How To Have AI Tweet (x) Your YouTube Videos
Let's build with Zapier and AI (100+ videos)📅 2025-05-05
Transcript
[00:00] Let's learn how we can start automatically creating tweets like this where we have a personalized tweet that's about the video, the link to the actual video itself, timestamps for the video, a thumbnail to the video, all done automatically when you schedule on YouTube. Let's do it. Let me show you how I set up that automation so we can automatically follow any YouTube channel or our own YouTube channel. A new video is posted and when that new video is posted, we automatically make a tweet. Sound good? Let's jump in. To get started here, we'll jump over to Zapier. a create new zap. Now, one thing you
[00:30] noticed in that tweet was not only were we able to create a custom description of that video. It had context of what even the video was, but also we were able to get timestamps. In order to do this, we're going to leverage a new software here called bumpups. With this bumpups integration here, we're able to access any YouTube video and do whatever we want with it. Whether that is sending a chat to it, so think of it like a chat GBT chat, you're just prompting at it, or generating timestamps, creative descriptions, and there's going to be a bunch of more stuff that you can do with this kind of tool. For now though, let's go to integrate this. So, I'm going to go ahead and rename this zap to YT
[01:03] tweet, maybe YTX because it's X now. So, we'll say YTX, formerly Twitter. I'll make sure I leave this automation in the description down below as well. So, I'll share it. You can click it, add it to your project for free. The trigger, obviously, because we are basing this off YouTube. It's going to be YouTube. And what's really cool about YouTube here is we're able to do a bunch of different action events when it comes to a new video. Therefore, in theory, we could set up this automation to follow any YouTube channel. It doesn't just have to be your own. So, for example, we can just do new video or we can do new video and channel
[01:34] or new video by search or new video and playlist. For the sake of this video, we'll do new video and channel assuming that we set up this automation so that anytime you upload on your YouTube channel, it will automatically exit. Is that what it's called nowadays? Exiting. I'm I'm not too sure. We're going to continue. The channel name is actually going to be your ID. Don't do like the at@ whatever your username is. is it's going to be a channel ID. To find this, typically you can just type in like YouTube channel ID finder, like straight up just going to Google and type in YouTube channel ID finder and clicking something like this. So, I went ahead and found my channel ID right here and
[02:04] we are good to go. So, we're going to test this trigger with the current way Zapier integrates with YouTube. What you'll notice is that we get very limited information about the video. So, we'll get the title, we'll get the underlying description. Coming down here, we'll also get stuff like the URL for the thumbnail or the actual link to the video. So, let's go ahead and see how we kind of mash this all together and leverage it. So, the very first action we're going to do is let's first off get the time stamps for the video, right? It's always helpful when watching a video to have timestamps so you know what the heck's going on. So, to do this, we're going to type in bump ups. Click bump ups here. We're going to choose the event of generate timestamps.
[02:36] We hit continue and all we need to do is put in the URL of the video. That's it. So, we come down here embedded URL. You'll notice that we have currently one model here of bump 1.0. That's the model that's going to be analyzing the video for these timestamps. You get to choose your language as well. So, if you don't want an English output, you can ask for a Spanish output, a Portuguese output, etc. And then the style of timestamps is up to your preference, whether you like really short timestamps or very long timestamps for description. I'll make sure I leave a link in the description down below that shows you how to
[03:06] actually integrate bumpups with an API key. So, it's very simple and you get the actions right away. Also to maybe answer some of your questions right off the bat. Yes, this is going to work with other types of URLs in the future. Whether you want to put an X URL there, a LinkedIn URL, or maybe good oldfashioned Facebook. If you're still using Facebook, let me know in the comments down below. Let's get back to the video. With this done, we're going to continue here and test this step. Taking a step back, we actually want to choose the YouTube.com URL, not the embed URL. The embed URL, what that's used for is typically when you just go to a random website and you see that
[03:37] nice little YouTube video like that's integrated or I guess shown in the web page. That's the purpose of the embed URL for us. Make sure you select YouTube.com URL. We are generating timestamps with the bump 1.0 model. And here we go. With the simple addition of just a URL, we got our whole time stamp list here. Pretty cool stuff here. We get two different types of outputs. Whether you want to do a timestamp list or alternatively a timestamp string, just different ways of formatting it. Notice how we separated the data points here, which we'll leverage later. Let's go and add a step here. We'll do another bump up step. Now, what's also cool
[04:07] about bump ups is we kind of have a free form prompt. So we can actually ask the video for our X tweet. This free form prompting just think of it like any type of chat you've ever had of an AI model. This is just a AI model specializer video. So I'm going to add the URL here of the YouTube.com. We're going to keep our models here. We'll keep the language of English. The format we'll do a text output. And we're going to say this. I'm going to go ahead and prompt it and say context. This is a new video on my channel titled and then I can provide the data from the trigger which is the title of the YouTube video. Based on what you see in the video, generate a good tweet that will engage the user to
[04:38] watch the video. Pretty cool. Format max of two sentences, just the tweet, no text before or after. This helps with scaling when doing a ton of different tweets. This should be good to go. We're going to continue here and test this action. What I'm expecting here is the bump model to watch the video, get context of what's actually happening in the video, and give me a very lasered tweet to it. Test that. And the video I uploaded had to do with refactoring code. I know it's it's a little nerdy, okay? It's a whole separate topic that I do on this channel. This is actually pretty spot-on, though. Stop drowning in huge React files. Learn to refractor your code into clean parent child
[05:10] components and/redundant fire store reads in one go. Watch how AI powered refactoring can level up your front-end game in minutes. For context, this is the video right here. And as you'll notice is we'll identify important skill for AI coding refactoring code. But that's it. That's the only additional context this chat had. But you notice in the tweet we reference React files, fire store reads, and everything that's actually identified in the video itself. Pretty cool. So now that we have that, let's go to the last step of the puzzle here, which will be a buffer step. Now, this used to be a Twitter block, but
[05:41] Twitter got deprecated. And as you probably already know, it's called X. Therefore, there's like a whole situation there where in order to access the API, it's a little wonky. So, we're going to use Buffer to make this tweet. Action event, add to Q, select your account for context. Buffer is free to use. Whole another software. It's like a social media scheduler. I know there is a ton of those, but this one connects with X. We're going to choose our X profile here. And here we go. We can actually make our tweet. So, for the first step of this tweet, I'm going to actually take the chat from bumpups here. Scroll down and we'll do the output. There we go. Next thing I'll do is I can add like fixed text here like
[06:12] watch video at semicolon and then put in the link from the trigger youtube.com URL. In addition, we can even add the tweet itself of the timestamps. Now, in theory, we could upload the video through an automation as well and have the timestamps associated with that X tweet. So, I can come over here to timestamps and we'll say timestamp list. That's fine. and do time stamps here. Websuff also has the ability to give you creator hashtags that we could add to this X as well to get more exposure. For now though, for the media, it's pretty cool is we can do image here and we'll be able to add our underlying thumbnail
[06:44] URL. So, the image option available here, we go to add here, new video, and let's get the high res. We're going to do max res here, y'all, because I don't know about you, but we don't need 240p or 120p. We need that 1080p, preferably 4K, but 1080p for the thumbnail. The option we're going to choose here is max res URL. Add to Q or we can just schedule the post right now. So maybe if you want to schedule this later, you can. But if you want to do it right when the video is triggered and your video is uploaded, we'll choose share now. Continue. And let's see this X. We're at my X profile. I went ahead and
[07:15] reformatted the timestamps a little so it's line by line. But for now, we've successfully created a personalized tweet that actually watched the video. Got a thumbnail to the video. We also have a link to the video. All done automatically through Zapier. Lastly, I'll make sure to leave this in the description down below. You can click it. Nice little template for free. That just about does it, y'all. If you felt like you learned something, make sure to leave a like. That's also free. I'll make sure to do more videos like this. I plan on diving a little deeper into Zapier, automations, AI, no coding, all that good stuff. Without further ado, those are two random videos. That is my
[07:46] face and I'll see you in the next