show you I am working on a way for us to make 90 images then automatically put that into a video editor with 11 Labs voice over all in one go so it's all automated and if you want to see that comment down below AI video generator and like this video that shows me that a lot of you are interested in it and I'll be happy to spend more time so we can get an even better tool that we can use together all right with that out of the way let me show you how to set it up so all you need to do is first go to replicate dcom click on your account settings billing and set up a payment method we're going to be using the flux Snell model which is the fastest and the cheapest because it costs around 0. 003 per image so 100 images would cost around 30 we'll go to this black forest Labs flux Snell model and click the API and then we'll click on the python and copy our replicate API token just click here now just go to code dovisual studio. com because this is where we're going to run it now that we're on the inside all you need to do is click on file new file and then we're going to name it something like script to image.
piy and I recommend that you save it to your documents folder now in the link in the description I have given you access to the entire code so all you need to do is copy then just paste it in here now let's take the replicate API token and if you have a terminal in the bottom right here that's perfect perfect but if you don't click on the plus icon to create a new one if that doesn't work you can go to the top click on Terminal and click on new terminal here you have the environment that we're going to run this script so I'll just copy and paste the token into the terminal click on enter and that should be fine now all I'm going to do is save these scripts to image by clicking contrl s and then I'm going to write Python 3 script to image. Pi and click on enter just like that you will see that it's going to start creating images for you but you might run into a problem and you actually almost guaranteed going to run into a problem so that's why when you run into a problem you just copy the error code that you get into chat GPT and chat GPT will rewrite the code for you and then you just copy and paste back and forth and if you have any problems there just ask for more information or what to do and follow chat gpt's instructions by just copy pasting the code and copy pasting the error message and I promise you that going back and forth a couple of times will get it to work for you also to make this run you would have to write pip install replicate and run that as you can see I already installed it but again make chat GPT your best friend and you will have it up in no time then like I said before all you need to do is triple click this you can write your entire script in here choose the style choose how many parts that you actually want and we can now use script to image pie file to get some amazing results so all that's left to do is to actually create 90 images so we can make the actual video so I'm going to save my P file write Python 3 script to image dop enter and I think we're going to do a Time lap I have to say that this is absolutely incredible I guess there are some things that that are kind of wonky like I'm going to show you this one for example you see that an arm is coming out of his head so there is a way that you can actually make better images and that's to change the flux Snell to the flux Pro but the flux pro model is 5 cents per image which means that 100 images is around $5.
Copyright © 2024. Made with ♥ in London by YTScribe.com