If you make YouTube videos and are not currently using AI to help you do so, you're missing out. With the advancement of AI problem solving tools like ChatGPT and DeepSeek, which was the new platform that has blown up this week, AI has the power to help you make better videos at every stage of the process, from ideation to helping revive old videos that are literally years old and giving them a second life.
So in this episode of Full Time YouTuber, I'll share the top 7 ways that I use AI to help me make some of the best videos I've ever made. ever made recently, as well as three things you should avoid using AI for that will kill your channel before it has a chance to take off. Hey, I'm Ben and I help YouTubers who are entrepreneurs at heart grow their channels into profitable businesses.
This podcast is for creators who want to help their audiences through informational content, not to chase fame or subscribers, even though that may come. But rather to build a business that generates revenue while you sleep. So you can live your dream lifestyle. And right now the tech and AI world has been buzzing with the release of this new platform called DeepSeek, which is kind of like a chat GPT competitor, but done on one 100.
Of the budget and without having to use lots of processing and Nvidia chips and so on. And it's really got a lot of people talking about how fast AI is now going to accelerate due to this technology now being cheaper than ever, as well as more accessible with these companies sharing their code publicly.
So AI is. Only going to gain more and more momentum from this day forward. And this platform DeepSeek seems to be the one that is starting this brand new era of AI. I'm not going to go too nerdy. Cause even I don't understand the majority of what goes on behind the code and the programming of AI platforms.
But in this episode, I want to talk about AI from a creator's perspective. How can you use AI to produce some of the best work you've ever done? Because as YouTubers, there are so many different decisions we need to make at every single stage of the video creation process. And with each of those decisions, you have to choose from literally hundreds of potential options.
And this is where AI can be massively helpful for creators in helping you make the best decisions that lead to you creating the best and most successful videos possible. Now, when I talk about AI and the ways I use AI to make YouTube videos. The only way I'm talking about here is using these research tools like ChatGPT and DeepSeek as those decision making tools to help craft and refine each video and your general channel direction.
So it's. Going in the right direction at the fastest speed possible. I'm not referring to any other kind of AI like generating YouTube videos and so on. Personally, I don't believe that AI is there to make a video for you because that just takes all the creativity out of it. And you can't really call yourself a creator if a computer is doing all the creative work.
But again, AI, in my opinion, is better used as a decision making tool to help you then make all the creative choices. So with that in mind, I want to go through the seven ways I use AI to make better YouTube videos. I'm going to go chronologically across the YouTube process from start to finish. And after that, I'll share three ways you should never ever use AI that a lot of people do use it for, and it totally sabotages their channels.
So let's jump into it. The First way I use AI for YouTube is to come up with winning video concepts. As creatives, often our minds are highly stimulated and we have hundreds of ideas floating around there, all competing for attention. Sometimes some ideas will come to the surface and others will just fade into oblivion.
And every time you commit to making a new video, that's because you made a strategic decision That, that specific topic and concept was the best one that stood the best chance of doing well at that point in time. And this is always an educated guess. It's really hard to know how an audience is going to respond to your videos.
Sometimes no matter how much research you do, you might have a hunch that something will perform and then you make the video and it completely bombs, which is why using a more. Scientific approach to this will save you so much time and so many videos going nowhere. So at any one point in time, I have a list of at least 10 to 20 ideas that I wrote in Apple Notes that I just accumulate over time.
And if none of them stand out to me on any given week, because I try to make roughly one YouTube video per week, I'll put my entire list to ChatGPT and I'll ask of all of these ideas. Which one stands the best chance of doing well, which one has the highest demand, and which one is likely going to be the million view video that I'm looking for.
Then from that point, the AI searches all the keywords and concepts, and will come up with a list of what it thinks are the best performing ideas. And more often than not, it's right. So if you've got multiple ideas floating around in your head, and you can't decide on which one you're going to pursue next, Put them to chat GPT and ask it to rank those ideas from best to worst in terms of the potential they have to reach lots of views.
However, that also presupposes that your ideas are actually good to begin with. I know we all think that our ideas are great, but. Unfortunately, this isn't always the case. I know I've had my fair share of dud ideas, so that's also why you need to go back to chat GPT and ask it to suggest another similar kind of video that's still along the lines of what you started with.
But has a higher chance of going viral. And it will then come up with a list of similar ideas, hopefully some good ones. It doesn't happen all the time, but more often than not, it will give you at least one idea that resonates and that could be worth pursuing. It's always a bit of a process when you workshop ideas, so don't be afraid to give it feedback.
Tell it no, that you don't like some of the ideas, just so it knows what not to do. And try going back and forth with it until you find a clear statement. Okay. The next way I use AI for my YouTube videos is once I've chosen a winning video concept. So I like the general direction of where it's going next before I do anything, before I write a script.
Plan, film, anything. I will always write my video title in advance. It's always a good idea to do this before you film the video because there's nothing worse than making a great video and not being able to find an appropriate title for it. Which then sabotages that video from ever finding success since you didn't start with an enticing title that people were likely to click to begin with.
So I will usually workshop this with ChatGPT and I'll say something like. Give me the top five best performing video titles for this concept. Then it will go to work. It will think about it and come up with the five. Let's say there's one or two of them that I like, but still no standouts because often it makes these titles too long, too keyword rich, or just not speaking in a down to earth way that people can relate to.
So often I'll have to tell it to make it shorter and use more. Everyday words that more people can understand or another very common one is to make it more intriguing because people won't click on your video if the title or thumbnail isn't intriguing. This leads me to the third way I use AI for YouTube and that is for my video thumbnails.
Just like with the titles, I will always workshop this to try and come up with the most clickable thumbnail possible. While, again, a thumbnail is something that you can have a quote unquote hunch on, it's really not worth leaving this to chance. I can't tell you how many video thumbnails I've had to go back and redo.
I think for probably 90 percent of my videos, I've gone with my hunch idea. And it hasn't worked despite the videos being really good. People didn't click them because the thumbnails weren't that intriguing. So I'll ask chat GPT to come up with a winning thumbnail concept. Now I won't ask for multiple concepts because there are so many different small elements that make up a general thumbnail.
So I prefer to do thumbnails one concept at a time, and I'll tell it to come up with. This concept for the thumbnail and to make it as simple as possible, while also being highly intriguing and getting maximum click throughs. Even if you don't word your instructions perfectly, it gets the gist of what you're saying.
And then it will usually spit out a written description of how you should make your thumbnail. For example, I'll read you the thumbnail description it gave me for an upcoming video. I'll be releasing where I compared ChatGPT and DeepSeek in depth. It says thumbnail concept. The background features a dynamic split screen design with contrasting colors, one side in ChatGPT's signature green and the other in DeepSeek's blue.
Place the ChatGPT logo on the green side and the DeepSeek logo on the blue side, both slightly angled toward each other. As if in a face off, in the center where the two sides meet, include a bold white versus symbol to emphasize the comparison, and then incorporate an image of yourself at the bottom center looking intrigued or contemplative to humanize the thumbnail and connect with viewers.
So that's a pretty vivid description, and spoiler, I went ahead and made this thumbnail exactly as it described, and I'm not gonna lie, it's a pretty damn good thumbnail. It's not the exact thumbnail that I I would have made personally, but it was very good. So if you're ever not sure about what to do for your thumbnail, or if you think you have a hunch, but you just want a second opinion, head on to chat GPT or deep seek and see what it thinks.
All right. The fourth way I use AI to help me make better YouTube videos is in researching the topic that makes the substance and the information within the video. So if I'm teaching a topic in depth. I want to make sure I'm covering all the key points that people need to know about that topic. And often when we make videos, especially with very little preparation, it can be very easy to miss things, very important things.
And inevitably there's always that one person in the comments that says. You missed the most important point. It was blah, blah. And that's something you can't undo once you've posted a video. You can't go back and re edit it. That is, unless you take the video completely offline. So finding holes in your structure is really important to do before you make your video.
So it can be complete. In its coverage of your given topic. So the way I'll go about this is I will say something like what other top five things people need to know about x topic, or it might be break down this exact process, step by step in a simple, logical. Bullet point format. Then that step by step checklist essentially acts as your video outline, where you can then teach those points step by step.
Now you want to be careful not to overdo it and try to include a million different points in every video, because with every topic out there, you can go into such granular detail that it totally derails the Focus of the video. So it's important to find the balance of covering everything, but not rambling on too long about unimportant stuff.
And in this research also comes the experience of you as the creator, learning things for the first time. For example, this platform, DeepSeek, I had not known from a bar of soap just a couple of days ago, and then I saw people started posting about it on Facebook, and then YouTube, and I still had no idea what it was.
So of course I checked it out for myself, but ironically, I actually went onto DeepSeek and asked it to describe itself, and why there had been such Buzz around this new platform, deep seek and why it's being touted as such a significant platform. And then I added a very important line to this prompt and that was describe it.
So a 12 year old could understand, because like I said, the start technical jargon often goes over my head. I really need things to be dumbed down to a 12 year olds level sometimes, especially with topics like this. That I really don't understand the nuances of. I'm sure if you had a tech or AI channel, this would just be the most standard thing ever and so easy to understand.
But when a topic isn't your core topic, often you have to go and learn it as a complete beginner. And using these AI tools are a great way for you to learn in a very short period of time, so you can then go teach that same thing to other people. I really can't tell you how many new topics I have learned over the last 10 years, which I had absolutely no clue about before doing the research.
And afterwards I was then teaching hundreds of thousands of people about that thing. So it's really important that you always do your research, especially when teaching others. And again, if you want to save time and not have to read the encyclopedia. on the topic. Using AI is a great way to summarize new topics in the shortest and quickest way possible.
Now the fifth way I use AI to make better YouTube videos comes in the scripting and structure stage. So we talked a bit about knowing the must know bullet points about a given topic, but just because a process has, let's say 10 steps in total doesn't mean you'll.
I've made this mistake far too much over the years and that is, without realizing it, teaching a process out of order. For example, with video editing, usually the process is rough cut, fine cut, final cut. Export. Yet sometimes you'll see a tutorial where they talk about export settings in the opening 60 seconds.
Whereas really most people watching that video are nowhere near ready to export. They want to know what steps 1, 2, and 3 are before they go to 4. So using AI is a great way to structure your video so it's chronological. Because I found over the years that the more chronological you can teach information, the more effective your presentation is going to be and the more people are going to understand it.
Because they're only taking baby steps one at a time to get from A to Z. They're not having to go A, W, X, G, B, M, Z. You get the point. Speaking of like the Scripting and structure stage. While I won't script my videos using AI, sometimes I will need a one off line, whether it's the hook of the video, AKA the opening 10 or 20 seconds, or maybe it's a sales pitch within a video, and I'm not quite sure how to word it to be as effective as possible, that's when I will get ChatGPT to help me write a persuasive and effective line or two.
Especially at high stakes moments within your video, sometimes it's really important to get your wording right for maximum effect. If you're trying to get people to buy something from you, follow your affiliate link, or even just stick around a bit longer, your wording plays a big part in that. So for key moments and payoffs in a video, Sometimes I will script individual lines.
For example, with this podcast, the vast majority of every episode is completely made up off the top of my head and a bullet point outline. However, I do script the opening because I want people to know that, Hey, I'm Ben and I help YouTubers who are entrepreneurs at heart, blah, blah, blah. And I can't always remember that word for word.
So what I did was I wrote it down at the start. I workshopped it a bit, so it could be as effective and powerful and easy to understand as possible. And now all I have to do is read that every time I make a new episode, instead of having to wing it and try and communicate the same thing each time, which is hard.
The next way I use AI for YouTube is workshopping my underperforming ideas and trying to revive them and give them a second life. Because you absolutely can bring back videos from the dead. This has happened to me many times, where I've gone out and made an amazing video, but given it a crap title and thumbnail, and as a result, it performed poorly.
But then what I did was I went back and brainstormed a new title and new thumbnail for the same video that was far more appealing and click worthy. And that has suddenly brought the video to life, and the algorithm picked up that more people overall were clicking on this video, then it gets shared to more people, and then clicked by more people, and that's all it can take to make a video go super viral.
So if you have any older videos that you know are good ones, but they're underperforming, try giving them a new title and thumbnail. Again, you'll want to use a tool like ChatGPT or DeepSeek to come up with better title and thumbnail combinations. Of course, you need to be honest and consistent with what the video is actually about, but you totally can give videos a second life from doing this.
In fact, I do this on a monthly basis. I'll always go back over the last, say, six to 12 months. And often there's videos that I thought would be my next big video, yet there were total crickets. And I will just go and edit a brand new thumbnail with a new pose, new photo, new text. So I can then give that video a better chance of taking off.
So I definitely encourage doing that on a regular basis to give all of your videos a chance to shine. The final way I use AI to make. Better YouTube videos is brainstorming my channel's direction. At any point in time, you are pointing in one direction with your YouTube channel, and you have the option to keep going straight, turn slightly to the left or slightly to the right.
You could even do a complete 180 if you really wanted. But choosing a niche is not something that you just set and forget. Often people make video for years and years, and they end up unhappy because they don't like their original topic anymore, or maybe it's getting really low demand, but they feel obligated to stick with it because that's what people know them for.
I've been going through this a lot over the past year, since I pivoted my channel from covering 360 cameras to now covering YouTube and video creation. And this was a very big decision for me. I had to really think it through all the different aspects that go into pivoting my channel and direction and to help me.
Do that. I use chat GPT to talk it through logically. For example, I'd have deep conversations with it about what my greatest passions and knowledge was, where I wanted to take my business over the next five to 10 years, as well as the sustainability aspect of being able to make videos without burning myself out every single time and having chat GPT talk this through with me, helped me make this decision that I ultimately know.
is the right one. While it was a tough one, you should always be open to pivoting your direction just a little bit in order to get to your end destination, because that's really what it's all about. You need to be creatively fulfilled and it needs to be financially sustainable so you can keep doing it for a long time.
So workshopping your channel direction with chat GPT can help you find a happy medium between the two. Okay, so that's seven ways I use AI to make better videos. But there are also a few ways that I see other people using it for that makes me cringe and back off slowly. Because sometimes too much of a good thing is not a good thing.
And I can say for sure with AI that's the case. Here are three examples of that. The first one is using AI to script your videos word for word. And even though you can pretend that an AI script was your real script that you wrote yourself and convince people of that through a good natural performance, just keep in mind that AI doesn't have a personality.
And it doesn't have your personality. On YouTube, people follow other people because they like them as a person. Either because they've got a good personality, or they're bringing lots of value to them. Whereas when you generate scripts word for word using an AI, while the value might be there, the personality is going to be completely void.
And it's gonna sound like a news presentation that blends in with every other generic video out there. You really want your videos to stand out and people to like you and follow you based on your authenticity. And there's nothing faker than speaking words that aren't your own, saying things you don't believe, using phrases that don't sound normal and natural.
So I would definitely avoid scripting your videos word for word. Like I said, I will script the odd line using a tool like ChatGPT, but that's only after a bit of back and forth. And once ChatGPT gives me what I want, I will then edit that so it's more in my own words. I will never ever read something that was 100 percent written by AI.
Because I think it would be very obvious to people watching and that would harm my reputation. The next AI use you should avoid is using avatars or fake voices. I know it's a bit of a trend right now to do a quote unquote faceless channel, but just keep in mind that if you want to bond people to you as quickly as possible, it helps to Showing your face and speaking your own voice.
Real human beings do not connect with AI or robot voices, no matter how realistic they sound. And fake avatars of people are just a little bit too perfect. They don't stuff up, they don't look away from the camera, they're just not relatable. So if you do use editors like CapCut, and I know OpenAI have their Sora app where it generates videos for you.
Just keep in mind that generating fake video is not going to be the thing that explodes your channel. Especially compared to real authentic you, sharing a real authentic This is another feature I have seen within CapCut where you can literally describe the video you want it to make. So you could type in a prompt like make a video about how to cook spaghetti bolognese and it would then go and search for the process of making spaghetti.
It would write a script, generate the AI audio, and then generate AI video to accompany that just thinking about this makes me cringe hardcore because it's just so devoid of a real human connection with the audience. Not to mention that the video quality is really low. We're not talking cinematic shots here.
It's really just the most. Basic shots of what the ingredients might be. If it's a pasta, it will generate what it thinks roughly pasta looks like, as well as someone putting pasta into a pot and so on, but none of this footage is groundbreaking and a lot of it won't make sense because AI is still in the learning phase and doesn't always understand what not to include when generating images and videos.
And this is why I don't use AI much at all when I'm editing at any stage, even to cut up my b roll and remove the silences. I still like doing that myself because my creative decisions are what make my videos my style, whereas an AI couldn't cut them in the same way, at least not yet. Alright, that's it for the episode.
I hope you enjoyed it. Definitely check out this new Deep Seek platform because I think it's going to be super exciting and totally disrupt the AI landscape in 2025 and beyond. And at the very least, if you're not using ChatGPT yet, I'd strongly recommend it because it's great for helping you make better videos through the seven things I talked about in this episode.
I wish you all the best of luck with your channel. And hey, don't forget, you're just one video away.
