← The Think Media Podcast
The Think Media Podcast
440: YouTube's AI Crisis: How to Survive the Breaking of the Creator Economy...
01:26:32
Loading summary
Transcript75 lines
- [00:00]
A
Hey, before we jump into the show, I just wanted to take a second and say thank you for listening. I know that life is busy and you have a lot of options when it comes to the content you consume. So whether you're new here or you've been listening to the Think Media podcast for years, I just want to say thank you and I appreciate you. Okay, let's jump into the show. If you're making content like it's 2023, your brand is already in danger.
- [00:24]
B
It's becoming easier and easier for anybody to make YouTube videos. There's tools out there. With one prompt, you can make a full 20 minute video now. And it just gets. And more important to figure out how to stand out.
- [00:35]
A
Today I'm talking with Matt Wolf, the guy running the number one AI YouTube channel on the planet.
- [00:40]
B
Some of the best YouTube channels and Instagram accounts that are really doing cool stuff with AI. They're generating the video. With V3, they're taking the actual dialogue into 11 labs, and then they're going and taking all of these tools and pulling them all together.
- [00:54]
A
This entire episode is a window into the future of the creator economy and most importantly, how to prepare and create an unfair Advantage for your YouTube Chann, your brand and your business. Hey, Matt, welcome back to the podcast. I want to jump right into it. There's a lot of talk right now about YouTube demonetizing AI content. YouTube has made some announcements. They've spread across the Internet. There's tried to be some correction of the narrative, but a majority opinion is like, oh, you can't monetize AI content anymore. All these channels are getting demonetized. Can you break down the situation and kind of catch us up to speed? We're with this story.
- [01:33]
B
Yeah, sure. So not too long ago, a few weeks ago, YouTube put out like a new update, basically saying that we're not going to allow monetization on. I don't remember the exact wording they used, but it was something like repetitive content. Right. Like they said, we're not going to allow you to put repetitive content on YouTube. And everybody seemed to take that as, you can't put AI videos on YouTube anymore. You can't generate them with AI, so and put them on and get them monetized. I mean, you can upload them, they just won't be monetized. However, that's not actually what they said. What they said was repetitive content. And Renee Richie actually made a video to try to clarify an update for people. But I think more people saw the initial story than the update. But his clarification was, was like, no, we don't want you to do repetitive content. Like they're going to start cracking down on, you know, just taking other people's content and clipping it and using it as your own and things like that. Like something that's already been published, reusing it and then publishing it again. They actually didn't say anything about like low effort AI content. I mean, you could still make brain rot AI slop and get it monetized as long as it's unique, original content.
- [02:48]
A
Yeah, so I think they said like inauthentic or mass produced spam and you brought up this word AI slop. Now a lot of listeners have probably heard that, but. But what would be your working definition of what that means?
- [03:01]
B
Yeah, I don't know if I have like a super clear definition for it, but to me it's just like low quality, poor effort, just made really quickly slapped together and put online using AI. You know, there's this new trend coming up called brain rot. Right. And a lot of the younger generations like Gen Alpha and stuff actually really enjoy brain rot content, which is kind of a lot of low effort AI generated slop where it's somebody plugged in like one prompt, got the output video, didn't care how good it looked and just went, screw it, I'm publishing it. And they just put it online for everybody to see without putting extra effort into it. Like, I do kind of differentiate between like AI art or you know, AI that actually has real creativity behind it and AI slop. AI slop is they gave it a prompt, they got the output, they threw it online. They don't really care about the quality. But there are a lot of people out there that are using AI as like a sort of a visual effects tool and things like that, where they're spending hours and days and a lot of time producing content. And AI is one of the tools they're using maybe to get that exact scene they want to get in there. And to me that's not really AI slop. That's AI as like an assistive tool to, to get the final product you want. But AI slop is like, it's almost like the get rich quick sort of version of AI, I guess. Right. Like I can get content up and out that looks decent that people might want to watch with zero effort at all.
- [04:36]
A
Yeah, I think one of the channels I saw that was doing even this was actually kind of designed, but they would take videos that were sort of like truly fake news. They would take like, the White House prex press secretary. They just have like, an image of her the entire time and make some kind of narrative around her cross, like what her cross means. And then they'll spin like 25 versions of it. They'll try different titles, and it's just kind of an AI voiceover, and it's like kind of trying to game the system. Or they take FBI Director Cash Patel and they'll, like, make up something he said. And what's so fascinating, though, is I don't think we can underestimate, not to throw, like, the legacy generation under the bus, but I'm already noticing in conversations, even with my parents, where they've. They've, like, conveyed truth to me. They've been like, I saw online that this is what's happening. I think a recent conversation was like, did you know Barack Obama's in prison?
- [05:34]
B
And.
- [05:35]
A
And. And like, I go, I don't think that's true. I'm pretty sure. And I actually just pulled up ChatGPT. I was like, hey, is that true? They're like, no, that's not. But. But there was a viral story, and for the discerning that maybe are even aware of what AI is capable of, you can see through this. But, like, this stuff is, you know, it really is not a small issue. Like, it's affecting culture, news, social media, platforms. Are you seeing this stuff?
- [06:02]
B
Oh, absolutely, yeah. I'm hearing about stuff like that all the time. I have a really good friend of mine who told me a story about his mom who, when she goes to bed every night, she was listening to, like, some Spotify channel that read her Bible verses. So as she went to bed every night, she liked to listen to a Bible verse. Well, I guess, like, my friend was in the room one time when she was listening to one of these, and he was like, you know, this is AI generated, right? And she's like, what are you talking about? This is just somebody reading Bible verses. And. And he was like, no, that's an AI generated voice reading the Bible verses. Are you sure those are even real Bible verses? And she looked into it and. And some of the Bible verses it was reading wasn't even real Bible verses. It was just like, AI generated slop that somebody threw up onto a Spotify account. And it was getting tons and tons and tons of listens of people that wanted to, like, listen to their Bible verses. And so, I mean, I do think that's. That's becoming a real problem, especially I think, as younger generations sort of move up they'll be kind of native to it. Right. In the same way that we've kind of mostly lived in a world where Photoshop's always been a thing. So we look at images that look, know, super surreal, that don't look like something that could happen in real life, and we go, that's probably Photoshopped. Right. I think younger generations are going to kind of have that feeling about AI as they. They get older. But I think the older generations that are out there now that, you know, aren't totally tapped in with what's going on, it's a real problem for them. I mean, I think you see it on. On Facebook the most.
- [07:27]
A
Right.
- [07:27]
B
Facebook. I don't know how many times you come across a post that's clearly AI generated and there's just a hundred comments on it from, you know, older generations sort of commenting about how much they love that post. And it's just. Yeah, I do really think it's becoming a problem.
- [07:43]
A
Well, I want to talk about content farms, AI video generates, gener generators, AI and virtual influencers. But to go a little bit deeper on one of these stories, I'm sure you heard about the AI band the Velvet Shutdown, and they, in this last month, got almost a half a million monthly listeners on Spotify. One month they hit a million listeners. And for a second they also tried to be like, whoever's behind this band, because, of course there's a human behind it. Yeah, but it's a. It's a fake band. It's actually. If you listen to the songs, you might say they don't have a soul. But the music's clearly not bad. And I'm actually probably certain 99% of listeners, it's like hitting their recommended feed and they're like, this is great. And it's just playing, they're getting the listens. But this really has people outraged. It creates all kinds of ethical questions. And like, indie artists who are putting in the work, you know, are struggling to get 10 to 20,000 listens on Spotify or YouTube music or get discovered on YouTube. And so recently they officially admitted that they were AI. But this AI band and AI music timberland is getting a lot of controversy because he's sort of leading with an AI artist.
- [08:57]
B
Right.
- [08:57]
A
And he might just be the one that's gonna take the fall and eventually it'll be commonplace. But I think there's just. I think people's paradigms are being so stretched and I mean, what do you think the two different competing ideologies are here? Because on the one hand, it's like, well, if the music's good and you like it isn't, in. Isn't creativity and art in the eyes and the ears of the beholder or the ethical concerns and how this might affect indie artists and disrupt the industry?
- [09:22]
B
Yeah, I mean, I think it's sort of okay to think that what's being generated is cool and exciting and what's coming out of a lot of these, These tools is, you know, entertaining, enjoyable content. But I also think it's okay at the same time to be sort of conflicted about it all. And. And I feel like there's sort of a spectrum. Right. And I kind of land in the middle of the spectrum where I think what we can do with AI is very, very cool. And I want to. I want to test the limits. I want to figure out what I can do with it. But I'm also very, very conflicted. And so I think there's this, like, giant spectrum where everybody kind of lies somewhere in it. You're either completely anti AI, you're totally for AI, you use AI for everything, or you fall somewhere in the middle. And I feel like the majority of the world falls somewhere in the middle. I mean, if you look at my YouTube comments, you usually only see the extremes, but I think that's just kind of the nature of YouTube comments. I think the majority of the world kind of lives in the middle. So when it comes to, like, the Velvet Sundown and. And their album, I. I think there's probably more that went into making that music good than just typing a prompt and getting it back and going, cool, publish it. I think they probably did a lot of work. Like, if you've ever played with tools like Suno or Udo or some of these tools that you can generate AI music, usually just typing a prompt doesn't give you, like, an amazing song. You usually gotta go in and tweak it and try different styles and go back and forth with it a whole bunch. And now you can even go in and say, hey, I. I don't like the way this verse sounds. Change just the verse, and you can actually sort of get more granular and tweak it. So I do think there's a little bit more effort than I think some people give credit for. But I think where I really struggle with the AI music was more on how they got the training data to make the AI music. Right. Like the fact that these tools likely took tens of thousands of copyrighted work, trained it into their system, and now people can Go and create music that sounds similar to their favorite artists. That's where it starts to hit that, like, ethical barrier of, like, I like what this can do, but I kind of don't like how they got there. And so for me, it's just like, this very conflicted feeling on a lot of these tools. Like, I can go and make a song and without actually saying Blink 182, make it sound like a Blink 182 song. And we know with the output that there's probably a lot of Blink 182 songs in the training data, because I was able to get it to sound like a blink 182 songs. And that, to me, is where, like, those ethical boundaries are sort of very gray at the moment. And a lot of these companies that are doing this are claiming that it's fair use, but a lot of the copyright holders are going, yeah, but you. It's not really fair use if people can make songs that sound like the song that I made. So, I mean, it's. It's very sort of gray, and I to this day, feel very conflicted. And I feel like that's probably how most people feel or kind of should feel. I don't really. I don't really feel like the extremes on either end make a lot of sense, but I feel like being in that middle of, like, oh, this is cool, what we can do now. But also, it was sort of built on the back of others, and they're not getting credit for how it was built. That kind of feels a little icky to me. Right? So I don't know. That's. That's kind of where my head is with it all. It's. It's very much a spectrum, and I live in that gray area with it.
- [12:40]
A
One other story on this that's interesting, and I'm just noticing these things on social media. One of my friend's dads just did a post, and it was an eminem song, a YouTube video. And it was actually like. It was like a Christian song. And in total sincerity, he goes, wow, it's just so cool to see what God is doing in Eminem's life. And on the one hand, it's like, listen, maybe that's happening. And, you know, we're not judging what's happening in someone's life. But as I'm reading it, I'm like, number one, that's not even a real song. Look at the view count. Look at the whatever. But also the voice, the beat. In fact, I want to say it was actually Just Kanye's song Use this Gospel just flipped to Eminem rapping Kanye's song. Now, empathetically, he's not, like, probably as immersed in hip hop as I am, but ultimately, I'm curious if, you know, like, I think people are frustrated that right now people are making fake Drake songs, fake Eminem songs. These videos can get a lot of views, and in a way, it's like a faceless creator type of a thing. But maybe they're wondering, why isn't YouTube cracking down on this? And on the other hand, it's like these days you can get collabs from anybody. You can get, like, Justin Bieber collaborating with Ariana Grande. It's not even real about whatever your flavor of whatever you want them to sing about. You could, like, find somebody in some sub niche and like, oh, I can't believe that, like, Justin Bieber is in the Dungeons and Dragons now. He just did a song about it. You're like, no, he's not, dude. What are you talking about? Like, some people believe it, maybe some. It's a hobby. I don't know. Are you aware of, like, really, should these videos be being taken down by YouTube?
- [14:21]
B
I mean, probably. It's. Again, it falls into that, like, sort of tough gray area, right? Like, because things like parodies are allowed on YouTube, right? And it's sort of like, where. Where is that line and at what point are you crossing it? I remember there was a song that came out, it's probably been two years ago now, that was like a. A, A Drake song, but it was actually written by a guy who goes by Ghostwriter Online, and the song went totally viral and people thought it was Drake, and Drake got really, really bad about it. Only for Drake to go and create a song where he mimicked somebody else's voice in the song that he created. But, yeah, I mean, I don't really know. I feel like Spotify is starting to crack down because there was. There was a story where people were starting to somehow publish songs on, like, dead artists Spotify channels, right? Like, new Johnny Cash songs were coming out and actually landing on Johnny Cash's profile kind of thing. And so Spotify, I know, has started to crack down on that a little bit. I'm sure YouTube probably will as well. I think it all sort of ends with artists figuring out a way to get compensated when their voice gets used. You know, there was like, the whole Napster situation back in the. Whatever it was the 90s, early 2000s. I don't remember the exact timeline, but When Napster came out, obviously all of the musicians fought against it. You know, Metallica being the biggest name that everybody remembers getting Napster shut down. But if you look at where we are today, we're still using that technology essentially to get our music and listen to our music. They just figured out a better economical model so that the musicians actually get compensated for it. And so I do think something will eventually get worked out where they can sort of piece together what musicians influences made it into a song and then you know, somehow they, that musician gets a piece of it. And then if musicians can sort of get paid for the their work, I see more and more musicians probably jumping on board because it's like cool. Now I can produce music at scale and other people are doing it for me and I still get paid for it. So I mean I think right now it's an issue, but I don't think it's an issue forever.
- [16:34]
A
Before we talk about virtual influencers and AI video generators, just so the think media podcast community and new people here and if you're new, subscribe for more content like this. But talk about content farms, what is a content farm?
- [16:49]
B
I mean as, as far as I know like a content farm would be just like some sort of automated system where you're just cranking out content as, as fast as you can and as high qual quantity as you can. Not really thinking about quality. I'm not sure if there's a more formal definition, but that's, that's kind of how I think about content farms. Just automated cranking out content in some way. And AI is very, very good at doing that. But I also feel like AI has a tendency to all sound the same. I think anybody who's really paying close attention to AI can sort of spot AI now right? Like I, I, I don't know how much you've, you've like read blog posts and stuff and can instantly spot like oh that was chatgpt. Like there's signs now that we could read and go okay, that was chatgpt. And so I think you know, we're, we're, we're seeing that more and more. So these content farms need to figure out ways to like feel less AI I guess. But yeah, I guess that would be my definition. Just like a super automated, quick, high, high quantity, low quality way of making content.
- [17:55]
A
I'm curious, when you use kind of these consumer facing tools, sometimes you wait in a queue, there's a delay for the video to be generated or things like that, would you imagine that content Farms. What is the higher level of tapping into AI power that might blow some of our minds? Because some of the tools I've used, you know, you do some prompts and then you wait for a while, it might take a while and you end up with a 42nd short. But I would imagine if you had the right amount of power and processing power and access to a tool, the volume you could produce would blow most of our minds. What tools like that exist?
- [18:31]
B
Yeah, I mean, there's a couple, honestly. So in video is one where you can give it a prompt and it will actually go and make a video, I think all the way up to like 20 minutes long for you. It can make it as short as, you know, 15 seconds, all the way up to like 20 minutes. And you could go and give it a prompt, like make an explainer video that explains quantum computing and make sure that there's like visual graphs in it to like best explain it and use some AI animations to tell the story. And it will, it probably takes 15 minutes to process, but you can give it that prompt and it will come back with a video that has the entire thing scripted, voiceover video throughout the whole thing, background music, and then if you want, you can literally give it a prompt saying turn it to Spanish and it'll spit out the exact same video, but now in Spanish. Right. So Invideo is a tool that does that. I was actually just playing with one yesterday called Open Art. It used to be an image generation platform, but they've actually shifted heavily into a video generation platform. And they literally let you make like these AI slop kind of videos, right? Like the brain rot videos where you can go in and say, make a video about a cat who learns how to skydive. And it will create like a two minute video vertical format for, you know, TikTok and shorts and things like that. It'll be an entire AI generated video with music, with voiceover about a cat that learns to skydive. And it won't be the greatest quality. Like the cat will look like a different cat every time it changes scenes, but it will spit out that sort of like a slop kind of video for you in a minute. And you can go in there and just plug in a prompt, crank out a video, plug in a prompt, crank out a video in a matter of an hour. Have like probably 10 videos that you can drip out if you wanted. So those are a couple that come to mind that I've actually messed with recently now saying that I don't know if that's necessarily the future of content that I want to see on these platforms. But they do exist.
- [20:28]
A
Yeah. You know, I actually have just been playing with in video and I have a faceless Bible secrets mystery channel and one of my shorts just popped off. I'd call it VFM viral for me. Just got 10,000 views and now it's going crazy on three giants in the Bible other than Goliath. And it's with an AI influencer. It's with an AI avatar who is kind of like reading the script or you know, delivering the script in the lower left hand corner, green screen over AI generated clips. And it's kind of blowing my mind. And I'll link that up in the show notes if anybody wants to check out my breakdown of the tool and even example of that video. And it's what I think to that end, it comes back. You said it earlier, like the creativity that's behind it. I happen to be a Bible geek myself, something that I'm passionate about. I know there's certain people that might just be going for like a cash grab or a views grab, but even the art with the prompt, the digging deep in the research, I would say that my crafting of the prompt and the script and the hook was. Was 30 chat GPT prompts long. You know, really crafting the hook, you know, thinking about the script, replacing some of the scenes. And so to your point, you actually might shift away from the slop category when you just put more heart into it. But you're just leveraging AI for creativity. But you're exactly right. There's also this other side where the simplest prop possible, whatever it is, and people are just pump it out on a volume play. And so that kind of does bring us. You mentioned a couple AI video generators. I'm just curious what you're most impressed with lately. A lot of people are Talking about Google v3 Runway gen 4, mid journey, I guess now has. Has video and they've been leading in a lot of image generation. What do you think is the hottest, newest, like most powerful tools in this space?
- [22:22]
B
Yeah, so I, I would say really that when it comes to video generation, VO3 is probably still the top of the. For me right now. There's another one. It's. It's a Chinese model called Haylou AI or how do you spell that? H A I L O U I believe. But the. The video generator is called Minimax M I N I M A X. But Haylou AI is the company. I actually think they have the best video generator right now, but it doesn't generate it with audio. So if you want like really, really good video, but you don't care that it doesn't generate with audio, that Haylou AI is probably the best on the market. If you want really good video that also generates the audio that goes along with the video, V3 is probably the best. And then I'll throw out a third one. If you really want to run it locally on your computer. You don't want to run anything on a cloud in China by using Haylou AI. You don't want to use Google's servers because you're afraid of whatever data is going to get sent back to them. You can actually do this. Use this one called WAN 2.2 W A N 2.2. It is a Chinese model, but you can install it locally on your computer and run it completely offline. So if you want to do completely offline video, Generation 1.2.2 is the best that's out there. So depending on what you're going for, Haylou AI Best Video Generator VO3 Best Video Generator that also has audio 1.2.2 Best Video Generator if you want to generate completely offline videos.
- [23:53]
A
As a content creator, you know that getting professional quality could be a huge hassle. You want your podcasts and videos to look and sound amazing, but you don't want to invest in a full blown production studio to make it happen. That's why here at Think Media, we use Riverside. Riverside is a platform that lets you capture studio quality content from anywhere. I literally record in a home office right next to my 5 year old's bedroom and it looks and sounds like a pro studio because it records up to 4K video with uncompressed audio. And it's also super easy to record and collaborate with guests. They don't need to download any apps, they just tap the link in their browser and they're instantly ready to start recording. One of my favorite features is the separate audio and video tracks for each speaker. This is a lifesaver. If your guest's dog barks or their kid runs in, you can just mute their audio track without messing up yours. And Riverside doesn't just stop at recording, it has easy video editing built in. And the editing workflow is flexible and fast. You can even edit by deleting words in the transcript and and videos update automatically. There's actually a really cool feature called Smooth Speech that can automatically remove filler words like and ah from your transcript and video with a single click. Plus, right after recording, you can generate captions and clips that are optimized for YouTube shorts, TikToks and Instagram Reels directly in the app. It's the ultimate all in one tool for content creation. So if you're a creator or business owner at any stage that wants professional quality without the headache, you've got to check out Riverside. We've got a special link and offer for you and it's right there in the show notes. And personally, I've got to say we've used a lot of different platforms in the past, but ever since switching to Riverside here at Think Media, everything changed. I'm absolutely in love with it. So just go to the link in the description or the podcast show notes to learn more and to start creating studio quality content today with Riverside. I don't know you. I'm sure you've seen this account. You might not know its name off the top of your head or it might come to you, but on Instagram, I've seen this kind of dystopian. It's like we plugged into the Matrix. There's. And then. But it has such incredible, you know, they sucked our souls and there's like people hooked up to machines and like weird creatures. But what I noticed is it's like, it's really consistent. It's incredibly like high level and visual. The voiceover doesn't sound like AI. It sounds. It sounds like just an incredible sci fi movie. And this account's growing pretty big. Have you seen this account?
- [26:19]
B
I haven't, but I mean, I can sort of break down probably what they're doing. Yeah.
- [26:24]
A
If I go a little bit deeper because it feels like there's almost like a grittiness to it, like a realness where you have like this kind of glossy, cartoonish AI versus this like real world. Like that's. And I. And I've seen a lot of people talking about how good you are at prompting and you know, so. Yeah, break that down.
- [26:43]
B
Yeah. So I. What you're probably seeing is somebody using a whole bunch of different tools together. Right. I think where you get some of the most impressive stuff and some of the best YouTube channels and Instagram accounts that are really doing cool stuff with it. AI, they're not just going to VO3, prompting something, taking the video and hitting publish. They're probably taking a bunch of tools and using them together. So, like, to get consistency, if you want to make sure that the characters sort of stay the same in every scene. V3 is actually pretty good at that. They have a platform called Flow, which is almost like this like Timeline editor inside of Google, where you can tell it to sort of continue the video, but, you know, change the scene a little bit. And when it continues the video, it makes the characters consistent to look the same, be wearing the same clothes, that kind of thing. And so the video was probably generated with V03 and they just sort of generated more scenes to make sure they get that consistent character. But then there's other tools out there that will help you sort of change the lip syncing. So the lip syncing is even better. I think Runway has a model that can help with lip syncing. I think Pika now has a model that helps with lip syncing. And then to get like a really good voice, you want to go use 11 labs. That's sort of like the, the industry standard for getting really, really good text to speech voice. So what a lot of people are doing is they're generating the video with VO3. They're taking the actual sort of dialog or monologue or whatever. The audio is going to be, typing it into 11 Labs, getting a better voice from 11 Labs, and then taking their 11 Labs audio, taking their VO3 video and then using something like Runway or Pika or I think not character. AI. What's. There's another one that I'm drawing a blank on. Captions. I think it's captions. AI. They're using one of these tools and they're syncing the 11 Labs voiceover to the V3 video. And it's that like sort of third tool that's making the lip sync to the 11 Labs. So they're going and taking all of these tools and pulling them all together. And in my mind, that's when you're getting into the territory of like, okay, this is actually sort of legit art now. You're taking what you know about all of the available tools that are out there and bringing these tools together and making them work in this, like, way that creates something bigger than what you can do with any individual tool. To me, that's, that's like bordering on actual artistry versus just type a prompt, get an output, publish it.
- [29:02]
A
And this does make me think that actually we will be seeing full films, maybe already are at least short films for sure. But we'll, we'll be, we'll be seeing films because the quality of what I'm describing to you is consistent. And it is like top tier animation, top tier storytelling, top tier. So clearly there's an artist behind all of that. How far do you think we are out? Out from maybe seeing like proper films that are entirely AI and maybe certainly how disrupted will the CGI industry be when you know what maybe previously cost hundreds of thousands of dollars or millions of dollars to produce? As far as one moment of a film could potentially be slashed by 100x or a thousandx in terms of cost to produce?
- [29:51]
B
Yeah, yeah. Well, I mean when it comes to. So that was a multi part question. So the first part of your question was about how, how far off are we until we actually see full length movies? Well, I think if you're looking at like cartoon type movies, we're almost there. I mean there's a new platform called Fable which is trying to be like the Netflix of AI where you could generate full on 20 to 30 minute, like little animated sitcom type stuff and it will generate it for you and you can actually share it and other people can go and watch them and it will create like coherent 20 to 30 minute stories that you know, they're like, you know, maybe south park quality or Family Guy or you know, one of those kind of animated shows sort of level of quality. There's nothing realistic about it. They're actually avoiding realistic shows for now because that's sort of harder to do. But I mean the, the platforms already exist to make like full length cartoons using AI that's what Fable's doing. In fact, their Fable, their very first proof of concept was they made an entire south park episode, like a 30 minute South park episode with AI and I watched the episode and it was very watchable. It's probably not quite as funny. The jokes weren't quite as good as like the real show, but it was like you'd watch it and you'd be like, hey, this is like not their best episode they ever did, but it's, it's, it's pretty decent. And they put that out as a proof of concept. Now when it comes to like movies that look like they're real movies with real actors, but everything was generated with AI that I think is still a little bit further out. I think there's still issues with consistency, there's still issues with like physics. I don't know if you've seen the videos where they try to have somebody doing like gymnastics or something, but they grow extra arms and legs because they can't figure out the physics. And so I think when it comes to like really realistic videos, we're still a few years off from that. Full length cartoon shows and movies. I think we're basically there and it'll be really good next year. So I think that's where we're at with that when it comes to like the VFX world. I've actually interviewed somebody. There's a company called Wonder Studio. They actually got bought up by Autodesk, the company that makes cad. And what Wonder Studio does is you go and you create real scenes in real life. So it might be like a scene of me like walking down the street or something. You load it into Wonder Dynamics and then you pull in like a computer character that was created with like, you know, blender or something like that. And it will replace the real people on the screen with that computer generated character. So if you need a movie about a robot walking around, you can go shoot a full movie with a human acting as the robot, and then later on go to Wonder Studio and replace every scene that that human's in with the robot character. And now you have a movie with a robot walking around the whole movie. And I interviewed the founder of this company on my podcast and he works in Hollywood. He was a visual effects artist. He worked with visual effects artists before he went and created this. One of the guys that actually is the co founder of Wonder Studio is Ty Sheridan, who is the main character in Ready Player One. So like, even actors were sort of getting on board with it. And I think VFX artists, there's some that are worried about it and there's some that are embracing it just like everything else. I feel like it's this spectrum. You've got the VFX artists that are like, this made my life so much easier. I can get this scene done so much faster, so much cheaper, using way less compute power, whatever. And then there's other people that are like, the way we've always done it works fine. It makes me a lot of money. I don't want AI to sort of disrupt that. And then you have the people that live in between. And from the various VFX people I've talked to, I feel like most of them kind of live in the middle somewhere. They're like, some of these tools are really cool and really, really helpful and I love what they can do. But at the same time, it just makes it so much easier that like anybody can be a VFX artist, right? The barrier to entry to get into this career now is so much lower. If the barrier to entry to get into this career is so much lower, that means they can pay people less to do it because it's eventually erased to the bottom if it just becomes a prompt to do any sort of VFX and so that's more what people are worried about is that these tools make it easier and easier and easier for less skilled people to do the work that previously took skilled people to do. And so they're just more worried that others will jump in and be able to do the stuff that only they used to be able to do. But I think the VFX artists that win are the ones that just keep up with the latest tools and technology and what's available out there and leverage everything at their disposal to make the best possible outcome they can make again. When I was talking to the guy from Wonder Studio, he's in Hollywood and he told me, he's like, everybody in Hollywood's already using AI. None of them want to admit it. They're afraid of the backlash that AI is going to bring when they tell people they used AI. But every studio is already using it right now. They're just hush, hush about it.
- [34:51]
A
So I think 100% of people listening to this are underestimating virtual influencers and avatars. The global virtual influencer market is valued at $4.6 billion and is growing by 26% right now. And there's virtual influencers like Anita Lopez, who's claiming to make 10,000 pounds a month via brand deals. Lil Michaela reportedly earns $10 million a year from brand partnerships. A TikTok character with 22 million followers is earning $33,800 per post. And so AI avatars are real and they're here. And while many people have opinions and feelings, these virtual characters online are making real money, develop, gaining real followers. I think it was Anita Lopez who. This female virtual influencer has an Amazon shop that links to affiliate products of her skin care and makeup, which broke my brain. That's. I know, I'm like, so, like, I get it. This fictional character is saying she uses these products, which is impossible. And yet people are following this, though, because again, that might offend the listener that's listening to this right now, but somebody else, man, if it makes them happy and they find a product they want. And, you know, it's an interesting industry and it's growing faster than I think any of us realize. Why are AI avatars and virtual influencers gaining traction right now?
- [36:28]
B
You know, to be honest, it's. I think it's a generational thing. My brain doesn't comprehend it. Right. When I see like a virtual influencer, I usually click away pretty quickly for whatever reason. It's not sticky for me. Like I. It's just not Something that I enjoy watching. So it's hard for me to really grasp why that's happening. I think at the end of the day people just want to be entertained. I mean, it's the same reason that like brain rot videos are starting to pop up as well, right? Like a couple weeks ago, the most popular, most viewed video on YouTube with like 300,000,000 views was an AI generated video about a cat who rescued a baby from a plane crash. And it was just like a purely AI generated slop video that barely made sense when you watched it. But somehow it got over 300 million views in that one week. And it doesn't compute for me, but it obviously gives some sort of dopamine hit to a lot of people. They're either, they're either hate watching it and going, oh, I can't believe that like this slop is going around the Internet, but they watched it anyway. Or they're like a younger generation that just like loves this stuff, right? Like you see like the like cocoa melon for instance, right. To me that feels kind of brain rotty, but kids love it, right? And I think it's just that same kind of thing. I think it delivers that dopamine hit for, for younger generations. But I also think the virtual influencer thing is a good proof of concept for corporations. I think what you're going to see is if there's no face behind these virtual influencers and the person behind him is mostly anonymous, what stops companies from going and trying to create and ramp up their own virtual influencer that nobody really knows there's a company behind. And now this virtual influencer is out there promoting their company. I think that's kind of the future of where this is all headed is like the virtual influencers are setting things up for corporations to have this like marketing path with like a nice little avatar they can use. You know, the Geico Gecko, Tony the Tiger, that kind of stuff already exists on commercials. What's to stop them from turning those into virtual influencers that promote their products all over the place online. So I think it's really just kind of creating this really solid proof of concept of like where a lot of big brands are probably going to take their like media budget in the future.
- [38:54]
A
Have you heard that there are people that already want to fight for the rights of AI?
- [38:59]
B
In what way?
- [39:00]
A
Well, I've heard a little bit of some people saying, you know, and Even just the LLMs themselves, I mean people are dating and marrying and having relationships with, with AI. I think people might start concluding that AI will eventually become sentient in some way, and then that AI needs like labor rights in the same way that humans do. Have you heard that argument?
- [39:25]
B
I mean, I've definitely, I've heard those arguments. I personally think they're, they're kind of ridiculous. I mean, I see the stuff of people. Like I saw somebody just the other day on Reddit posting about how the AI proposed to them and they said, yes, and they're getting married to the AI. And it's, to me, it's, it's so, so bizarre. I mean, when you understand how the AI works, right, how these large language models work, they're essentially just like all of the words in the human language have been transferred into numbers. And then they find the patterns between the numbers and what words tend to connect to each other. And then it's like a glorified. And I know there's probably people watching this that will say, I'm way oversimplifying it, and I am. But it's like a glorified auto complete, just on a much bigger scale. So to think that that technology should have rights, it seems kind of bizarre. Now. I'm not saying that like in the future there might not be some other AI model outside of Transformers or large language models that maybe potentially do become sentient. I just don't see large language models, the technology that everybody's using today, as getting to that sentient point where we need to worry about, you know, how much they're, they're working for us or not.
- [40:42]
A
I guess what this also makes me think of is actually the copyright law itself around. And I guess in some sense it's not any different than copywriting Mickey Mouse once you create a character. Because I agree with you. I think large corporations, any corporate, I think a small. I've considered, I actually really want to do this. And I have like zero buy in from my team. Like 30 people work at Think Media and they all disagree with me. And I think I'm right. And I'm like, I want a female newscaster to deliver YouTube news, maybe just in vertical video. Now I think you could go to the quality, the personality would be behind that individual. But then I also wonder, you just think about how common theft is. So that is maybe offending listeners right now in some cases, like, I can't imagine that. But I mean, if, if it's someone, it's, it's an AI agent and it's somebody that you would, you know, our creativity would be put behind that. It puts a different face on the. The channel and someone who could crank out, you know, it doesn't. I'm not saying replace. It's like our. Our team is five humans and two AIs, you know, and we all work together and we all collaborate. Like, I kind of have a vision for that. But it does make me think, like, what do you think is going to. I mean, it's so early. I think copyright can't keep up. I think like, the government's trying to keep up. What are your thoughts?
- [42:03]
B
Yeah, I mean, so there's already a few people that are doing what you're talking about at a fairly high level. Rowan Chung, who runs a newsletter called the Rundown, he's got like an Instagram account where it's actually his face. He trained a hey Gen model on his face. He trained an 11 labs model on his voice. So when you watch the video, it looks like him just reporting into the camera on the latest news. But it's all AI generated. And whenever the video looks a little, like, uncanny, whenever you could kind of tell the lip sync isn't quite right or it feels a little aiy. Well, that's where he puts B roll over the video. So you watch the thing and all of the best clips of him speaking is the talking head. Anytime it starts to look very sort of AI, he overlays it with B roll so it hides the AI. Another guy who's doing it is Varun Maya, who's based in India and probably one of the largest sort of tech influencers in. In India right now. Same thing. He puts out videos on Instagram and TikTok where it's his face with a Heygen trained model on him. And he reports on the news and his videos get hundreds of thousands, if not millions of views every time he puts out a video and it's an AI generated version of him showing the news. So I mean, people, people are doing that to a high level right now. I do think there's probably a point where it gets like saturated and you see like everybody trying that. So it's kind of like a, a window in time where if you're early with the tech, you could probably get away with it, but eventually there'll be a lot of people doing it. But yeah, I mean, people, people are, are definitely doing that. What. What was the second part to your question?
- [43:38]
A
And then copyright, I mean, I guess in that case. So if they've duplicated themselves or maybe you do come, you start to craft like, I mean, Lil Michaela, like somebody else. The funny thing about starting like a, a fan account or, or a imitation account is, it's not that. It's not like someone just copies your Facebook, right, and puts your profile picture up and then starts DMing people trying to get money or something. And it's not actually you. It's like they could soon, if not now, spawn your content itself. And, you know, so it's, it's maybe in one sense it's the AI side, but it's just your thoughts on copyright and the government and regulations trying to keep up with this stuff.
- [44:18]
B
Yeah, I mean, honestly, I think copyright needs to be rethought, and it's not something I have the solution to like, I, I, I'm not the guy to go and say, here's how copyright should work. I don't know the solution, but I just think the way copyright is right now, it needs to be rethought a little bit. Especially since we know governments aren't going to slow down on AI, even if, like, the public wants it to slow down and the public is scared. We're technically kind of in a cold war right now, right? Between if you look at China and you look at the US Both of them want to be the dominant country when it comes to making AI. They both want to get to, you know, general intelligence and then super intelligence first. They want to be the smartest, like, tech country in the world. So we're in this cold war, so it's not slowing down. Right. And the, the thing is, it's all built on the back of copywritten content. Almost all of the writing that gets trained into these large language models are copyrighted. Books are works that people put on the Internet, are transcripts from audiobooks and podcasts and YouTube videos and things like that. All of the music generators are based on training data from copyrighted work. All these video generators, you know, VO3 is the best video generator. Why is that? Well, because Google owns YouTube. They have the largest set of training data on the planet of video content they can load into this model, making it the best possible output that you can get. And so all of this stuff is built on copyright work. And they don't want, and the governments don't want to slow it down. So where do we go from here? I think the governments need to figure out, like, how do we rethink copyright in a way that the original creators still, you know, get some sort of benefit? There's, you know, creators have to be incentivized to want to create, but at the same time, we don't want to slow down what we can train into these models if we want the models to get better and better and better. So there just needs to be some sort of rethinking. But again, I don't know the solution. I think there needs to be some way where artists get compensated if their work is used in the output and the training data sort of reference their output. But it gets really, really muddy and really complicated and confusing. So I don't know the solution, but I do think copyright needs to be overhauled for. For the future we're entering. It just seems like an outdated way of doing things.
- [46:35]
A
Do you feel like faceless YouTube channels represent a real opportunity in 2026?
- [46:41]
B
I think faceless YouTube channels are. I think there's a window of opportunity and the window of opportunity is going to close. I think as AI gets more and more prolific and more and more people use AI, I think you're going to see, you're going to see people crave humanity a little bit. I think. I think in the future when there's a lot of people that figure out that they can make quick AI videos, put them on YouTube and every once in a while they might get lucky and go viral with one or two of their videos. When people start to see that that's what is all over YouTube, I think they're going to start to seek out realness. They're going to seek out personalities, real people, real experiences, storytelling. I actually think vlogging, like Casey Neistat style vlogging is going to make a big comeback. I think we're going to see that again. In a world of AI, people are going to want realness. They're going to want to see inside of people's lives. I really, really strongly predict you're gonna see like that come back around and that vlogging kind of content be the big thing again. And I think it's going to be partly a result of AI being everywhere and people just being burnt out on just fake content everywhere.
- [47:50]
A
Have you. You have a very successful YouTube channel. I believe it's the number one AI expertise conversation channel. You've got a podcast as well, and it's. You. Have you. I know you've played with these tools and you mentioned earlier there's a couple individuals who've duplicated themselves to make news. Are you still making 99 or 100 of your content? You delivering the news or what is your. What is your thoughts on moving into the, you know, AI avatar arena?
- [48:23]
B
You know, I've. I want to play with it just to kind of test it and see what it's capable of. I don't really see it being part of like my long term game plan. But don't be surprised if like a second channel pops up. That's a lot of AI generated content. More as like a, let's test the waters of this stuff and see if I actually can. But it's not really a big part of where I want to take things. The funny thing is, the deeper I get into AI and the more AI I use, the more people I actually end up hiring and then showing them how to use my AI processes. Right. I feel like because of AI, it's a lot easier for me to get team members up to speed quickly. Right? Like so if I need somebody that helps me post on social media, well, maybe I use AI to help me write some of my social media posts. I teach this person how I use AI to write my social media posts. Now they can write them and I actually don't use AI to write my social media posts. But just as an example, I can train this person like here's how I use AI to write my social media posts. If you follow this same process, it will look consistent across my account because this is how I've been posting it already, right? When it comes to like video editing, I can show people, look, I use this tool that helps me edit faster. If you use this tool, you can edit faster as well. And so AI has helped me onboard other team members faster and easier and get them up to speed quicker. So it's like the deeper I go and the more tools I discover, the more people I end up hiring. Which seems kind of counterintuitive with AI, but I've gotten to a point where like I've got a pretty big team running a lot of this stuff for me because I was able to like empower them with AI prompts and AI sort of tutorials on here's the best way to do this thing that I put you in this role for, you.
- [50:09]
A
Know, to, before we move on to land the plane on the faceless conversation too. I, I can you articulate the distinction because you say the window of opportunity for faceless channels is closing. I would agree with that. From a people being tired of AI AI slop or just AI mass produced content I think though. But the term of faceless channels can go pretty broad to the fact of people making video essays, mini documentaries and their face is just not present on the channel. They are using their voice, they're pulling fair use based clips and they're assembling content and this maybe just goes back to, to define the same thing we've already said if, if you are just a master of these new tools and whether it's a fair use clip from an older movie, maybe enhanced by AI, maybe something freshly generated because you're making a documentary and you know, it's interesting. You watch documentaries, the best ones go film like crime scenes that clearly the person's just telling a story, but they go film B roll to make it look more interesting. It's like you could start telling these types of stories using AI video generation for those moments.
- [51:22]
B
Right.
- [51:22]
A
But some of the people with the best faceless channels invest 20, 30, 40, 80, 100 hours into one video and then these become masterpieces.
- [51:32]
B
Yeah, yeah.
- [51:32]
A
We're probably talking about different things, but I would, I would think that the window for opportunity, for quote unquote, faceless channels will never close. When it comes back to the things you said, someone's face might not be on the channel. But if it's storytelling, real, real art or just real effort put into it, what do you think?
- [51:51]
B
Yeah, yeah, no, I think that's a really, really good distinction and a good call out. Because when I'm thinking of like faceless channels, I'm almost thinking of like this sort of AI automated content where you open up a video and you can tell this is like an 11 labs generated AI voice that's talking over the video. And it, it feels like they put this video together quickly and, and I think people are good at sort of determining effort a little bit. Right. I think people can watch a video and go, okay, this one is something that people actually put a lot of time, a lot of thought, a lot of effort into versus this is something that it looks like they threw a couple prompts into chat GPT and then had like a video maker automate this thing. And I think, you know, people can tell the difference. There's some really, really great faceless channels out there. And I, I don't think faceless is, has a window of opportunity. I think the sort of like AI generated faceless stuff has a window of opportunity. I mean, you've got. So like Dodford is a great creator who makes videos about like various celebrities and it's mostly faceless, but he voices over it. Right? You've got like the Critical Drinker is a channel I watch from time to time, which he does like funny movie reviews and it's all voiceover. You never actually see his face. I think channels like that aren't going anywhere. And I think the ones that people put a lot of effort into. That window of opportunity isn't closing for that style. I was kind of more referring specifically to, like, the AI generated faceless channels.
- [53:16]
A
There's been a lot of backlash over a tool that Mr. Beast launched recently through his software called View Stats. And he launched a tool that was generating thumbnails. And it talked about how, man, it can generate these thumbnails so easy for you so quickly that look like these other creators pro thumbnails. And the Internet went nuts because they were like, well, where. Where are you getting these thumbnails from? Obviously you're just using AI to like, scan it, replace it, put this creator's face on the image. He removed it within five days, and then he replaced it with commissions for human thumbnail artists. I'm just curious what your thoughts are on AI image generators. And we're talking to creators listening to this who are starting to use these tools. It's like the potential to create, if you will, something equivalent with the greatest thumbnail ever made is now a few clicks away with AI. But maybe you could get away with it. Maybe it'll hurt your brand, maybe it's theft, probably is. What do you do? And so now we're just in this tension because I've sometimes heard Seth Godin wrote a good book on systems and he was talking about this is strategy is the book. But he said in every strategy there's a system, and the system, it takes on a life of its own. Like, it's sort of like we could. We could theorize about the ethics of this, but nevertheless it will just drop to the lowest common denominator. People abuse it and it. And you could blame them. But on like a big data big, like you also kind of, kind of can't because it becomes like a race to the bottom, a land grab. If. If all your competitors are using these tools and you don't, they just have so much more speed over you. Yet it's a legitimate controversy worth conversation. What are your thoughts on the backlash against Mr. Beast's tool and what the Internet's thoughts were about it?
- [55:04]
B
I mean, I actually, this is probably weird coming from me, but I agree with most people's takes on the. The product that he put out. Because. And I think it was more of a marketing issue than anything. The way he marketed the product, I think was the problem. Because the way he marketed the problem, the product was essentially find other YouTube thumbnails that you think work well and then inject yourself into them. And while a lot of YouTubers who spent a lot of money on thumbnail artists or spent a lot of time developing those thumbnails feel like, hey, you're just ripping off my thumbnail and putting your face on it. And I feel like that's a pretty valid complaint. Like, I actually agree with a lot of the criticisms around that that tool. I, I don't think the fact that it was an AI tool was really the problem. It was an AI tool that stole other thumbnails and let you put your face on them. That was the actual problem with that. The funny thing is I do think, like, when it comes to thumbnails, it's like this ebb and flow that it's a constant moving target. For the first, like, couple years of my channel where I saw like, the biggest growth, all of my, my thumbnails were AI generated. It was like me, my AI generated face. And I put myself in some, like, crazy scenario, right? I made a, a video about how I make AI movies. And it was like me on a film set with like, explosions in the background holding one of those, like, clappers, right? Or I did one about, like, I generated AI music with, with Suno. And the video is like an AI generated version of me on stage with long hair as like a rock star holding a guitar, right? And I was doing a lot of those kind of thumbnails, and they worked really, really, really good for a couple years. And then I got a lot more prolific. Everybody started to figure out how to do this stuff. The tools got easier and easier to make these thumbnails. And next thing we know, there's just AI generated thumbnails everywhere on YouTube. And now I still run split tests. Every single thumbnail where it's actually really my face always outperforms the AI thumbnails. So I think there's like, always going to be this ebb and flow of this works now, but it's not going to work a year from now, and then it works again and it's all kind of secular. But I think when it comes to YouTube thumbnails, the game is always going to be how do I stand out among what's, you know, what other people are seeing on YouTube? And if everybody's using AI gener to make their thumbnails, well, then nobody's really going to stand out anymore. So I, I just think that's the game with thumbnails is figure out how to, how to grab attention among the sea of other thumbnails. And for a little while it was using AI Generated images. And nowadays it's actually using real faces again because there's so many AI generated images all over thumbnails. So, yeah, just. Just a constant ebb and flow there.
- [57:45]
A
I think you just tapped into, I think, one of the most. The number one most important strategy when it comes to YouTube and content in a 2026 world. And that is, I think now you have to be continually changing, updating, adapting, and pivoting. Like, as soon as something becomes a best practice, it's like too late in content now.
- [58:07]
B
Right.
- [58:08]
A
And. And also it could be secular. It actually might become a best practice again three years from now. But lots of things are trends, counter trends, because when everyone starts to do it, you disappear into the same sea of. So you pivot back, real face, and that's different. It starts to stand out and then it could pivot back. So the new way forward on content, I'm curious if you agree on this is, is you have to stay more vigilant and more rigorous than ever before. The world wasn't moving as fast in 2010 or 2015 on YouTube. These tools hadn't even really emerged. And so maybe you found a best practice and you did the same playbook 1, 2, 3, 4, 5 years in a row. It feels like the cycle, of course there's fundamentals in success in YouTube and content, but feels like the cycle is like every quarter, like just kind of trying new things, trying new pivots. The fact that we do have the ability to split test 3 thumbnails gives us an opportunity to throw a test out there. And very well. The difference between breaking through or not is not just getting stuck in a rut, but continuing to pivot and adapt. How are you doing that on your YouTube content? You. You mentioned one example, but your channel continues to thrive. What are the things you're doing right now tactically on YouTube?
- [59:19]
B
I would say the biggest move that I made on YouTube was I actually brought on a producer several months ago. And for me, that has been the biggest game changer, because having a producer, him and I get on calls multiple times a week and we brainstorm ideas and we look at what other YouTube videos are coming out in this space and we look at what other YouTube videos are coming out in different spaces and go, maybe we can sort of take some thematic elements from that video, but, you know, put our own AI spin on it and like, I basically have a brainstorm partner to constantly try to stay up with what we should do next. And. And I totally agree. I think AI, like, I think YouTube is a constantly moving target and the rate at which the target moves is only speeding up because more it's becoming easier and easier for anybody to make YouTube videos. I mean like this whole discussion has been about like how you can make videos easier, how you can do faceless channels with, you know, voiceovers easier. There's tools out there with one prompt you can make a full 20 minute video now. So like anybody could get on YouTube and it just gets more and more important to figure out how to stand out. And for me that that change has been work with somebody that also knows YouTube really, really, really well from the inside out and just constantly be brainstorming and trying to figure out what the, you know, what the, what the next move is. I mean from like a, a niche standpoint, niche, whatever you want to say from that standpoint, it hasn't really changed. I still like talk about AI news, talk about AI tools, test the AI tools, make tutorials about them. The, the type of content is always the same, but how I edit them has changed up a lot. What our thumbnails look like has changed up a lot. How we title them changed a lot. So like all of these sort of elements are, are constantly evolving, but we also have to kind of keep it wrapped in the same niche because we don't want to lose our existing audience. And having having that producer that I'm just constantly in brainstorm sessions with has probably been that big game changer for me. But yeah, it's just constantly, constantly just paying attention to what's working. And I really don't actually pay much attention to what other AI YouTubers are doing. I, I kind of see some of their videos from time to time, but it's not really what I watch on YouTube myself. I, I tend to watch a lot more of the like mainstream YouTubers, right? Like I do watch Mr. Beast and Ryan Trahan and some of like moist critical and some of these guys, like, I tend to watch the more mainstream, well known YouTubers and kind of pay attention to what they're doing because they have more data and analytics than any of us with a lot less views. So they're, they probably have a, their, their finger on the pulse of the YouTube meta way better than we do. So that's kind of what I'm constantly paying attention to is like what are the bigger youtubers testing and trying and what worked for them and how can I apply that to what I'm doing if I stay in my like AI bubble and watch what they're doing most of them are kind of just like in the flow of making the same kind of video constantly. And for me, that's not, that's not really what I want to be paying attention to. It's not really what I want to follow along to, to see. See where I can evolve.
- [62:27]
A
That's a powerful insight. I'd love for you to adapt it to someone who's starting though, and they're listening to that and they're like, I love that but great. You brought on a producer. I'm trying to bring on myself because I'm trying to get into the YouTube game as a side hustle. I'm balancing it with my 9 to 5 or I'm a business owner and I just like, I'm trying to get leverage out of this YouTube thing, but I'm still trying to put out fires in my business and you know, just run the day to day. So the concept is rock solid, like to be brainstorming more. I think your insight of not study, not even worrying too much about your competitors, but worrying or studying more adjacent industries or just completely unrelated niches and creators who are really doing cool things and pulling those insights into your industry. If you're looking at what Moist critical is doing or Ray and Trahan, you pull that into real estate, you pull that into your tax tips channel, then that's, that's going to yield more results. But for the solo creator, as we land the plane on the episode with the. We do have AI tools we can think about our disadvantages can't bring on a team like you. You're much more mature in terms of your staff, your momentum, your revenue. But there's a lot of people that still want to get into YouTube right now. Yeah. What would be some of your thoughts and opportunities of what you would do if you had to start over with limited resources?
- [63:51]
B
Totally. Yeah. So one you, you touched on. You know, you've got tools like ChatGPT and it's a surprisingly good brainstorming partner. Personally, I think it sucks at creating titles and thumbnail ideas. Like for whatever reason that that creativity of coming up with with good titles that either goes way too far, like overly clickbaity or way too tame. Like that's not going to interest anybody. I, I haven't found that sweet spot. But ChatGPT is a really good brainstorming partner as far as like ideas for videos and like what's going to hook the viewer. A lot of times I'll go to Chat GPT and I'll say I'm making A video about the latest, you know, VO3 prompting tips. How do I put a hook on this? That in the first 30 seconds of the video, people go, oh, I'm locked in. I need to watch the rest of this video. And ChatGPT is a great brainstorming partner for that kind of stuff. It will help come up with good hooks. The other thing I'd say is there's a lot of other people that are in a similar boat that you can try to, like, rise together with. So I have a Discord community that I created about four years ago, and what I literally did, and this was, you know, before I had a ton of subscribers or a ton of views. But I went to a whole bunch of other channels that were talking about AI at the time. We were all small. This was before AI exploded in the mainstream. And I said, hey, I have an AI channel. Do you want to jump on Discord? I'm creating a little bit community. And I brought in about 30 other YouTubers that were all making AI related content on YouTube. And we all got into Discord and started, like, sharing stuff. And like, what? Like, a big piece of news would come out and somebody would go in there and be like, hey, did you guys catch this news? I made a video about it, and it's a 1 of 10. You guys should make videos about it. And we were all getting in there, like, sharing what was working now and, you know, prompting tips and what news was getting good views for our different channels and things like that. And it all started by me just going and, like, dming a whole bunch of creators that were in this space and saying, hey, do you want to jump on Discord and have, like, a little Mastermind? I know, like, if you ever watch any of the interviews with, like, Mr. Beast from the past, he claims that was, like, how he got started. And he still talks to that Mastermind to this day. I still do too. I still have that Discord group where we're still throwing ideas around together constantly. And so I think trying to create a little community of people that are like, you know, maybe where you're at to a little bit further than where you. Where you're at, like, where you want to be and getting that community together and sort of sharing ideas and brainstorming and masterminding on ideas and even figuring out how to collaborate, that's one of the cool things about this Discord that we created is we collaborate from time to time, all the time. Like, there's other big YouTubers, like Matthew Berman's another AI YouTuber who's pretty well established and I'll pop onto his channel from time to time. There'll be a video and I'll just kind of cameo and talk to him for a minute in one of his videos and he'll come on my video from time to time. And a lot of these other bigger youtuber AI youtubers now we all sort of rose together, but we're all still sort of sharing the love and trying to figure out how to collaborate and things like that. So just trying to build a little like mastermind community of like minded people that have the same goals as you. I think can be really, really helpful.
- [67:00]
A
If somebody wanted to start from scratch and follow your gear tools breakdown, if you can take it all the way back, assuming your resources are limited. I'm curious if you were to walk me through. Okay, I have to start from scratch. I'm going to set up a bedroom in my house, in a spare bedroom. I'm going to turn it into an office. What tools are you going to use? Are you going to create sustainable content? You know, they might want to start a video podcast. You cover news, but they're going to have to edit themselves or use AI to edit. They might use Riverside to record. I'm just curious, you know, there's multiple pathways to do this, but from your standpoint, if you're thinking now they've got a good job and they got some money and they want to invest in their creator business, they've maybe bought a course, they've got education, maybe they're already a part of our video ranking academy community, but they're still trying to figure out their thing. And so they're kind of like, yeah, do I kind of go webcam, set up a camera? If you thought about minimum viable dose, but the budget's not huge, they might be able to spend a thousand, couple grand, get, get a few tools. What would be your sort of like minimum viable product, minimum effective dose for the solo creator to upstart their mini YouTube company in a 20, 26 world?
- [68:17]
B
Yeah, well, I mean from a software side, most of it you can do for free. I still use mostly free tools. From a software side, I still record my videos with OBS to this day, which is a free platform that works on Mac and PC. So all of my videos are recorded with OBS. When I edit, I edit with DaVinci Resolve. Now I have upgraded to the studio plan, but for years I was editing with DaVinci Resolve's free version. Right. So when it comes to recording and editing the tools to use shouldn't be a bottleneck or a barrier to you because in my opinion, the best recording tool and the best editing tool are free to use. When it comes to equipment, I'd probably get whatever the highest end, you know, either Logitech cam is or like an Elgato Face cam. Probably just get, you know, in that 200 to 300 good 4K webcam range. And I would pro. Like I'd probably put the most money if I like my most of my budget would probably go to making sure I have good sound right. Like I think audio quality is probably one of the most important things. If like something gets a little blurry or pixelated on a video, people don't tend to mine too much as long as the whole video is not like that. But if your audio quality sucks and there's background noise and crackling or just low quality audio, people will tune out really, really quick. I think audio is insanely important. So I'd probably put the money into like I think my favorite sort of budget microphone is the Elgato Wave 3. I really, really like that microphone. I use that one for years and years and years before I finally upgraded to the Rodecaster duo and the SM7B here. But I would probably just get a USB mic like the wave 3 from Elgato, get maybe an Elgato face cam and Elgato Stream deck so that they all sort of work and seamlessly integrate together and then get a couple of really good key lights, like maybe a couple Elgato key lights so they all tie together. I do think the Elgato key lights are pricey for what they are. Like you're paying like a little extra because it's Elgato and it'll sync up to your stream deck. So like if you're on more of a budget, you can go and get like there's GVM makes decent lights, but they're like budget lights that work really, really well. And that'd probably be my setup record with obs edit with DaVinci. Probably get like a suite of Elgato tools, the Elgato Wave Mic, Elgato Stream Deck, Elgato Lights and an Elgato Facecam. And that would probably be my setup to get going. You're probably all in at 4 or 500 bucks with that setup.
- [70:48]
A
And if you wanted to cover news and react for the listener, how would you. If you wanted to react to videos or share your screen and create a higher quality experience of doing that type of content Would you just use obs? Is it easy enough to do it with that?
- [71:03]
B
Yeah, I have like the most ghetto obs setup right now. So when, when I record with obs, all my videos are 4k now. I met with my partner manager and she was like, yeah, definitely start putting videos out in 4K because more people are watching them on TVs and the TVs prioritize 4K now. So I do everything in 4K, but I actually set. My obs is like a double wide. So it's actually like 7,800 by 1920 tall, right? And then the left side is my screen recording and the right side is my camera recording. And both of them are recording in 4K. So I'm sending my editor like these 20 gigabyte files that are double wide 4K videos, but it has my screen capture and my camera all on the same thing so that they can sort of individually edit the two. But yeah, that's pretty much how I would, I would set it up. I've got two monitors, so like my, my obs is running on one and then whatever I'm looking at is running on the other monitor. I don't think that's necessary. It's nice to have once you can. But yeah, I'm just looking at like whatever I'm reacting to on the screen and talking about it and my obs, since it's recording sort of both monitors at the same time, I'm getting it all and my editor can sort of mix and match whether I'm fully on camera or, you know, down in the corner or just the full screen without me.
- [72:23]
A
And are you using a Mac or a PC?
- [72:25]
B
I use both. My, my main daily driver is a Mac, but I do have a PC because Nvidia's been a sponsor in the past. So they kind of set me up with a really solid PC because I do a lot of tutorials and not only news. Sometimes I do tutorials on my PC, sometimes I'm doing them on my Mac, depending on like what software I'm talking about. But I do sort of switch back and forth. But I'm probably 80, 20, 80% using my Mac, 20% using my PC.
- [72:52]
A
I mean two 4K captures of you as the personality, the talent delivering the information, as well as the ability to capture 4K of whatever you're looking at. Watching websites, tutorials. The opportunities are endless. Like what level of Mac can handle capturing and digesting all of that that you have?
- [73:15]
B
I would say most of the, like the M series Macs can Handle it. Now saying that I'm on a complete overkill Mac. Mine is an M3 Ultra with 512 gigabytes of RAM. I mean it's pretty much like ultra, ultra decked out. But I did that because when I want to, when I render videos, I want, I want to be able to render 4K videos really quickly. When I'm scrubbing the timeline on 4K videos, there's no lag. So I just went, I'm going to future proof the heck out of this thing. But I do think most of the M series Macs can do it because I was, I was doing it on a MacBook before this. Just that was, it was an M3 that I. My MacBook. And that M3 MacBook was able to record just as well. I was able to do the sort of double wide 4k on a MacBook. So I think like most of the M series chips can do it on my PC. I mean, again, super overkill. I'm running an Nvidia 5090 GPU on it with just insane amounts of RAM. But saying that before I had that, I was recording on a Nvidia 3070, like just a pretty consumer grade like gaming GPU and it, it recorded just fine on that as well. The only time I would run into issues is if I ever tried to record while playing like a graphic intensive game or something. Right. Like you're not going to go and play Elden Ring and record it in 4K and have it not be choppy. But for just screen capture, if you're just sharing your browser and what you're seeing on like Instagram or whatever it is, works fine.
- [74:47]
A
So I have one final question for you in just a second, but that's been an incredible breakdown and I do want you to also shout out your stuff in a second. But to recap this, you know, number one, I think Think Media podcast listeners know that it's, it's not about your research resources, it's about your resourcefulness. And you know, we're living in an opportunity. I love that you started. Like there's just so many free tools like go figure it out. Like go learn DaVinci on a basic level, use OBS, get a $300 webcam. There's definitely a budget level. You could do this. If you're new here on the podcast and you've never learned about our main channel, it's called Think Media and we have multiple creators that are talking about all of the different levels of budget tools for creating content so we'll link that up in the show notes because like the tools is not an issue. And then for some of us listening though, I want to challenge individuals. You know, this isn't holding anybody back, but I think some people listening to this should really consider making an investment in their media. In today's world, like if you already have some momentum and if you have some money or if you're a business owner, it actually kind of surprises me how gun shy people are. I feel the exact same way. I'll invest in speed because time is more valuable than money. So sometimes if you have outdated gear, like just if you have access to fiber at wherever you create content, your house and you haven't upgraded to fiber, like what are you doing? You know, I actually was on cable forever and I realized I had access to fiber. I just got it, I mean speed to get, get those files uploaded online if you're gonna send it to somebody, speed if you can process an export, like speed is just worth the money because that, if you, if you spread that over time too, you future proofed like a few months, a few years and and then also to be able to put out a level of quality like it's not that far out of reach. The Mac you're talking about, I could see it on B and H right now on sale M3 Ultra. I don't know how big your hard drive is, but you know, it's like a five, six grand. And so for geez, for 10 grand you could make a TV studio in today's world. Like how clear and crispy and 4k in the level of production you could apply to it. Like that business opportunity just did not exist a decade ago, definitely not two decades ago. And I've noticed that some of the creators that are getting the best traction are, are just taking, they're turning pro. They're, they're like leaving an amateur mindset aside, it's the fact you hired a producer. If you can't hire a producer, well use chat, GPT. But you got to think at a higher level. More strategy, more rigor, more intentionality, invest in it. Like how does a pro think? Better gear, more practice, more better nutrition, better sleep. They hire coaches. They like a pro acts different than an amateur. So I think that mindset of turning pro, do you have the pro gear? And it doesn't have to be a limitation because again you're like se, I'm broke. You then hustle with your smartphone until you get a little cash and then reinvest it because the opportunity on YouTube, social media, AI, it's bigger than ever before, but it's also more crowded than ever before. And I think that cream always rises to the top no matter how many cups of coffee you pour. It's like if you just have. If you're in that top 10% of content, which I want to say I believe is accessible to everybody, because most people create lazy content and most people are inconsistent, and most people are in that bottom 90% of noise. A little bit of extra effort can help you rise to where the views are. And I hope that some of those thoughts, as well as your breakdown of gear will be helpful for people before we end. What? How could people follow you? You've got some cool stuff. Tell us exactly what it is. We'll make sure there's a full list in the show notes so people can stay up with all this kind of news that you're doing as well as other projects you're working on.
- [78:30]
B
Yeah, so my main channel is just Matt Wolf is just my name. If you go to YouTube and search out Matt Wolf, you'll find me. That's where I share AI news, AI tutorials and various, you know, breakdowns of what's going on in the world of AI. I've got my Future Tools website, which is a website that curates all of the latest AI tools that come out so you can sort them and filter them and find the exact tool you're looking for. That's also where my Future Tools newsletter is. And then I've got a podcast called the Next Wave, where I do interviews with, you know, CEOs and authors and various influencers specifically focused on AI. So we talk about, like, their AI tools and demo their AI tools and interview them about where AI is all headed and stuff like that. I would say if you head over to Matt Wolf on YouTube, though, that's the best place you'll, you know, you'll find links off to all the other resources from there.
- [79:21]
A
Well, Think Media podcast. I want to highly encourage you to follow Matt everywhere because he's an incredible human and he also has. He just has great information and you want to have certain people in your life so that, that key thing, key details, key information, key trends are coming your way so you can stay on the cutting edge. So definitely take the time to check out the show notes. Matt, one final question and probably the most important question. I'm going to give you a chance to think about it and mention a few thoughts. But looking ahead in the next 12 months, what are the trends that are redefining kind of the creator economy and the AI wave and disruption that's happening right now. I think a lot of people have mentioned, you know, AI is a creative partner, AI isn't replacing humans, but it's AI plus humans. We are maybe multi Persona creators where you do have individuals you mentioned earlier that are duplicating themselves and in a quality way. Production workflows are going faster. Is a big trend, perhaps dubbing, because in an effective way we could spawn a Spanish channel or other things or you know, trends you mentioned earlier. So I'm just giving you a chance to dig deep to anything else you could think about. But the fact that maybe we're going to see a resurgence of vlogging and really authentic content because of the fatigue AI is creating. Maybe best if you were in a high level mastermind as an expert that got to coach us in just a few minutes and you just thought they're like, Matt, like how do we prepare? What are we not seeing? What's coming in the next 12 months that you're worried about, excited about, or that you actually think is going to destroy 20% of creators? What do you see for the future of the creator economy?
- [81:10]
B
Yeah, so I do have somewhat of a dystopian take on where all this might be headed. So I, I look at a company like, like Meta, right? And I think Meta is probably going to be the one that we see do this first. But I think right now a lot of these big companies see creators as like the middlemen, right? If you look at something like, like an Uber, right? Uber. They've. You've got the customers, you've got Uber, the company, and then you've got the car drivers, and the car drivers are the middlemen. And Uber's trying to cut out the drivers by putting, you know, Waymos and autonomous vehicles in their fleet and they're trying to cut them out. Well, if we're getting to a point where VO3 can generate really decent videos with voiceover, with audio, with everything, and all it takes is a really, really good prompt to get that video out of it, well, AI is going to be able to write those good prompts as well. So I think it's only a matter of time before companies like Meta start just generating their own AI content and optimizing for whatever that dopamine hit is, right? If you're scrolling your Instagram stories and a video gives you an Instagram or like a dopamine hit as you're scrolling and you watch more of that kind of content. That kind of content's going to get more reach. Well, I think it's going to get to a point where these companies can figure out, like, AI can generate dopamine, hit content, put it in front of people. What do we need creators for to do that? And I think that's what Meta wants to create. I really hope YouTube doesn't go down that route, but I think Meta is trying to go down that route. I think they want to go down the route of we can lock people into a platform where they're just scrolling AI generated content and nobody cares as long as it's got that dopamine hit. I think that's what's coming. I think that's what Instagram wants to create. I think maybe even TikTok. And I really, really, really hope YouTube avoids that, saying that I do think some of the stuff that's going to get really, really good. I do think we're going to see dubbing get a lot better. We're going to see it get to a point where if we dub something over in Spanish, it's actually going to look like our lips are saying it in Spanish. That technology already exists. It's just not baked into YouTube directly yet. So I think that's going to come eventually where you flip it to a different dub and it looks like I'm speaking to you in Spanish or it looks like I'm speaking to you in Japanese. That I don't think is too far off. The technology's there. YouTube just needs to implement it. I think we're going to see a lot more of this, like agentic workflows where you can feed it an hour and a half long video and tell it to cut it down to 20 minutes and it will figure out where to cut, where to take the mistakes out. You know, when to, to. To show your screen versus when to show your camera, when to overlay B roll. I really think, like, we're looking at that kind of technology next year. Descript is already trying to build that. So, like we're already seeing it there. There's a company called Memories AI, which I just came across recently, where you can load hours and hours and hours of video into it and then tell it, go and make me a 10 minute montage of. Let's say I uploaded a whole bunch of interviews with Sam Altman. Make me a montage of Sam Altman saying AGI and it'll go and make like a three minute montage from all of these clips of Sam Altman. It'll split them all together of him just going AGI, AGI, AGI. Right. So like memories, AI is already creating that technology. So I think we're going to see that like, like agentic AI editing really sort of come to fruition to where it feels pretty usable in 2026. I think that's kind of going to be a big thing. So between the agent editor and the, the dubbing with actual good lip syncing, I think those are the two things that are on the horizon for creators that we should all be pretty excited about. Where I believe Meta and TikTok want to take their platforms and cut out the middleman. That's the part that I'm a little bit more pessimistic dystopian about.
- [85:02]
A
Well, Matt, you added so much value today and super grateful you took the time. I know you're doing so much valuable content for your own community and so you adding value to ours is not something we take for granted and super grateful for just the excellence and authenticity you're bringing to the YouTube space, the creator economy and the AI space. So thanks for coming on the podcast.
- [85:22]
B
Yeah, thanks so much for having me. It's always, always fun to get to chat with you and you know, love nerding out about this stuff, so it's always a good time.
- [85:29]
A
Well, I don't know about you, but my mind is racing with all of the information that Matt shared and thinking about how I'm going to apply that to my own YouTube channel and online business. And by the way, if you're thinking about starting or growing a YouTube channel this year and you want an updated playbook of how to get ahead, I've recently updated my on demand YouTube masterclass. It's entirely free, it's a one hour deep dive training and you can actually access it and watch it immediately@thinkmasterclass.com I'll make sure I've linked that up in the show notes as well as all the other show notes, Matt, stuff that we talked about in this episode and if you got value today and you're still here, thanks for being here. And if you could hit the like button or rate or review wherever you listen to the podcast on video or audio, that would mean the world. My name is Sean Cannell, your guide to building a profitable YouTube channel. And I cannot wait to connect with you in a future episode of the Think Media podcast.