Holy Grail, or Poison Chalice? (Ep. 33)

Hello and welcome to the GameCraft podcast. I'm your host, Michalasky. I'm Blake Robbins. And this is a podcast about the modern history of the video game business. Episode 8, Holy Grail or Poison Chalice. That's a great title. All right, we're going to wrap up this season. And for the final episode of this season four, Blake and I looked back across the episodes and the topics that appealed to us this year and those topics hopefully coalesced into the first seven episodes of the season. And we wanted to find a way to bring together the threads that we've been weaving throughout this season, sort of in this final episode. And we wanted to look for something that was representative of that sort of culmination of the thing at that at the intersection of those threads. And we'll get back to what that is. So what were the threads that we found so significant this year just to summarize? The big one was obviously artificial intelligence and the impact that it will have on the game's business, what things that it will enhance, what things will it destroy, and where are their potential business opportunities? Love it or hate it, AI is going to dominate conversations in the game's business for years to come. And so we felt it was significant enough to devote a lot of time to this season. Another thread was the scaled casual mobile companies like Roblox and the huge Turkish puzzle masters at peak and dream. We talked about how these companies operate, how we understand their competitive advantages and the threats to their continued success. We also looked at the broader threats to the game's industry from the attention drain of feed-based services and gambling, how these engagement engines have sucked so much time out of our human world, seemingly at the expense of time spent on games. We talked about what these services learned from games and what games can learn from these services. And near the end of the season, we came back to the thing at the center of it all, play. A theory of fun, the experience of play, and why that experience is so valuable to us as business people when we're thinking about what comes next in the game's business. So for this final episode, we wanted to focus on a new crop of companies that sit at, like I said, that intersection of a lot of these trends. The AI enabled user-generated casual game content portals. It's a bit of a mouthful, but why do we think they're so interesting? Because these products have the bone structure and the hype structure of something that is total-cap-nip to the investment community, cheap, near-infinite content, creator tools, platform dynamics, IP capture, and the promise of new aggregation layers for games. These things are going to get funded, probably too many of them. And so we felt it was useful to think about them and to talk about them, really, to talk about them more broadly in how they'll function in the video game business. And what they ultimately mean for the video game business as a whole, are these things good for the game's business or bad? Yeah, this is a man. I mean, when we were drafting up potential episode topics, this is one where I think you and I have seen variations of this page over the past decade, but none fully captured the AI elements, which is really, like you said, as the bone structure and hype structure that is going to cause this category to explode. And so I'm really excited to be diving into it because I think one, everyone is going to hear about these type of platforms, but two, there's a lot on pack in what these platforms are actually trying to enable. Yeah, and how they work and will they work, right? Because I do think that a lot of them come from people with games backgrounds, a lot of them don't. And so it's going to be interesting to talk about. So you know, what I think, rather than what we would normally do, which is sort of go through the history of casual games portals and look at antecedents going back to the early days of the internet, et cetera, I think we're going to do it in reverse order. And Blake, I think it'd be useful if you just start, and because you have a lot more experience of this than I do, and take us through this current wave of AI-enabled casual game micro platforms for lack of a better word, and explain to us kind of what they are, who funded them, how much money they've raised, why they're significant. And then we can go back and look at them in the context of history, which may be a sort of not unusual approach, but maybe an interesting approach for this particular set of companies. Yeah, I mean, I think you hit to know on the head, which is like, we're starting to see a new crop of companies that are trying to build this almost like prompt a game. We've used vibe coding, we've used agentic engineering, we've used all the terms, but these platforms, each of these that we'll discuss are all trying to basically act as the aggregator and the creation platform for these games. And there's a few that, like, there's, look, I've been pitched dozens in the past year. And so this is not meant to pick on any in particular or call out any in or shine on any in particular, much more of these ones I just think cover a pretty good grasp of the overall answer. These are representative. Yes, that's exactly right. And so one one that we have to talk about is there's a company called Astrocade, which is actually founded by way more of your technical or or AI background founders. So way more on that side than games business side. Actually, Feifei Lee, one of the famous researchers on the I side is a co founder of Astrocade. And Astrocade just came out publicly as raising 56 million from Sequoia and a few other major funds. But Astrocade really is, they described as a UGC gaming platform where anyone can turn an idea into a playable shareable game and days and not months. And like, if you go to this platform, it's all web and primarily mobile web, honestly, it feels reminiscent to some of the other platforms that we're going to talk about later on of Dekongriez or Antikton games or many clips. We've seen this before, right? The differentiation here is that the integrated creation platform operates off of a chat box window, right? Yes, it's basically a prompting window. So you're not having to kind of download tooling or use your own, roll your own, whether it's, you know, flash in the old days or HTML5 or whatever it was that was required for these sort of in browser game experiences. I think this one is an integrated experience where play and creation are kind of co-existent. Yeah, exactly. And actually, they frame this, which I think is worth calling out. They explicitly frame it as think YouTube, but for games. And I think that captures a lot of what we're touching on here, which is the other product that we're going to talk about here is Sakai, which is a competitor that has raised 30 million from Kosoa and Andreessen and Mayfield and a handful of others. They frame themselves as we're TikTok for games. I mean, we're going to hear a lot of these elevator pitch, you know, shorthands because again, I think if you think about who the audience is for this is, it's as much for the investment community as it is for the creator community or even for the audience that's intended to play these products. So I don't, I'm not surprised to see the YouTube of games at the TikTok of AI mini apps. It's like, yeah, we're going to see a lot of that bullshit as we sort of go through this. But I, but in some cases, it's useful to sort of think through that. And I think we're going to try and do that a little bit later as to what they mean by that. For sure. And so Sakai, I would encourage everyone to try out Astrocade and Sakai is Sakai almost looks like a one to one copy or a clone of Astrocade. Like the experience is feel very similar, but Sakai does have mobile apps. And they put a little bit more emphasis on the remix side, but also driven by chat bot, go and talk to it. And that's what game you want to make. I think these platforms in their current state look a lot closer to hyper casual type games way more of these bite size snack size games for lack of a better way to frame it. But they all are, both of those are getting traction unclear where the traction is coming from. But those are, you know, have raised quite a bit. Yeah, it'll be interesting. I mean, just comparing the two. It is kind of interesting to look at them side by side because they are so similar. And they do take on this idea of a limited amount of kind of categorical curation. I'd be really interested to know how they curate within that because it doesn't seem like they're doing it based on the number of on numbers of plays. Because as you sort of look across, you can see they're not really rank ordered in that necessarily in that order. But, you know, like, you know, there's a survival or an arcade category. And every, you know, it's a bunch of games that look really similar. Yes. And maybe worth calling out is Sakai is founded by a couple X by dance and TikTok engineers. And so to the extent that they worked on anything related to the algorithm or curation, it will be interesting to see if that shines through. But just the contrast of astrocade, obviously, being non-game games team way more on the technical AI side, Sakai is similar, non-game team that is X by dance track. Big by dance being the TikTok company, at least in the media space. Yes, yes, exactly, exactly. Another one to call out here that is not explicitly games, but it's called Gizmo, where they would frame it as, again, another new creative platform where you can make sure and discover interactive experiences. This is the one that Facebook bought. Yes, yes, yes, so meta acquired this about six months ago. The CEO was X snap and That is still actually out there meta acquired it in one of those classic like Aqua higher license detect type deal. I suspect it not trip antitrust, but a very interesting really look at What you could create and how far this expands if you don't just stay in pure games or what is the idea of a game? Actually look like a lot of their experiences feel very much like art Yeah, which is if you look at their homepage the sort of self-running demo starts out basically with what looks like a kind of Microsoft paint program from the 1980s. Yeah, and so like actually the the founder at one point as I've spent time with the team This is early on he almost framed it as like clawed artifacts if you're familiar with that as like the the thing You can share out from clawed, but like packaged into a feed, which I just think is an interesting take on this So maybe the most interesting stuff that you would you would create that you would want to share to your friends But that's just another one that I very high polished the product, but in a little bit more broad direction than pure games Any idea like what that traded for? so actually first round led to seed round and it's unclear how much it it actually existed for but if I had to guess I would guess like in the blow hundreds of millions Yeah, yeah, yeah, and The the other one that's like maybe the most ambitious or most hardcore version of this that I've seen is a company called spawn their site is spawn.co They are doing it's a the founder's name is Jacob and he had previously worked had a company called Pluto that he sold the Robin Hood And like was very high up on the day I cited Robin Hood and sort of the early 2024 type bearer and left that job to go and do spawn and spawn is Way more focused on how do you actually build like a true real blocks competitor? Let's say deeper games multi-player games. I think in few asked him you would be closer to trying to build forever games through prompts This is you know for a whole bunch of reasons may like the triple black diamond of Companies you could be trying to build in at once, which is you're solving multiplayer real-time all of those things Well also building a vibe coding platform on some level is Extremely difficult, but there is some interesting remnant already on that site Well, I would I would say just to you know hearing you to walk through these things. I'm Sensing a couple of trends, right? So one is obviously all of these things are basically, you know AI prompt to mini-game or AI prompt to micro-game or whatever, right? Two is it doesn't feel like there's a whole lot of game DNA in these companies Yes, actually almost no game DNA, which is either really bullish or really bearish Exactly. I'm not exactly sure how to take that. That's right I think and maybe speaks to the fact that we've we've unfortunately Discussed at length throughout the seven episodes that have led up to this episode, which is just the The lack of of real leadership from games industry people in the in the entry of AI into the game space Yeah, and I think there's a little bit of In the same way that this is attractive to venture capital as I imagine it's attracted to a founder where the prize feels very big We and we'll go into all these throughout the episode, but People have been funding some version of Tik Tok for games or in this world for a while decades Yeah, yeah for decades YouTube for games. I'm sure you got that bitch probably in your email 20 years ago You know from mini clip in 2001 to congregate in 2007 to you know I mean there they have they have come and those were the successful ones, right? There are a whole bunch of other ones. This has been a you know an area since really Originally in browser games, right in browser experiences with flasher with HTML 5 those were enabling technologies which Originally engendered a wave of these casual game portals and then clearly then the at the launch of mobile There were a number of other other shots on goal there to try and create these little portals with multiple games that sort of could keep you occupied But with with a fairly limited depth. Yeah This is very clearly a like Trying anytime. There's a new technology or if you could be the you know We talk about distribution and vision so much on this podcast But I think this is the one that a lot of people come to is how do you try and aggregate This specific play style which is almost more of these micro games like you're talking about although I think it's still an interesting open question sort of what the competitive advantages are here, right? Because if as you say really these are distribution plays fundamentally What is the you know and everybody basically has access to the same or similar Content creation engines through AI Yes, like where is the real competitive advantage? Like why do you choose one over the other? that is Well, there's there's you know Maybe the the thing that we haven't discussed here is costs is obviously a big factor Yeah, I can imagine yeah, these these are probably not positive gross margin businesses So yeah, so it might it might be there might be a bet and maybe you ask for Katie who's raised the most is gonna argue that if we subsidize game creation at a certain point will break away from the competition But I it is it is definitely the right question to be asking which is Yeah, what is the j curve look like on these things and then ultimately once the I mean you know again I think there it wouldn't be You know all that outrageous to bet on a grow on a on a negative gross margin business In this early stage of AI right when cost has still not really normalized Although I do think we are starting to see just as an aside the kind of contours of that starting to Look a little more solid, right? I think we're before everybody was kind of like oh Computes go into zero or near zero and you're gonna be able to have infinite computer They were all all this bullshit that we heard early on in the AI revolution about you know Just throw dollars at scale here. I think we're now starting to see even corporations that have really made This a centerpiece of their strategy pulling back a little bit on you know the tokenomics if you will yes Yes, yeah, I mean look and and we'll see when when opening an anthropic actually have their S1 drops We'll see how much is there subsidizing on their on their subscription plan so I so these are really interesting I think primarily because in this environment in which very little is getting funded in the games business Except later stage stuff that's already kind of proven and dynamic or things in Istanbul or things like that These companies are getting funded and funded significantly. Yeah, yeah, and I think it's worth really going back to Talk about whether this wave is actually like going to endure I think it's actually worth talking about more specifically the web casual game platforms like the mini clip and aggregate Congregate and then maybe we talk about even some of the other pre-AI UGC platforms that we've seen over time. Yeah, so I'll start off on the on the mini clip and congregate site So you know mini clip was a really interesting company for us when we were doing jammed at back in the early 2000s because they were sort of demonstrating on the web the viability of Casual gaming and the scale of casual gaming. I mean mini clip in those early flash days pre-mobile got to like 75 million monthly active users Right, so which you know again, we're talking about 2001 to 2008 Those are significant numbers right for a company in that period and you know, they did it with a bunch of reasonably well-curated reasonably fun small games They had a little not a very interesting social layer, but a little bit of a social layer and a little bit of a kind of identity and Yeah, I think like you said they got they got to a lot of user scale But they never really felt like they broke through the monetization ceiling Yeah, yeah, I mean look at web is For better worse and possible to to monetize the team. So I don't know where we're like 30 40 years into the web games And it's still not clear that there's a great way to do it. Obviously you can do advertising But it's not anywhere near what a mobile casual game is making so true And then mini clip had a second life as a mobile portal Yeah, yeah, and in and in that guy is got to some real scale I mean, I think I've heard rumors you know in excess of 250 million monthly active Maybe as many as 400 million monthly active and you know 50 60 million daily's I mean, that's like those are big-ass numbers Wow, oh my gosh, and I have to believe they're monetizing that more successfully than they were able to monetize the web portals And I think probably the reason why we're seeing a lot of these sort of gizmo Astrocade type companies launching on mobile primarily as opposed to on the web or elsewhere Yeah, yeah, and and then Mini clip is still technically I think a like a private company is that right or I think so I don't know that much about it I haven't followed it that that that closely Congregate on the other hand started by Jim Greer great guy. We work together. I think I published his first casual game out of the University of Texas or Princeton when With a number of a couple of partners and they made a game That was it we ended up publishing as a PC game, but it was really probably a mobile game well before its time And Jim ultimately launched Congregate, his sister Emily also. served as senior executive, maybe even CEO at one point. And they never reached quite the same user scale as MiniClip, but I think they had kind of very high quality products and maybe a better approach to curation. I remember what I was telling you the story before we started recording, but Reed Hoffman at Greylock. I remember called me almost the first day I started at Benchmark to kind of ask me, he was about to write their series A check and called me to ask me kind of what I thought about it to do a bit of a gut check with him on the team and on the strategy. So I go back quite a ways with this company and a very interesting company, again, ultimately got sold to GameStop. Yep. And I think it's probably still kicking around out there somewhere. I mean, I grew up playing congregation MiniClip, but I congregate with my main platform that I played growing up. I loved it. And it's funny to like now look back and see maybe where the flaws of that model were. But again, it's likely it was all around monetization more than anything else. I think it might not have reached MiniClip scale. But if you had monetization, you could justify a lot more of these other actual UA and all those pieces. But it is very like, I think it's a cautionary tail on some model for this, which is these are platforms, congregate and MiniClip that aggregated. We're talking tens of millions, if not hundreds of millions of users at scale. And they had good to great outcomes. But not, you know, what Mitch and I would define as generational outcomes in GameStop. No, I mean, I think we've talked about this season on a number of different occasions, like the idea of profitless scale. And again, not necessarily suggesting that these companies are not profitable. But they haven't really monetized commensurate with their scale. That's exactly right. And it's, they will be a question that we will have for all platforms moving forward, which is they haven't added risk here, which is they're actually taking on the cost to make these games, whereas in terms of the ones we talked about earlier, the AI and the New World. Yeah, the New World is, they're actually incurring that. Whereas MiniClip and congregate sort of got the right to wave of both flash games and HTML, five games, and really mobile, yeah, and really mobile. Like they didn't, that was just, there truly needed to be an aggregation and discovery platform. But I think it raises the question. And I think it's a really interesting question. And one we're going to then turn to in this next section a little bit when we look at some of the things that are swimming a little bit upstream, is what are people willing to pay for in this particular area? Right, and like what is necessary in order to shake the money tree a little bit more aggressively and get some of these things that could be potentially operating at scale to monetize like some of the scale plays that are, that we've seen in other areas of gaming. Yeah, it's, oh, but yeah, there's a lot on back in this episode. I guess the other, the other thing worth calling out, which I think we said a little bit earlier is VCs were funding this idea. Oh, yeah, absolutely. I mean, look, Roblox is this idea on a certain level. Right, UEFN is this idea on a certain level. My craft modding is this idea on a certain level. We should go through these and talk about them because I think they're, like, I would call these sort of pre-AI UGC creation plans. Yes, yes. And the ones that we know that are at scale today are the Roblox and UEFN and Minecraft and like even Quake on real or version of this. Yeah, you could maybe look at these sort of engines. Yeah, I mean, like substitute AI for game engine and you're sort of talking the same language. Yeah, the modding dream, the overall pitch has always been compelling. Just like, if you can explode the supply side of game creation, can you build a platform? And so those are the success stories. I think there's quite a few that are-- There's definitely a graveyard there. Yeah, sure. And you know, we're not even talking about graveyard ones. We're talking about ones that are still out there. And you know, there's high pipe, which was had come out to gate trying to do TikTok for games. And I think by all accounts has not found any real success. At scale, mainly I think because they-- which we'll talk about all of the learnings that I think from now, which is a bunch of different control patterns. There's definitely trying to-- you end up in Roblox territory pretty quickly. Yeah, let's come back and talk about that. Because I think it suggests something more fundamental about how these things work. But yeah, just to full disclosure, high pipe is a benchmark investment, although not a one of mine. It was one that was made after I had retired. Yeah. And then there's play bite, which is actually-- was originally pitched as TikTok for games. And that is not like that company is as long to add. And I think very similar learnings to maybe high pipe. And then there's a handful of others that are out there as well. Yeah. I mean, I did one kind of UEFN adjacent called Manticore. It's still kicking around out there in the middle of a pivot. But that wasn't originally intended to be something very similar, a UGC engine on top of the Unreal engine that allowed you to create very quickly and build sort of smaller experiences around 3D games. But really, you think about what separated the ones that kind of worked from the ones that didn't work. And I think one of the things that these things have in common because remember that this dream of TikTok for games-- let's just use that dumbass metaphor and sort of explore that a little bit, right? What the way that TikTok works, you're scrolling through these short form videos, you don't have to relearn how to watch a short form video every time you scroll. But for some of these game portal type platforms, that metaphor of TikTok really breaks down because every time you're scrolling to a new kind of experience, you are having to relearn some degree of control, play pattern, et cetera. And that friction is really limiting to that idea of the TikTok for games or the YouTube for games. Again, we have been trained since we were tiny children watching cartoons on Saturday mornings, or at least in my ancient era, to watch and decode linear media, to extract stories from mise en scène and cutting and editing. We don't have that same vocabulary, necessarily in the game space, at least as fundamental and as deeply ingrained. And so in a case of something like high-pipe or manticore, every time you went into a new experience, you had a relearning that you had to do. And a lot of people bounced off that. Yes. And I think it's just so important to call out because there's another piece of this, which is if a video is bad on TikTok, it actually worst case scenario, you watch till the end, and you swipe after. If a game is bad in one of these platforms, you've opted in in a way that is different than opting into a TikTok video. Yeah. You have your lips it out. It is definitely a heavier lift. At that point, you might just exit out of the whole app. You're like, what was that thing? What was I supposed to be doing? Why was that buggy? So why did Roblox and to a limited extent, lesser extent perhaps, UEFN or Minecraft work, where these other platforms were less successful? And you know what? I think that they have something in common. And that is in the case of Roblox, they did a really interesting job. And this, as you said before, was not necessarily by Fiat. This was something that emerged kind of organically, which is that a lot of the games adopted the same controller system, and as well, the same kind of camera structure in terms of what the view was of the action. So in terms of relearning mechanics, they really limited it to a certain extent by allowing you to leverage the knowledge that you had about how the control system worked. And to know that the view you were going to see was kind of standardized. Yeah, it's one of these like really, that maybe hindsight is 2020 for Roblox. But like, it is amazing where you can jump in almost any Roblox game. You'll at least know how to walk around. You might not know the core intent of what you're supposed to be doing in that game, but you're going to at least be able to walk and move into the basic objectives of those games. And I think that is a huge differentiator for almost all of Roblox games. And actually, we've talked about in the Roblox episode of there's some of these maybe more hardcore games and when you move from that variation or you go in the, there is some first person games that exists on Roblox and those largely struggle. Yeah. And I think you know, you've talked to me also before about like how Roblox's use of, or the Roblox creators use of the blocky characters and how you have kind of a common character or whatever is more amenable to like decoration and cosmetics and other things like that, right? Because you do, again, you don't have to worry about that. Okay, the stuff that I bought for a character in one game isn't going to work in a horror first person shooter 3D game or whatever. And maybe part of the reason why some of those games are struggling on the Roblox platform because they don't leverage the existing play patterns and visuals. Yeah, it's, Roblox is so fascinating. It's specifically around this vector of like, whether intentional or not, a lot of demonetization comes from these avatars and the amount of time they invest in their character. And if you don't let someone use that character, like what happens? Like do people clearly don't want to play the game? games where they didn't get to use their character. And that is like, for better or worse or whether it was intentional or not, the core monetization and the user experience like are super tightly tied, which like that is to build one of these platforms, you need to nail every single aspect of that. But you really need to nail that. I think you hit it right on the head, right? Which is like that has to be really, really synchronous, that experience of play and monetization. And I think that's really harder to do than a lot of these people who are coming in from outside the games industry fully understand. And you know what, look at the end of the day, they're probably going to end up being advertising supported. But in order to do so, they're going to have to get to massive scale as we've seen, right? You don't get a big advertising business with a small user base. Yeah, and well, I mean, it also would just be a shame that that's like the way that you make money in games. I don't know if I totally agree with that. I mean, I don't, I don't particularly like ads as a user, but there are plenty of people out there who'd rather watch an ad than pay for something. Well, maybe, maybe, maybe it's a better way to friend that would be you were just under utilizing the medium, which is, which is, this medium is so rich in ways that you can do cosmetics and in-game purchases and all of that. And to have it be reduced down to ads feels sad. Yeah, but before we move on, though, I want to talk a little bit about UEFN and Minecraft, because I think those are two really interesting examples. UEFN, as an existing example in Minecraft, as I will say a theoretical example, UEFN actually has a, goes further in the sense that like, it almost feels like everything's a Fortnite mod on a certain level, right? Even if the play patterns change a little bit. And I think, again, while it hasn't reached the, the scale that Roblox has reached and probably has an older, a little bit more core user, I don't really know. I've never seen the demographics, but I would just postulate. It does share some of that same bone structure around consistency, which allows for user-generated content to not pollute the system with too much of that friction and heavy lift that we described earlier. Yeah. One of the things that really opened up UEFN even as a, and they made this clear as a bet for them was they made it so all of the monetization and revenue that came from the skin shop, also get shared with UEFN creators. And so much of this is how do you incentivize creation? And for them to have the monetization already tied in, basically goes back to this other point that we just made on Roblox, where, of course, you need to make it this third person show your avatar and we'll play these games because they want you to go into the skin shop and buy more these things. Yeah. And then as we've talked about a million times, like getting from zero to one, from no monetization to one dollar, is way harder than getting from one to 10. So the fact that they've already got these credit cards on file, the fact that they're already monetizing these skins or weapons or whatever and other ways or dance moves or, you know, there's already, there's already a monetization engine at work underneath it. The idea that you're tapping into that as a creator on a certain level and basically I'm sure getting a pro rata share of it based on how much engagement slash, you know, scale you're able to achieve within the platform. That's really significant because that it's again, what it's, it's, it's more difficult than I think a lot of people understand for these new AI enabled platforms to get to that initial sort of zero to one. Yeah. And look, I think it's, it's worth calling out that Roblox and UEFN and Minecraft and all these aren't going to just sit on their hands when it comes to. No, I mean, Roblox is already building AI too. I mean, they've already announced that they have an MCP that you can, you can use with cloud code like you can use Roblox Studio. And so I think while this is a very juicy area, we should assume that on Reels go into allow, you know, and Fortnite as a result of that. Well, allow prompting of some version for the games and look Minecraft. I mean, with, you know, with Matthew Ball now going in as the chief strategy officer and with Asha bringing in a bunch of her team from the core AI group, I have to believe that Minecraft may be heading in that direction as well. And frankly, I welcome it. I think it would be great if Minecraft opened up and had and was more sort of explicitly sort of UGC oriented. Yeah. And so I like, I think the as we talk about these new platforms, the incumbents, they're different incumbents than even, you know, some of the like traditional tech incumbents in that they already have this player and they've, they've already aggregated, you know, in the case of Roblox or UEFN, they've already aggregated user on a UGC platform. Yep. And they've solved a lot of these hard problems that, you know, you and I are discussing most, maybe most importantly, they've solved monetization at, at scale. Right. So it really raises the question, what is the core and competitive advantage of those companies we discussed at the beginning, right? What, what makes the idea of just building a games portal around AI as a creation engine fundamentally that great? And I mean, obviously you get more content faster, right? And so, but the question is, how do you figure out what of that content to prioritize and what and and foreground? Yeah, I think the it's the other piece of this is, you know, we talked a lot about Tiktok's algorithm and how Tiktok's algorithm was amazing at figuring out and and ingesting like, oh, they actually liked that or they they interacted in a certain way. In games, it's so hard to tell what is what a player finds fun. And especially in a UGC context of like, I don't know if Roblox has a read on whether they think I'm having fun when I play certain Roblox games. Yeah. They can use time and engagement, but we also know some games have short times and they can still be fun. Looking back at the YouTube example, for example, you know, we know that YouTube is what greater than 10,000 to one consumption to creation. Yes, right? And if that's the case, then does broadening the base of the pyramid of creation, which is really probably what the pitch is from these these AI enabled portals does increasing the size of the base of the pyramid in terms of number of creators actually result in like a better platform. Yeah, this I mean, look, this goes back to something that you and I have discussed a bunch, which is like, do we have a like number of good games shortage? Like, I don't like is there a shortage of like enough games in the work? As you've said to me, even like pre AI, we were we had a surfeit of games, right? We had too many games for the for the available distribution networks. And you know, as we've discussed in the context of some of the AI tooling and AI companies that we've we've mentioned in earlier episodes, it's not as if like more is necessarily going to solve that problem. Yeah, I'm certainly not. And so maybe maybe the to answer your question, like, what a moat looks like your maybe is curation and that like me like, if I was to play out a version of this, it's like something that maybe doesn't look close to a game that we would think of today and they've curated that at core, but are at scale, but yeah, it's interesting to think about. I actually think the curation is the right answer, right? I believe that it may actually be efficacious to do one of these in a particular genre with a particular kind of like limited range of art styles, like really like learn from sort of Roblox and UEFN rather than, you know, you go to the homepage of something like the companies you mentioned earlier and there's, you know, it's just a dog's breakfast, it's like there's tactics, there's zen, there's chess, there's memes, there's simulation, there's farm, there's trending blah, blah, blah, blah, it's like that's not curation, right? That's just categorization. And there's a big difference between curation and categorization. And I think that look, this market in general is crying out for curation. Yes, yes, I couldn't agree more, which is go and pick a really juicy genre or maybe a place where it's underdeveloped and from a number of games that can be made there and go and explode that. But I think focusing on like a specific version of that is absolutely key. My favorite version of this is actually Nintendo of all people have a game called Super Mario Maker and it's like sold, I don't know, like 50 million copies, something crazy. But the whole thing is it's just UGC, you know, platformer Mario levels. But my favorite twist of it is the only way to publish games to the masses is you have to be able to beat the level. Like that's how they solve the like, oh, there's not a bunch of games that are like incomplete. And so like it actually becomes like a pretty hardcore like platformer game, because it's just all like expert level, like flexing of, oh, like I made this really crazy level. And I also beat it because like the only way I'm able to publish this is what it would a great psychotic way to like sort your user base into creators and consumers, right? It's like, oh, let's let's do a really, really hard level. And you have to beat it. It reminds me a little bit of the scene in Halt and Catch Fire where the programmer had the tech head, she puts an adventure game on the network and then ends up firing the people who can't solve it. Exactly, exactly. It's just such a clever also psychotic way to do it. I wouldn't maybe recommend that if you're not an intelligent, but it is like a, because that's the other piece of this is like if someone's just going to publish a half published game. You also need to make sure that like there's actually a wind condition and I just always appreciate that it's like a clever way to solve it and platformers are maybe a perfect example of what you could expand on and make millions of different variations. Yeah, I think that another way we may see the similar kind of metaphor expressed is or similar kind of pattern expressed is at UGC engines inside existing games, right? So I think you know, we looked and we talked about in previous, you know, pre-AI episodes in previous seasons games like Aggie Party or Fall Guys where you have the ability to kind of create social through level creation and then inviting your friends. And you know, it has it has Roblox like elements, but it really is just like making new levels for an existing game rather than making entirely new game experiences. And again, maybe there is something there to extend the engagement time for some of these more casual games by allowing them to be reconfigured, but still preserve that sort of core fun. Yeah, I think there's going to be I couldn't agree more with that. I think there's going to be a really fun twist on this, which is maybe single player games actually adopts like an AI modding platform where you could certainly imagine adding extra missions that are UGC. And like that feels like a very fun twist on this if you just make infinite missions within GTA even. And you're like, okay, now like go and pick up stuff at this place and oh, now you have to do a bank robbery. I just I think there's going to be something around all of that too, but likely of starting at like this is the core mechanic, this is the core gameplay and expanding on that. So I think like, you know, we can we can sort of summarize. I think the bull case for these is somebody creates a roblox out of this, right? Somebody creates a UFN out of this, the next real aggregation layer. And it's AI creation driven. So it opens up, I guess the optimistic view of this would be that it opens up creation to a much broader audience because even the existing tooling or in places like Roblox or UFN requires a certain amount of technical skill, artistic skill design skill to realize. And if you could essentially farm that out to AI, which I think as we've discussed in previous episodes, while we're very optimistic about AI, we do believe that there are limits and that those limits tend to be around sort of the consumer facing parts of it, right? Our style, taste, design, things like that. And so I wouldn't be super, super optimistic about whether or not that AI alone is going to drive a sort of new kind of aggregated portal here that's going to work just because it has AI creation tools. That said, I'm always optimistic about the idea that like having new people come in and use those tools who may have been put off by the tooling of the previous generation of products may produce some really interesting and outwiring kind of content. Yeah, I mean, like it really does have the bone structure of like something that's here as a symbol of like, like you have this like infinite contents. The last medium of like a feed that hasn't been conquered, you're going to have network effects. Like all of all of the things distribution and innovation, all of it is packaged into one. It's just catnip, catnip. I mean, you're going to have a lot of people going after this, including the giants. And so I totally get why this category is so massive. I think it's worth maybe a just position of that. Like it's worth just reminding people that people are playing like fewer games. You know, that all of the data shows like they are going and staying with their counter strike or their League of Legends. And that's the game that they're playing forever. And that's actually who you're competing with. Unless you're trying to go super broad. Well, I would say Roblox may be a counter example, right? Because as we discussed in the Roblox episode, you know, the average Roblox consumer in a month is playing, you know, a dozen, two dozen. I call them experiences rather than games. So in some sense, I think Roblox is an example of a counter example, if you will, to that sort of, you know, playing fewer games. But I do think that in a lot of cases, it's sort of positioning the portal as the destination almost as the app itself, right? It's like, I'm going to go play Roblox. It's not so much that your kid tells you they're going to go play Blocksburg or they're going to go play, you know, grow a garden. They say they're going to go play Roblox. And I think that the goal for these portals is to engender and experience similar to that. Yeah, I think it's worth calling out that Roblox has seemingly taken the position of trying to build higher quality games or certainly trying to incentivize higher quality longer play duration games. And this current iteration of products that we're seeing being built in the space is completely counter to that of like what's flood with millions of games and maybe you'll find the one experience for you or go and play. I don't think these platforms are thinking of you're only going to play 12 experiences. I think they're thinking of you're going to play hundreds of experiences in one session. So what's the bear case? Matt, there's a lot of bear case here. I mean, one is obvious, right? They can't monetize. Nobody's nobody's going to be willing to pay for let's let's, you know, we don't have to put lipstick on the pig. Nobody's going to pay for a bunch of AI swap. Yes, yes, yes. Well, I think the bear case is just like does more games equal like better like an actual way to build a platform. And so far, I'm not sure that there's any signs that that's true. Yeah, I mean, although I think again, if they were really forward looking and maybe as an evolution of their current positioning, if they were like, look, we're going to create an engine that's going to allow, let's say, 10,000 games to be made. And then we're going to use AI to call that herd down to a thousand really good ones and then maybe a hundred that we're going to promote. Maybe there is some kind of strength in numbers there, but it is not really typical of entertainment businesses that that kind of an approach works. Yes, yes. Yeah, I mean, look, and I think it's, I think if you were going to play devil's advocate on that, I think it's they need to take a stance of maybe reinventing what a game is. I think my biggest knock on the current ones that we're seeing other than maybe Gizmo is they're still trying to build what looks like traditional games. And I think we should be like redefining what a game is. In the same way, Roblox, no one, no core aimer looks at Roblox and it's like, that's a game. Like I shouldn't go into some of these sites and see the equivalent of a mastery game on there. I think we should see total and new experiences. I agree. I just don't think that's going to happen. So, you know, I think getting back to our initial question, is this a holy grail for gaming, zero cost of creation and, you know, infinite scale? Or is this a poison chalice? And I guess I'm slight coming down slightly in the poison chalice quadrant here. Look, I don't think TikTok, I mean, I look at the TikTok for gaming thing doesn't ever work on me because I actually don't think TikTok was good for linear media. Right? I think TikTok commoditized both the content side and also the consumption side in a really kind of base way. And it worked. And it's really entertaining and really efficient. And it's just an amazing engagement machine at scale. But ultimately, did it really kind of improve the experience of linear storytelling? Yeah, there's, you know, been some examples, you know, the kind of remixes and some of those things that are new storytelling tropes that are kind of interesting or whatever. But there's a, you know, a lot of it is just kind of base. And I'm not really sure that a similar approach in games makes me very hopeful. Yeah, it's, when we were talking about this episode and you brought it up, I was like, man, it's like bleak if this works at scale on some level. Like it's unclear what that means for the games industry if there's a break down in platform. Yeah, if it's become, if one of these things gets to Roblox scale. Yes, yes. Yeah, I agree. A little bit strange for the industry. Awesome. Okay, Blake, so we're at episode eight of our eight episode season. And just to reflect back on it, it's been a lot of fun this year. I think we took on some really interesting topics that, you know, we had talked about you and I in the past and we'd never really gotten around too. So I was really glad that we got a chance to talk about things like Turkey to really spend more time in AI. I mean, it was something that had been on our minds in previous years, but really broke out in the first quarter of this year. And in some ways, I'm kind of glad we ended up getting a bit delayed in terms of starting our season and episode one, not really coming out until kind of, you know, the end of key one, because had we started earlier, I think we wouldn't have appreciated the degree to which things were going to change. Yeah, I mean, look, similar, similar reflections for my end. Like this podcast is always a real treat for me. I think, you and I talk about the games industry constantly, but to actually, you know, for me, the prep and actually preparing for these episodes really helps me sharpen my thinking and our conversations makes me feel a lot better about like seeing the present clearly, which is so much of what my job is day to day. And yeah, look, I can't say thank you enough for all of that. - Oh, no, no. And I'm super grateful for your friendship and for the perspective that you bring to this because I sometimes feel like a bit of a dinosaur and taking on some of these new topics. And so I greatly appreciate it. And I thank you also very much. And I thank our listeners, many of whom I have met over the last four years. And I'm always astonished by who they are. We have just a crazy audience for this podcast. And one that we could have never imagined when we just started and we figured that there'd be 10 people who'd be interested in hearing us blather on about the games business. But in fact, we've reached a lot of people and a lot of people in really interesting positions in both the games industry and outside of the games industry. And I'm also super grateful for everybody who listens to the podcast and who approaches me and tells me what they liked, what they didn't like, who comments on the various services on overcast or on Apple or on Spotify and gives us feedback. And for all of the DMs that we get, I'm super grateful. And so thank you listeners for sticking with us through the season. And we will hopefully see you again, perhaps, even before our usual time next season. - Yeah, thank you again. (upbeat music) (upbeat music)