EP 445: Content's AI Revolution: Videos, Streams and Stars (2025 AI Predictions - Vol. 3)

This is the Everyday AI show, the everyday podcast where we simplify AI and bring its power to your fingertips. Listen daily for practical advice to boost your career, business and everyday life.

Everything you think you know about content is going to change in 2025 because of generative AI. How we think about content, how we create content, how we distribute it, how we consume it, everything around video, creativity, content, the media, it's all going to change very quickly. And that's what we're going to be tackling on this special edition of Everyday AI, where we are diving back into our 2025 AI predictions, volume three of five. Today, we're going to be looking at contents, AI revolution, videos, stream, dreams and stars. I'm excited for this one. Hope you are, too. If you're new here, what's going on, y'? All, My name is Jordan Wilson. This is Everyday AI. If if you're trying to keep up with what's happening in the world of Generative AI, you're in the right place. We do this every single day, Monday through Friday, with a daily live stream, podcast and free daily newsletter helping us all, hopefully not just keep up with AI, but how we can be the smartest person in our company when it comes to artificial intelligence. So, please, if you haven't already, go to your everyday AI.com Sign up for our free daily newsletter. And in that newsletter, we're going to be recapping not just today's show, but we're going to have links in there to volumes one and volumes two. All right, so maybe this is the first time you've heard this show and you might be wondering, who's this weird guy? Why is he making predictions? Well, I've literally, in 2024, I interviewed hundreds with an S. Hundreds of people for this show. Some of the smartest people in the world, from big companies like IBM, Google, Microsoft, down to small startups, the people who are building the technology that we're all using when it comes to content creation, media marketing, etc, right? And these things keep me up at night, right? Sometimes in a good way, sometimes in a bad way. But I've realized I have so much great information. Not just what, you know, you all here on this podcast or the live stream, but I've had hundreds of conversations not on those live streams, right? Whether we're helping enterprise clients learn or train their employees on Generative AI or just when I'm talking to people offline, right? Those same people from the same big companies. I have great relationships with all of them. So I said, hey, I need to take all of this knowledge, all these conversations and give them all to you. So as my thank you for tuning in to this show, we're going to go over our 2025 AI predictions. Volume three of five content AI revolution. So if you missed volumes one and two, we're gonna have those. You can go back, listen to them. It's going to be the last two episodes, but I'm gonna very quickly recap them. All right, not going to go into detail, I'm just going to tell you what they all were. So here they are. Number 25, we're going to go reverse. Ready quick. 25 agent orchestrators will be a growing position. 24 public companies will post jobs for AI agents. 23 company reasoning data collection. 22 high end professional services will go through a pricing crisis. And 21 UBI becomes a household conversation. That was volume one. Volume two, here's what it was. 20 open source surges. Open LLMs will temporarily overtake proprietary models. 19. Chinese AI will domina eight and cause confusion. 18. Perplexity will pivot, get acquired or get squashed. 17. API prices gonna drop like they're hot. And 16 embodied AI will be an exploding sector. Which brings us to Today, Volume 3 Content AI Revolution. Videos, streams and stars. I'm gonna give it to you right now and then we're gonna dive in. You don't gotta wait anymore. Ready? Number 15. AI video tools will one shot five plus minute HD videos and something I'm calling APM. The future of traditional Internet comes into question. Social media makes deep fake AI problems much worse. 12 first big copyright case is settled and 11 AI influencers start killing off UGC content. All right, live stream audience, as I did with the last ones. Go ahead, put the number of the one you think is the most likely. So maybe you think 13. So say 13 most. And then which one do you think is the least likely? So you know, whatever that is, put the number and then say least. Yeah, so if you listen to the podcast, we do this also as a live stream. Same exact thing. So if you want it first, join us on LinkedIn Twitter, 7:30am Central Standard Time. All right, let's dive into number 15. AI video tools will one shot five minute HD videos. And the advent of APM. That's what I'm calling a term. I'm calling advanced personalized media. All right. I kind of coined a term last year during my predictions show called second Computer AI, which actually became a thing. So let's just go straight into it. AI video is amazingly Good, right. If you haven't been following it recently, it's no longer, you know, janky Will Smith eating spaghetti. You can't tell the difference, right? Let me put this out there. Most of you all don't know this. All right? Way back in one of my former lives, both when I was a journalist for seven years and I was the executive director of a nonprofit for 10 years, I did a lot of photo in video, kind of like a pastime, hobby slash. Well, profession I've shot probably. I did the math once, I think close to a million photos in my lifetime. I have four DSLO cameras sitting in the closet. Someone here somewhere here in the little home studio. You can't tell anymore. Can't, right? AI images in mid journey. You can't, you know, at least those that actually know what, what, what we're doing. Right. If you're talking about something like Dolly or something like that, it looks cartoonish, right? But AI photos look very real. And everyone's turning those AI photos into AI videos, obviously. So it's, it's hard to tell what's real and what's fake. And with, especially with sora. SORA is very good, but it struggles with real world physics. What people don't understand the SORA version that we all got in December, well, that was Sora Turbo, which is not the full SORA model. And then obviously right after that, Google came out with vo, which I don't have words. It's that good. Shockingly good. It understands real world physics. But here's where I think it's going next. One prompt, five plus minute videos. Yeah, I gotta come with bold predictions. I'm not going to come here with boring, boring takes. I think that's where we're going. I think we will see truly personalized AI entertainment in 2026. But I think we're going to get our first glimpses toward the end of this year, 2025. Y' all remember those like Pixar shorts, right? Those are really good. Those are really cutesy. Think of those. But for exactly what you like, right? How it is now, you got to open up Netflix, open up Amazon, and there's, you know, thousands of things, but I'm like, who is, who are these for? Right? Yeah, there's these algorithms and they try to recommend to you whatever is best, but it's going to be different. I think it's. You're going to either, I think Google, as an example, you could just opt in and say, hey, take all my browsing data and make me a Video on that, right? That might be a good or bad thing, depending on what you do on the Internet. I would love that. I would love. It would be a bunch of dorky AI stuff. Stuff and you know, Chicago sports in North Carolina, men's basketball and random gadgets like, yeah, show me some videos on that. I think that's where we're headed. And I, speaking of that side note, I also see one of the major studios, right? Whether that's, you know, the movie studios, maybe it's Netflix, maybe it's Hulu, etc, I see one of them either acquiring or Aqua hiring one of these AI video makers, right? Maybe it's not Runway. I think Runway is probably, maybe too big or too established to be acquired right now. But maybe it's, I don't know, maybe it's one of these, these Pika Labs, you know, Luma, I don't know. But I think one of these AI video makers is probably going to get acquired also. What's going to make that happen? Well, literally weeks ago we started to see the first glimpses of multi scene AI video shoots. So that's where you get this consistency, right? So a lot of people, you know, text to video, it's getting better. It's not fantastic, right? I think you get good shots when you upload photos. So I do think by the end of the year there will still be text to video that's really good, but it'll generate up to five minutes. But you also have where you can now have the exact same scene even from a photo, and you get it from multiple angles. And I think that's ultimately what's going to help, you know, this kind of reality, this advanced personalized media, or apm, what I'm calling it, it's going to come to fruition. Mark it down. All right, 14. This one is good. Are you still running in circles trying to figure out how to actually grow your business with AI? Maybe your company has been tinkering with large language models for a year or more, but can't really get traction to find ROI on Genai. Hey, this is Jordan Wilson, host of this very podcast. Companies like Adobe, Microsoft and Nvidia have partnered with us because they trust our expertise in educating the masses around generative AI to get ahead. And some of the most innovative companies in the country hire us to help with their AI strategy and to train hundreds of their employees on how to use Gen AI. So whether you're looking for ChatGPT training for thousands or just need help building your front end AI strategy, you can partner with us too. Just like some of the biggest companies in the world do. Go to your everydayai.com partner to get in contact with our team, or you can just click on the partner section of our website. We'll help you stop running in those AI circles and help get your team ahead and build a straight path to ROI on Genai. Like, I'm good for it. It's. This one's a little ambiguous. We can't have a, you know, a yes or no on the scores on the score sheet. And I'll probably do like I did last year, y'. All, like in July, I'll do a little midterm report. So this one, yeah, is. Is there's gray area, but I think the future of the traditional Internet comes into question. All right, what do I mean by that? Well, the Internet's garbage, all right, For a lot of reasons. What a lot of people don't know, you know, chat. GPT wasn't the first large language model, right? The GPT technology was released in 2020, the precursor, right? So we had, I. I believe a GPT2, GP2.3 was released commercially. So a lot of the early Internet, right? A lot of the SEOs of the world, right, those that were using Spintax long before there was a G, they flooded the Internet with it, right? A lot of. A lot of garbage. But now it's like, who's not using AI to create content for the web, right? Because traditionally, creating content for the web, it was either expensive or bad, right? Now it's cheap and good enough with large language models, right? So even what we have online right now is already so much AI. So a recent studies show that 90% of new content posted online by next year will be AI generated. Let me repeat that. 90% of new content, new written content posted online will be AI generated by 2026. All right? This does a couple of things. Number one, it cuts down on the amount of original content produced by humans, right? We always talk about, oh, there's not enough data for these large language models to scrape and gobble up. That's true. I'm calling it AI regurgitation, right? It's already been happening for a couple of years behind the scenes, people just don't understand it, right? Even this training data, right? That, and this is why, you know, the smart people at the big tech companies are saying, hey, we might be running out of training data. Well, it's because it's already been regurgitated, right? So much of what's on the Internet is Just regurgitated content. You know, I would say the, the last kind of AI free scraping of the Internet was probably in 2021. You know, maybe before that. I could be wrong, but you know, I, I have a little background in SEO, so I always follow what's going on in, in, in the, in the areas of SEO content, right? Flooded. It was flooded early 2020, right? So it just creates this vicious cycle of just AI recycling. I just call it regurgitated because oftentimes you get the worse, you get the worst of it. Because the people doing this in mass and in bulk, they don't care. They don't, right? They're just like, all right, I'm gonna get a hundred thousand words on the website today, 100,000 words on the website tomorrow, right? This systematic AI SEO, not good. And this makes, well, this causes a couple issues, right? Number one, new content, it's bad for us humans because we're reading less human produced content. It is just regurgitated. So, you know, think, you know, a ship off 1%, right? 1% off the course of its target is eventually going to be completely in a different destination. I think that's the same thing that's going to happen, right? Compounding, number one, AI generated content isn't good because people don't know what they're doing. I obviously do. I'm a writer, you know, I used to win writing awards. I, I use chat GBT all day. I can, I can create fantastic AI content, but very few people can. That's not me being like, oh, I'm so good. It's just, that's just what I've been doing for a long time. All right, so we get a bunch of mediocre content. It keeps getting, getting gobbled up by the large language models. People keep spitting it out. So that's, we're getting less original, high quality data. That's a problem. Number two, we have answers engines now, right? So people aren't even going to websites anymore, right? So that's problematic as well. So what incentive do big publishers who want to be putting out original stories, new research, new information, they're becoming decent advice to actually do that, right? The way the Internet works online, right? You put content on your website, you either monetize from people who find your content on Google, they sign up for, you know, a subscription of yours, they give you their email, maybe you pixel them so you can retarget them on other things and you sell them things, right? That's not happening as much anymore because you have answers engines like Perplexity Chat, GPT search, large language models in general. But now you also have this thing that's, I think, a behemoth. Google Deep Research. I've gotten it to go to 1300 web pages to scrape, essentially scrape all this information. Google Research does a little different. It does it via its cache of the Internet. You know, other companies haven't, haven't really told a lot how theirs works, but that's 1300 sites I didn't go to. That's 1300 publishers that didn't get to Pixel Me. That's 1300 companies that maybe I would have bought something or given them my email. And now they get nothing. So the very hand that feeds large language models, they've cut it off. It's problematic. Mark it down, 13. Social media makes deepfake AI problems worse. Oh, gosh. Meta don't know what Meta's doing here. All right, so Meta had this disastrous. And that's putting it nicely. Meta had this disastrous experiment with AI profiles, right? They thought, you know, we're gonna, we're gonna do this thing, we're gonna have these, you know, AI profiles engaging with humans. They're gonna look like real profiles. Yeah, there's a little badge that says they're AI. But Meta wanted to make their platform stickier. Right? So literally, Meta, the company flooded their own social media channels with AI bots, y'. All. I love AI. Terrible idea. That's somewhere AI does not need to be. All right, so you had that. That's problematic. That's. They shut it down, luckily, because everyone's like, yo, we don't need this. You know, that's pretty bad, right? Between social media and AI, those are, you know, people who love those things. Love those things. So, you know, you are a little tone deaf if literally the whole world's like, nah, we don't need that. Don't do that. That's bad, right? We don't want that. Right. Yeah, you're trying to make the, the platform stickier and more engaging. Do it, do it in a different way. Another thing that's going to lead to deep fakes. Well, speaking of meta, Mark Zuckerberg. Mark Zuckerberg said he's getting rid of fact checkers and instead replacing it with a community note type system. So you could argue on the pros and the cons, I think ultimately that's a net loss for AI safety. All right. That just means, I think, you know, and I get you can make the argument right, because it's like, okay, well, you know, these fact checkers are humans. The Community note system, it gives it the power to the users and, you know, algorithms. Sure. But I don't think it's a good move. Next, Elon Musk's XAI is reportedly working on an unhinged mode in Grok. Yeah, so these. Yeah. Which is essentially just like Grok with no guardrails. So I'm sure that's going to, you know, start with tax, but also integrate with Twitter slash X photos as well with their AI photo generator. It's not good. Right. So essentially all the social media companies, they are literally guardrails. Now we're good, right? Let's flood it with AI. Let's get rid of guardrails. Recipe for disaster, right? All these, you know, 79 year old grandmas, no offense, if your grandma is 79. Right. They already can't tell the difference. Hey, Jimmy, is this your classmate? It looks like he robbed a grocery store, right? Yeah. We, we don't need fewer guardrails and more AI on social media, right? We don't. We don't. Those are supposed to be places where humans hang out and it's safe, right? Next, I think the combination is finally here of quality, Right. I think we're finally in 2025, gonna have real enough AI photos that get combined with real AI video. Right. You make the. But no even real photos, right. Because you can upload a real photo. And this is bad. This is where the deep fake comes, Right. I wouldn't want to be someone in high school because kids are going to get bullied this way. They're going to take photos of someone, they're going to create videos of things that didn't happen based on photos. Right? And that's very easy because every single, you know, teenager, college kid, whatever, they put their entire life on the Internet. It's very easy then to grab that, put it into an AI video generator, have it doing something that that person probably didn't want it to. And now you actually have lip sync AI. That's good. Recipe for deep fake disaster. All right. Also, yeah, you have to think. You have to think. So open AI and other researchers came out with a paper a couple of months ago to determine. They came up with a system to essentially say, hey, we need to be able to tell the difference between humans and AI, right. And I think something like that is needed, right? This is big researchers, right? It was OpenAI. It was Microsoft, it was, I believe, Oxford, some other big universities. And they said we need some sort of system, you know, And I know that's where people get like weirded out, right? Like, oh, I don't know, is it, do you get a number? Is it like a Social Security card? But it's just something that you put on an, you know, on, on Facebook or Twitter or YouTube to be like, yeah, I'm a real human. Because here's the thing. As we get closer to AGI, AGI will mimic humans, right? And as we look forward past artificial general intelligence to artificial super intelligence, AI systems are going to mimic humans. And we do need some sort of way to distinguish who is a real human, who is not. That's not me. I'm not advocating for. All right, we all need some, you know, number or something to say we're human. But the AI deepfake problems are going to get really bad in 2025, and social media is opening the door for that to happen. All right, number 12, first big copyright case is decided. All right. I do not know if it's going to be OpenAI versus the New York Times. That is the marquee copyright case. Right. And you might be wondering like, oh, well, for what? Well, let's call a spade a spade. For the most part, large language models scrape up the entirety of the Internet. Copyrighted works. All of them do. Right. That's why a lot of the people in big tech and big AI are essentially challenging, you know, decades old copyright law. And they're like, ah, this doesn't really apply anymore. So there's that whole argument. I'll say, I'll say argument, Right. I'm not going to get into that right now because, you know, yeah, it's essentially large language models. They don't one to one reproduce, but they take, you know, essentially, let's say someone writes about a product review, right? And maybe that's for a newspaper and that person got paid and you know, that newspaper copyrights that. Right. Well, the large language model is going to gobble that up, but also they're going to gobble up the other 50 people that did that exact same thing. And all 50 of those are copyrighted. But you know, the big tech companies say, well, when we just blend all 50 together and we kind of, you know, mix them up in a bowl and spit something out. Random. We're not reproducing copyrighted work. Right. But big news organizations, regardless, are suing them. And one of these cases is going to get decided, it might not be that one. That is the biggest one. And once a domino falls, a big domino. All right, I know, anthropic, I believe, did Settle a lawsuit with a music publisher over reproducing some lyrics. That's small potatoes. Once one of the big fries fries, things are gonna get spicy. All right. I don't know how much longer the Open AI New York Times is, is going to go on. I believe the New York Times is suing open AI and Microsoft for a lot of money, unspecified amounts, for allegedly using millions with an S of their pieces of work. Right? Regardless of what happens, right? Whether it's the, the AI tech company that wins or the publisher, you know, magazine company, newspaper company. Regardless of who wins, I think chaos, chaos ensues. The first big one, right, because then there's precedent and then everyone else is going to follow suit. But we're all waiting for that one big domino to fall or you know, maybe they just continue to get settled and you know, everyone's hush hush about it. All right, our last one. I'm trying to get all these done in 25 minutes by the way. We'll see if I actually do it. Last one, number 11, AI quote unquote influencers. So influencers that are actually AI are going to start killing off human UGC content. And you might be wondering like, why hasn't this already happened? Why didn't it happen in 2024? Well, the tech wasn't quite there. Now it is, right? There's literally all these open source models where you, you know, literally can get a picture from mid journey and you upload it and you know, they, they program all the, you know, tick tock dances and all these things. You can get these quote unquote AI influencers to sit there and you know, speak about your product. It's, it's nuts. And they're actually pretty realistic. When they first came out or when they first got popular about nine months ago, they were absolute garbage the past nine weeks. Scary good. All right, here's the thing. I think early on they didn't catch on because there was this over beautification of the kind of like the, the AI influencers, right? And it's like, no, that doesn't look real, right? But now they've gone this almost UGC style, right? User generated content. So why UGC was so big on social media is because, you know, sometimes it wasn't the best camera, sometimes it was a little shaky, right? Versus you know, a big commercial with a celebrity where they have a million dollars of video equipment, right? For the most part UGC and influencer culture took off because it's someone, you know, with a smartphone and they're Just talking into it. And that's all right. So I think now the kind of quote unquote AI influencer kind of sector is starting to explode because it's now just looks like real people, right? Not these, you know, hyper unrealistic versions of perfection that I think, you know, everyone was going for to begin with. It just looks like random people, right? In imperfections too, right? They don't have these perfect sounding voices. They're squeaky and they have weird cadences. Right? But I think that love them or hate them, they're here to stay. Right? And I don't think in some instances that's a bad thing, right? Because every single. Every single kid out there now, it's like, oh, what do you want to be? I want to be a YouTuber. I want to be an influencer. I want to do this. But it's like, okay, well, what? Right? Like, that's fine. All right. Are you doing that for science papers? Are you doing that for, I don't know, certain sports? No, they just want to, you know, go. Go talk. Right? So maybe it's better that, you know, hand some of that off to AI right now so the next generation can, I don't know, want to learn a certain skill, certain trade, go practice that. So I think they're here to say. All right, look at that. We wrapped it up in about 25 minutes. That was volume three of our 2025 AI predictions. Content, AI revolution, videos, streams and stars. I hope this one was helpful. Don't forget, we are now at the midway point. We already did volume one and two and I read those off to you. And we just wrapped up number three as well. All right, so make sure to join us tomorrow. We're going to be going over your. Our next set of AI predictions for the year. Your complete AI toolkit, tools, rules and global changes. It's gonna be a good one, y'. All. All right, if this was helpful, please do me a favor. Come back, Keep listening. We do this every day on our website. You should go check it out. We have everything sorted by category, like 450 podcasts. So whether you care about marketing, HR, legal, governance, whatever, go click on that category and learn from the world's experts on our website for free. Also, go to your everydayai.com. sign up for the free daily newsletter. We're going to be recapping this show and a whole lot more. Thank you for tuning in. Join us tomorrow and every day for more everyday AI. Thanks, y'. All.

And that's a wrap for today's. Edition of Everyday. AI. Thanks for joining us. If you enjoyed this episode, please subscribe and leave us a rating. It helps keep us going for a little more AI magic. Visit your everydayai.com and sign up to our daily newsletter so you don't get left behind. Go break some barriers and we'll see you next time.