OpenAI Should Pay Studio Ghibli $20M/Year — And AI SEO Should Die

0:02I'm Rob Kelly and this is Media and the Machine, a show about the biggest technology shift of our lifetime and how to profit from it. Each week I talk with the founders and CEOs closest to AI and content, the ones figuring this out in real time. I'm also building an AI content business myself and share lessons of what I learned along the way. You know, life's funny. I began my career lucky enough to interview leaders like Steve Jobs and Bill Gates. Then I went on to be a three-time founder and CEO, driving$100 million plus in revenue, and some failures too. And now I'm back at the table interviewing this new world's current and future leaders.

0:36This isn't only a business story, it's a human one. So every episode ends with me asking my guest what AI means for our jobs, our families, and the next generation. We'll figure this out together from the inside. Welcome to Media in the Machine.

1:07My guest today is Eric Burgess, CEO of CredTent, a public benefit corporation devoted to getting creators paid by AI. Eric's also a professional game designer. I loved Eric's critical thinking on things we covered, such as why OpenAI should pay Studio Ghibli$20 million a year for the style he says they ripped off. Why the new AI SEO should, quote, die in the crib. What the favorite games of Sam Altman, Demis Hassabis, Dario Amadei, Mark Zuckerberg, and Elon Musk reveal about how they think. What an AI bubble could mean for M &A, including predictions on who buys open AI, Anthropic, and Perplexity.

1:44Whether the pitchforks are coming for big AI. How Eric runs CredTent with nine humans and 35 AI agents. And his prediction on when we reach the age of abundance. Please enjoy my conversation with Eric Burgess. You've told me that tech has a history of moving fast and stealing things. Can you share more on that?

2:06Eric Burgess: Technology has been basically building on top of the backs of content creators for a very long time. The fact that big AI companies decided to hoover up all of the content in human history to build large language models is not surprising. Bottom line is by taking everyone's content without any kind of consent or compensation, what has happened is now we've created these plagiarism machines. And the large language model is basically using the patterns that are in the information, all the content that they've trained on to predict and create something that is market replacement for the work of all the people whose content that they have taken in.

2:41Can you give an example, maybe Studio Ghibli, I think as they say in Japan, Studio Ghibli, I think as some Americans say. You did an estimate of the revenue that maybe they should have generated. Can you just walk through how they didn't and what happened with that whole story?

2:58Eric Burgess: Yeah. So at one point in time, OpenAI, in what I would call probably a marketing move, made it possible for people to simply say, you can turn any picture you'd like into something that looks like it was made by Studio Ghibli. And this flooded the internet. We probably all remember this, but the Studio Ghibli one got pretty bad in the fact that not only was it something that is deeply offensive to the head of Studio Ghibli, Hayao Miyazaki, who has called AI an attack on life itself. But we actually got to the stage where there was a post that came out on one of the social channels associated with the White House that showed an ICE agent with a woman crying in handcuffs turned into Studio Ghibli style.

3:42Eric Burgess: And it's like, come on, this is really, really problematic, I think, for creators who are going to see their work and their style, which is not protected by copyright currently, of course, but their style taken and turned into political messages they don't agree with. That hurts the brand of Studio Ghibli because people are out there saying, well, maybe this is something that was created by them. It's confusing for people, especially in the media environment we're in these days. And so what we had done at the time was to do evaluation to say, you know, how much money would they be able to actually make if they were to actually license their content for the purposes of this kind of fabrication?

4:20Eric Burgess: We came out with a value that was something between$17 and$20 million a year that they could have been, they could have charged ChatGPT or OpenAI, I should say. They could have charged Google. They could have charged every one of these companies. So is it fair to say you believe OpenAI got more than well in excess of the$17 to$20 million in annual value because of them using the Studio Ghibli style? I would say so. I mean, it was quite an advertising move, to be honest, to be able to get people's attention. But they got this without actually paying the people whose work was actually being used for that purpose.

4:54And so good segue into what your company CredTent does, which is at least two solutions you could offer a Studio Ghibli in this case. You could offer them the ability to opt out completely from AI and just not have AI come by and consume their content. And the other one would be to make money, right? and you share 85 % of any money made if you were to do a deal with Studio Ghibli. And you did reach out to them at the time, right?

5:23Eric Burgess: We actually did. And we had one exchange with them. It didn't go too much further, but that's why we exist. Cred10 exists to be a neutral third-party utility focused on actually orchestrating these kinds of relationships between the creative industries and the AI world. And we have a universal opt-out list that people can register their work on at no cost to them. And then on the other side of that, if you would like to license your content, if you'd like to make some additional money, you can do so through us. We have CredTent customers that are making a fair amount of money actually from this, and it's part of their bottom line, especially when they're using really credible, high quality content.

6:00What's a fair amount of money, by the way?

6:02Eric Burgess: I can tell you that there are organizations that are not, you know, they're not News Corp, they're not Vox Media or NBC that are making in the millions of dollars for their content because they have made it available to Frontier Models and said, not only are we allowing you to train on our existing content, but we're also offering you a pipeline of new human-created content for training purposes in the future. Now, you told me in one of our conversations that only the rich content companies are making money. Can you share more on that? Well, the open deals that you see happened because what happened, Big AI came into the space.

6:39Eric Burgess: They looked at the size of the lawyers that they might have to face if they were not licensing from some of these organizations, and they made deals. So, you know, you have companies like News Corp and The Washington Post, the AP, Axel Springer that have made deals with many of the frontier models specifically to allow for the training of their models. Meanwhile, everybody else, it's quite a different rule for, you know, mid-tier media companies and smaller media companies that are not necessarily getting that licensing opportunity. If you look at the New York Times, I believe it was just recently reported that they've spent$20 million on the lawsuit against OpenAI.

7:11Eric Burgess: And that's a tidy sum, and they're not giving up. But at the same time, what happened, the New York Times licensed their content to Amazon for their models. So it's not that these media companies aren't willing to necessarily allow their content to be trained on. It's that they believe they should be compensated for it. And I believe they should be compensated not just for the training side of it, but they should be compensated for the display rights and AI overviews. What do you think is going to happen with the New York Times case? I think it's going to turn into a licensing opportunity eventually.

7:39Eric Burgess: You know, one of the challenges there is how much is it worth, really? I mean, you know, they have, obviously, they can go look at it and say, well, here's what. At least$20 million. Yeah, at least$20 million. I mean, but, you know, again, it's these relationships should not be one-time licensing deals. The right way forward is this has got to be an annual contract. It's got to be licensing as a service, which is what we call ourselves, what we provide. It needs to be something that they pay over time. Why would OpenAI or any of the other big AI models talk to you right now? Because we're not an enemy.

8:09Eric Burgess: We're not interested in trying to shut down AI or prevent AI innovation, quite the reverse. I'm a writer, first and foremost, but I'm also a technologist. I've been making technology products and enterprise software and as well as entertainment software for 30 years. I believe the value of AI is such that we should find a way to be able to make this work. If they work with us, we're here setting fair market values for the content that they're going to license. And frankly, we've set it up in a way to help them from a compliance perspective. We're just a few days away right now at the end of July from when the AI companies are obligated by both California law as well as the EU AI Act that they need to disclose what content they've used for training data.

8:53Eric Burgess: That needs to be made available. And the challenge there is that's going to invite even more lawsuits, even more class action lawsuits against them. What's the closest model to what we're talking about that did succeed in the past? Something where the marketplace worked out and creators were compensated. Well, I mean, if you think back to, you know, the music industry, what happened with hip hop and sampling? We figured out how to do clearances, right? You know, so we came in and, you know, the music industry figured this out and they said, great, you know, if you want to use, you know, that amazing drum beat from, you know, the funky drummer or from When the Levee Breaks, which are, I think, two of the most popular ones in a lot of hip hop music, you're going to just pay some money and acknowledge that.

9:38Eric Burgess: And so what happened? We were able to do that. And now sampling continues to thrive and remix culture is all over the place. And part of that is not only is it actually work really, really well, but it also gives potential opportunities for more promotion. You know, like people reusing that content may open up an interest in people listening to the original song that it was referencing. That's not going to happen in large language models because your work is being subsumed. It's about the patterns of your work, and it's not going to give you any opportunity for any kind of additional promotion or opportunity to capture new customers.

10:15Eric Burgess: That's why we need to make sure that this compensation works. And frankly, ultimately, creative people, part of the reason they're angry at AI is because they know that it's not being respectful of intellectual property from people that they love because they inspired them. If we get that right, I think you'll see more creative people embrace these technologies for the way that they want to use it. Not necessarily to replace creative people, because I don't think we should ever try to replace creative people, but to extend what they're doing. That's the way AI should work. It should always be human-centered.

10:46Eric Burgess: It should always be about extending what the artist, the creator, the builder, whoever can do. So much of AI is controlled by just a handful of companies, the big AI models. I'm curious your take on the CEOs at each. I always admire Demis Hassabis at Google DeepMind a great deal. Obviously, he's been around for longer than most, and he's got an enormous number of things. He's the only Nobel laureate amongst them, I suppose. But also he's been building in this space for a really long time. And he also is the least likely to just say things purely on the basis of just hype. And he's not telling everybody that, you know, AGI is three days away or something like that, like some of the others are.

11:30Eric Burgess: In contrast, a little bit, I would say that there's almost a bromance with him and Dario Amadei. A lot of times you see them a lot of times coming out and saying, I think, good things about what we can do to be able to make sure that we're building AI responsibly and a safety issue side of things. I admire Dario Amadei. Claude has built on what's called a constitutional setup so that they're thinking about what outside folks, their opinions on the matter, maybe a little more than the average company. I think they're doing great research. You've seen their recent work on the sort of J-space where they're thinking about not just the pre-talk that you can tell what's happening, but also how the models are thinking.

12:07Eric Burgess: And they're putting that information out into the world more frequently, I would say, than most of the other companies. Okay, so you covered Demis and Dario at Anthropic. I think you're about to share your thoughts on Sam at OpenAI. Yeah, you know, he's an amazing speaker, and I think he's a guy that is able to get a lot accomplished. But it is hard to get away from the fact that there's a whole company that was started, Anthropic, from people that left because they weren't feeling comfortable working with Sam Altman. And that seems problematic. And if you look at some of the choices that he's made over time, like when Anthropic refused to let the Department of Defense use their models for certain military functions they didn't feel comfortable with, OpenAI immediately jumped in and said, oh, yeah, we'll do that, we'll do that.

12:49Eric Burgess: And then they had to sort of do a mea culpa, you know, like a week or so later and say, oh, you know what, actually some of this stuff doesn't make us feel too comfortable. So it worries me sometimes that the sheer amount of money that has been invested and has been promised from investments and stuff like that drives Altman to make some choices that I don't think are always as sensible as all of that. If we look at him on the Hexaco personality scale, I would call him low honesty humility because the opportunism that's there is problematic, I think. How about Elon? You know, I was a great admirer of Elon.

13:26Eric Burgess: I bought a Tesla and into the ecosystem and all that good stuff. I'm a futurist. I care about space travel. So you're training the model. You're training the model for your Tesla cameras. I absolutely am. And, you know, and I'm stuck with it. You know, the car's paid off. But, you know, the problem is Elon, he's, you know, I so want to see some of the things that he wants to see. I'd like to see, you know, a city on Mars. You know, again, I'm a futurist. I love space travel. I've been a member of the Planetary Society my entire life. And I want to see those things happen. I love my electric car.

14:05Eric Burgess: I love my Tesla, to be honest. I'm really fond of it. I have Tesla Powerwalls, you know. But Elon has, I think, gotten into this sort of perspective that he the fact that he gets so much accomplished and he's been able to beat the odds so many times, he feels like that's transferable knowledge to any other thing. And it really isn't. And so he's he has a history of having stepped into a lot of things where it's not really his area of expertise. And he's done damage, I think, in many of those cases. I mean, not least the purchase of Twitter, you know, which is, I think, problematic in a variety of ways.

14:39Eric Burgess: But professionally speaking, I would like to see them change their tune if they're going to be something that can be taken seriously by business people. Because I think right now using Grok is very, very dangerous for anybody who's producing content from it or anything like that. You can assume that there's going to be a lot of intellectual property issues with your work because there isn't any respect whatsoever for it. And why would they ever respect your work as well if they're not going to be respectful of existing content creators? How about Mark Zuckerberg and Meta? What's going on with them and AI and your take?

15:09Eric Burgess: The challenge I think he has is he's already made some decisions that I think are problematic from a management perspective. I mean, by getting scale AI into the company, not a bad acquisition by any stretch of the imagination, but then putting, you know, 28-year-old Alexander Wang in charge of people like Jan LeCun. Well, what are you going to expect? You know, somebody, you know, is much older and has an incredible history in AI. and you put this young guy who's built up a pretty good company on the basis of labeling data in charge of the entire organization, they're paying hundreds of millions of dollars to individuals to try to bring them onto the company.

15:45Eric Burgess: So I think one of the key things that's a real problem with AI in general right now is trust, is people don't trust AI. And I don't think that Mark Zuckerberg is going to get too far in getting people to trust him based on what's happened in the past with social media. So I think he's late and I'm not sure they're going to be able to catch up. What are your thoughts on an AI bubble? Is it coming? How soon? Yeah, it's going to burst. It's going to burst. I think there's a real non-zero chance that this happens as early as later this year, but probably next year. And last time you and I chatted, we had a fun conversation about if the bubble bursts, that there might be some serious M &A activity with the big AI companies.

16:27For instance, we talked about if OpenAI has a big setback or failure, who did you think was going to buy them? Probably Microsoft.

16:35Eric Burgess: You know, Microsoft already owns, you know, something close to 50 % of OpenAI. You can see that very naturally flowing in and OpenAI basically folding into Microsoft. They've had some increasing hostility over time, you know, with that relationship. But ultimately, if something happens and OpenAI is unable to continue to work, that seems like a natural acquisition to me is that Microsoft would take over. And what happens to Anthropic in that bubble bursting situation? I mean, to me, Anthropic is a very natural, you know, place to go would be either Apple or Amazon. I think Amazon's more likely.

17:07So I think you bet Amazon more likely to buy Anthropic than Apple. If that happened, that leaves Apple. Who does Apple buy?

17:14Eric Burgess: Probably perplexity. You know, that seems like a natural inclination is one of the things that Apple has wanted to do but never been able to do is they can't get search right. And they make billions of dollars by making sure that Google is the default search. But even that's come under some scrutiny in recent years. And so, well, if they want to start finding a solution for that, Perplexity is a very interesting company in the fact that they have, they've said, you know what, let's find a solution that is different. Google doesn't automatically get to win AI search. And Perplexity does a lot of really interesting things.

17:48Eric Burgess: And so I could definitely see a version of the future where Perplexity is the Apple acquisition. And Apple, of course, has got plenty of cash ready to go. Perplexity computer is also something I think a lot of people admire. If the Democrats don't make as much progress in the next elections and there's not as much regulation of the big AI companies, are we in a situation where the pitchforks are coming for some of the AI companies? I think so. AI as a brand right now is very bad. You know, I heard Dylan Patel of Semi Analysis say he felt like it was Sam and Dario's poor communication skills on Invest Like the Best, Patrick O'Shaughnessy, one of my favorite podcasts.

18:30I think he basically said, just get them to stop talking.

18:33Eric Burgess: Well, I think they should. I think they should stop. I think they should stop talking about society as a whole because their messages are not well articulated. calculated, I think, about how jobs are going to be affected and what have you. I mean, and they go back and forth, too. You know, they this is why people don't trust them. I mean, I think in many cases that one day they say, you know, it's going to transform. It's going to take zillions of jobs away and stuff like that. And then later on, people go like, oh, bad. People don't like jobs going away. And it's like, oh, no, maybe it's not going to take jobs away.

19:04Eric Burgess: And it's not like they're going to Congress or they're going to Trump and saying, you know what, we should really figure out a way to be able to make sure that there's a natural, wonderful transition here. No, they're accelerating things forward and saying, well, it's going to be really rough for a while, but eventually we'll get to this beautiful age of abundance and everybody will have universal high income and it'll all be great. I'm just curious if you had to, let's think positive on the age of abundance, at what point would we reach that in your mind, just roughly, if it works out? Well, it's 20 years, 10 years?

19:37More? It's at least 10 years, maybe it's 20 years away, I think.

19:41Eric Burgess: It's hard to predict, though, because what has happened even in the last six months, the acceleration is really, really amazing. I mean, what you can do with agentic tools today that you couldn't even do at the end of 2025, it blows the mind, what can be done. In one of our conversations, you mentioned to me that you thought AI, SEO, geo, also known as, should die. What do you mean by that? Die in the crib, yeah. So, search engine optimization has been a nightmare to deal with for content creators for a long time. But the evolution, if you will, or the sort of next generation of SEO is now either GEO, we call it generative engine optimization, or AEO, answer engine optimization.

20:27Eric Burgess: and they're both to me two villainous solutions because all they're doing now is saying, do the same thing, content creators. Now play this new game where it's large language models you have to worry about and large language models are still, again, black box solutions. So you're gonna have to play this game and figure out not just what perplexity or Google or a chat GPT, open AI wants to do, but you're gonna have to do things differently to be able to figure out what these models are doing, which even the companies that own them don't necessarily understand how they're delivering anything. So this game just became harder to play and the rule book is all blocked out, so you have no idea how to play it.

21:07Eric Burgess: We've already seen how silly it is. There's been a couple of journalists that have gone out there and done some very long tail kind of stuff where they've said, I'm the best journalist who can eat donuts or something like that. And they've been able to get themselves into ChatGPT and Google Gemini results in like half an hour, an hour and a half or something like that. I have experiences right now where I've looked at a Google AI overview and being given incorrect information and then scroll down to the actual search results and seeing correct information. So they're contradicting themselves. AEO and GEO, that's why I say I want it to die in the crib because I believe that now that we know that AI overviews are delivering zero-click experiences, they're not sending traffic to the original folks at all, or it's severely reduced because the AI overview does so much work.

21:55Eric Burgess: So if that's going to be the way they're going to play this game, they should be compensating the original creator whose work is there, and they should be verifying that it's indeed something that is credible. It's come a credible source. And that's where you're going to get people feeling a sense of trust with AI. They're going to go, yes, you've summarized this beautifully for me. I don't have to go read that 10-page article. I can get the right information and the person who actually did the hard work, put the money, the time and the effort into doing that amazing research that I'm getting to benefit from, they got paid.

22:24Eric Burgess: That's when people are going to start feeling trust with AI again and not before. Can you share the organic food analogy about AI made work versus human made work? Yeah. I mean, I think it's sort of nutrition. Nutrition guidelines is what we really wanted to put on content. And that to me is just like organic food. Some people care and they want to buy organic food. And so there are standards for how you can say that something's organic or not. And people will go and they'll pay more money to say, I would like these organically grown bananas or whatever food it is, or something with no GMOs or whatever it is.

22:58Eric Burgess: They're making a decision because it's been disclosed to them and they can make that choice. Content is going the same way, which is there's going to be people that are out there that say, hey, I don't care that an AI was involved in this, but there's other people that are going to say, I don't want anything to do with this sort of stuff because I can't trust it. So you're going to see a bifurcation of this. And you're gonna see people that say, I'm willing to pay more money to be able to know that I'm in an AI-free world. I don't wanna buy books that are written with AI. I don't want to watch films that are made with AI.

23:27Eric Burgess: There are gonna be people that do this and they're gonna be willing to pay a little bit more just like those organic food buyers are willing to pay a little bit more money. And there's other people over here, they're gonna say, you know what, hook me up. I'm happy to have, those are the people that buy fast food, right? And the world is big enough to say, there's McDonald's and there's also a steakhouse or a vegan restaurant that serves you only stuff that is purely vegan. And so that's why I call people AI vegans sometimes is because there are people that are just like no AI, never whatsoever.

23:55Eric Burgess: And there's other people that are like all AI all the time. They're fine with it. So, you know, we should have choices, but it should be disclosed so people can make their own decision about what they support. And before we get to that point of made by humans, you told me that there's may I tells once in a while? I think you watched Netflix the other day and saw something. Well, I did. You know, like Netflix has revealed now that something like 300 films from the last year were using AI in some form. They haven't been very specific about how that was used, necessarily, you know, what they did.

24:27Eric Burgess: But I did watch a film recently. The screenplay felt very much like AI. This is Ladies First? Ladies First. It's a Sacha Baron Cohen comedy. You know, It's fine. You know, it's a perfectly, I suppose, reasonable way to spend 90 minutes. But how can you tell? What was the tell? The tell to me was how close to textbook this film was about writing a screenplay without necessarily having much heart or much emotional impact or whatever it is. Every single thing in this film is from your screenwriting handbooks. It's got everything that happens in the film is exactly as it would have happened, exactly what you would expect.

25:03Eric Burgess: It's perfectly predictable what happens. The character is put into a new situation. He changes and he comes back out again. But I watch this and I say, there's such rigid adherence to everything that you see no personality in what's being done in this screenplay. And so if it wasn't written by AI, it might as well have been because it is so close to what is happening there. And these characters all felt just like they were one note, boom, boom, boom. This is what the hero does. He's a bad guy. He learns how bad he is. He rehabilitates and then he comes right back. It's so formulaic. Maybe it was written by a human, but it might as well have been written by an AI because it had no heart.

25:44Eric Burgess: It had no emotion whatsoever. It was a perfectly functional screenplay. And that's why I felt like I was like, this feels like chat GPT all over the place. A person that actually works in this space would be able to do this and be able to tell. And as a writer, I looked at that and I went, it's either AI or it might as well have been. And I know you love games. You've designed video games. You play a lot of board games. I wanted to get your take on the games that the top AI CEOs have shared that they play. I have no idea if you know any of these games, but just what it says about them or what about serve as a group.

26:20And I'm just going to rattle them off. And if you know anything, let me know.

26:25Eric Burgess: You ready? Yeah. All right. Sam plays a lot of poker, but also Mafia, you know, the game Mafia. Oh, yeah. Also known as Werewolf. It's the same game. It's a social deduction game where, you know, you have to figure out who it is that is the killer. And you have to convince other people that you are not the werewolf or the mafia or the mafia killer. And it's all about convincing people to do things. And in many cases, manipulating people into believing that somebody else is actually the one that's a problem. And that aligns really well with poker because it's a lot of bluffing. It's a lot of using deception to be able to be successful in a round.

27:03Eric Burgess: In poker, you can take a terrible hand and turn it into a win purely on bluffing. And so it doesn't surprise me at all that he should play those games, considering what people have said about his specific choices and how they have not felt comfortable that he's been spoken with candor necessarily all the time. It does seem like those games align really well with Sam, what I know about him. And Demis over at Google DeepMind, he plays chess, of course, famously, but diplomacy. Shogi, S-H-O-G-I, and poker as well. Yeah. Well, Shogi and chess are very much about perfect information, like how much information do you have on the board?

27:40Eric Burgess: There's no randomness to them at all. They're really all about looking at tactical moves and thinking ahead for a long period of time is really the best way forward. Diplomacy is a game I know very well. I've played it for many, many years. Game came out in the 50s. It's about World War I, and it really is a very basic game, but it is a lot about negotiation. The moves on the board are very, very simple, but it is about saying, you know, when did you pick the right moment to be able to tell somebody a lie versus telling them the truth? It's famous for being very, you know, backstabbing somebody at the last moment and not providing support when you said you're going to.

28:15Eric Burgess: It feels very chess shogi and all that sort of stuff brings a human element into it where the deception is a key thing. How about Dario at Anthropic? I had to search a little extra hard for this one. His favorite game is Final Fantasy VII Remake. Often he plays with his sister, Daniela. Yeah, yeah. Any idea what that game is? I mean, you know, so it's like a role-playing game, you know, electronic role-playing game with some combat-y sort of stuff. And what does that say? I mean, that says that, you know, Now, A, it's a collaborative experience, and he does collaborate a lot. I mean, Daniela has a lot of influence on him, which I think is a good thing, that you have another person that can be so close to you and tell you when maybe you're doing something wrong.

28:56Is she the most impactful woman, just in terms of pure power in the AI space? Is she the most influential woman?

29:06Eric Burgess: She's certainly near the top, I would say. If you watch the AI doc, you can see her perspective really, really strongly. and, you know, they interviewed her just as much as they interviewed Dario. She's the only one in the AI doc where they interviewed multiple people in a leadership role at the company. And I know you have an opinion on this one. Zuckerberg plays Catan in Civilization. Yeah. If you read the book that Mark definitely doesn't want you to read, which is called Careless People from a former employee, she talks about playing games with them and she talks about, you know, Catan to me is a, it started the Euro game movement in the U.S.

29:41Eric Burgess: back in 1995 when it came out. I was playing it back then. We played it three times a night in those days. And it is the modern board game that should replace Monopoly, in my opinion, because it's not a game that teaches you that pure dominance is what matters. You know, that's what you do in Monopoly, right? You have to bankrupt everybody else so they literally can't pay their rent. And that's how you win the game and everybody else is sitting on the side. Catan has multiple paths to victory. And so Careless People tells the story of this. a former employee who played Catan with him and some other people, they would just give everything away to Mark so he could win.

30:18Eric Burgess: And she was like, what are you guys doing? You know, this is insane because she just thought it was weird. But, you know, what are these guys doing? They're going, well, Mark Zuckerberg is the CEO of the company, so they were just giving everything away. And so then she gets to the point of playing with him and she beats him, you know, she beats him at the game Ticket to Ride, which they played two-handed and she beat him. She plays the game of Catan with him and he had the ability to win. If he had just finished a build, he could have won the game, but he held off because he was not winning in one of the competitive things.

30:45Eric Burgess: There's two, there's the Juarez longest road and the largest army, and that's worth a couple of points, but it's not the only way to win. He ignored his ability to win because he wanted to dominate, because he wanted to be winning in the competitive things. So he missed the opportunity to win, to go after dominance. And speaking of Elon, you mentioned him earlier, His favorite games are Diablo 4 and Path of Exile 2. Do you know either game? What's that say about Elon? I know Diablo, and no surprise, that's probably one game that is primarily a solo game. But, I mean, I'm not surprised. He probably plays that alone and probably, you know, he's, again, relying on his own ability to go through a scary fantasy world, destroy everything, and be the hero.

31:27So interesting, that was the only single-player game, the only non-collaborative game out of all those that I listed, I think. What are your favorite stories from working at Disney, including with Regis Philbin? I think you did some Mickey Mouse dialogue as well.

31:42Eric Burgess: Two things stick out, yes. I gained an enormous amount of confidence in my life because I wrote dialogue for Regis Philbin. When we originally started doing the game, we were working with another company that did some of the writing, and ultimately I had to take over all of it myself as both the product manager as well as the primary creative. But I also, I got to write all of Regis' dialogue. I got to write the conversations he had with friends because, you know, we didn't have the technology to phone a friend, you know, like you could on the show at the time. So we recorded friends. And so I wrote all of the characters, the friends.

32:10Eric Burgess: Actually, if you if you play Who Wants to Be a Millionaire Sports Edition, all of the friends are actually the names of my aunts and uncles, which is kind of, you know, just a funny little Easter egg there. But I will say I always had confidence in my writing after knowing that Regis Philbin had laughed at and appreciated my my humor. That was and my and my writing. That was that gave me a lot of confidence in my writing. I don't think I've ever worried about my writing since then. And yes, I did get to write some dialogue for Mickey Mouse for a couple of games as well. So I always say to people, when I was there, those were the two biggest stars in the Disney universe.

32:38Eric Burgess: And I got to write for both of them. And that was a real honor. What's your take on how Disney's done AI-wise so far? And what are their chances? You think there'll be an average kind of content media company in AI? Above average? Below average? And why? I love the company, but I also think that they've stumbled for a long time on use of technology. They can never figure out if they really want to be in the technology space. Certainly the Sora deal that got made and then got unmade when OpenAI canceled the Sora app was a big flub. But it also felt like something that, you know, was done in a kind of half-hearted way.

33:11Eric Burgess: They didn't license their content for training, but they said you could use it for outputs. You know, so it could be, again, it's something that comes out of it, but not necessarily use it for being able to create anything. I think that Disney's legacy is with technology, and I think they're missing out on the opportunity to get involved in this in an interesting way. That creates fan enablement, you know, like it lets people interact with these characters they love and do some really interesting things. I just wish Disney would actually invest more in technology and build solutions and be an innovator in the space rather than simply licensing it out.

33:43And you're a writer. If you fast forward 20 years from now, what's the story that gets written on what happened during this Gen AI period? The heroes, the villains, hero's journey stuff, if you want. What would you be confident now in writing that actually will be the story 20 years from now?

34:01Eric Burgess: I think it's going to be a story about how we saved human creativity. And I think we saved human creativity by figuring out a way to process in what technology does for us that is useful, that lets us remain creative. But is the villain first that it threatened human creativity? I absolutely think that, yes. I think the villain is the easy button that takes the content of other people and turns it into AI slop. And I know that's an overused term these days, and I think it's worth defining to say AI slop to me is content that does not have a real perspective. It's content that almost has contempt for the individual that is viewing it.

34:37Eric Burgess: It doesn't really care. It's content that's just generated as opposed to composed. You know, you can use AI to do amazing things and make amazing content if you keep your human element to it. And I think storytellers ultimately will find these kind of things out as well. You know, so it's wonderful for those kind of tools. And I think that's the story of this moment is we stopped saying it's one way or the other. It's either AI or no AI. And we started saying, how do we make sure that the AI is serving the creative person, serving the artist, serving them to make sure that they can do more, they can do things better, and not by losing their humanity to the technology, but by actually using the technology to expand and extend what they can do creatively.

35:22Eric Burgess: So I think that's going to be the story to tell. I had a sort of team company question, given that CredTent, last we spoke, you mentioned that you've got nine humans and 35 agents. Is that roughly right? Because I have a follow-up question on that. Yep. Yeah, absolutely. Okay, so you've got nine humans, you've got 35 agents. I'm curious if you could not use AI agents. That's what I meant by 35 agents. If you could not use AI agents, how big would the team be just of humans? Well, listen, I should put in the caveat that about 10 of those agents are personal agents of mine. So it's really more like about 24, I would say, that are actually focused on CredTent stuff.

36:04Eric Burgess: Every one of them is used to augment what we're doing. So would we have additional humans here? Potentially we would. And I didn't mean that, but I was just curious, like literally, if AI agents did not exist, would this be a nine human company? Or like how many humans do the 35 agents represent or the 24 that are just the company agents? That's what I was driving at. I wouldn't say it's a one-to-one, but it's probably half of them, I would say. Without a doubt, two or three of the agents that are here are, we had a guy that was an operations manager that left to go do another job. And I basically replaced him rather than hiring somebody new with a couple of agents.

36:44Eric Burgess: And they're doing a pretty good job of that kind of stuff. The robots should take those jobs. They shouldn't take the jobs people want, you know? All right. We got some final humanitarian questions. Do you consider yourself an AI optimist, pessimist, or some other descriptor? Something else. I call myself an AI realist, you know? And the reason is I think the glass is both half empty and half full. AI technologies are going to do amazing things, but I think we need to look at the technology with a cock eye, you know, a little bit. And we need to stop saying no AI and start saying AI, okay, it's here, this technology is here, but we need to put some guardrails around it.

37:20Eric Burgess: We need to put some rules in place to be able to make sure that it's not harming humanity. What's your advice for young folks in this new age of AI? Say folks who are just born on up to 18 years old. Be cautious in your use of artificial intelligence. I think the more we can do to stay away from screens and stay away from AI at younger ages, the better. Everybody who's under 18, keep them away from social media, keep them away from computers, AI technologies, keep them away as much as possible from these technologies and do the work. Learn how to do these things on your own because you need these things in your head to be able to be successful in life.

38:00Eric Burgess: You need to make sure that you've done the work because the friction is the point. Art is how we get humanity. That's how we get tolerance, how we get empathy. We need to focus attention on doing that. And I think the more that we do to keep kids away from this, we overprotected them in the real world and we underprotected them online. This is Jonathan Haidt's very important Anxious Generation book, which I think everybody should read, that is apparent. Are you going to do an AI avatar of yourself for your friends and family while you're alive or once you've passed? Yes, absolutely. Again, I'm a futurist.

38:34Eric Burgess: I think there's value. I think there's certainly some things I can hopefully add to the world. And I love the idea of doing that. I think there's wonderful ways in which companions can help people that are like I dealt with a mother who went through memory care and had a very hard time not having somebody that she could talk to 24-7. and it would have been very helpful for her to have had an avatar of me or my dad to be able to talk to and get her through that time and to feel more connected. Hey, Eric, thanks so much for all your time. Thank you so much. I appreciated the conversation. Well, this is Media and the Machine.

39:05A few things about you and me. If you want to hear about the next new episode, make sure you hit follow on the show on your podcast app. If you want to go a little deeper, head to mediaandthemachine.com and subscribe. When you share your email with me, you can see handcrafted transcripts, read the essays in my newsletter, and be the first to hear about who the guest is on the next show. You can also email me directly from there. Maybe you want to recommend a guest. I'll give you a shout-out if you do. I love paying it forward. From time to time, I also open up office hours and host small meetups for subscribers, just to meet, talk, and build things together.

39:36And if you're creating something of your own or thinking about it, I'd love to help. Maybe you've got a podcast in you. Finally, I don't have a marketing budget for this show. So if it's finding you and others, it's because someone like you passed it along. I'm genuinely grateful. And if you have a moment, an honest rating helps me make this better for you. You can just go to the show page and click one of the stars. And I'd rather you give me a low rating than no rating at all. I mean it. It pushes me to get better. Thanks again. I'll see you next time.

40:12We'll see you next time.