State of AI and Games
39:48 Share

It feels like there's a big gap between the capabilities and a durable product. Even in a three-minute experience that you're having with somebody, there is a thing that excellence feels like. And we can't confuse short with great. What the industry could really use is a handful of people ambitious enough that they're trying to rewrite the rules of the games industry versus playing by the current rules.

If you shake the trees, a new innovation comes out. And that makes everybody think about where they might move next.

If my voice is a little gravelly, it's because it's been GDC week. So I have been talking 14 hours a day for four days now. Anything inspiring? Yeah, I mean, we could make the whole episode about games and AI. Given the processing that happened this week and the timing, if we were going to do anything, we could do a short episode on just the GDC debrief. Anything interesting, I'm trying to find the thing that I could say to be positive instead of whiny.

You know, I sent a tweet out that I don't love sending negative vibes out as it's hard to be a founder. It's hard to be out there, especially when the market for games is a little bit tough right now. Yep. But even more hilarious, I sent out some tweet that was something like, you know, why don't games take more risk and,

I'm not seeing the same kind of innovation and creativity in this world that I'd love to see, blah, blah, blah. And then I had a friend point out to me yesterday that at GDC last year, I basically sent some virtually the exact same tweet.

Yeah. So I had a very wonderful GDC dinner, AI dinner that I've been running for whatever it is, three or four years now, quite a while. And that original AI and games dinner was like four people or five people. Cause that was about everybody who

who was doing AI and games back then, most of them inside of R&D groups, inside of EA and other places. And, you know, it's now 25. I probably could have been 50 if you changed the bar a little bit, but for the people that I was trying to speak to, that's probably 25, which is great. That's like the positive is that there are at least 25 people

People at GDC, never mind the ones at home that are actually just working and so on and so forth and not going to a conference. Right. That are trying interesting things in AI and games that are not just about cost reduction. Right. This isn't just how do I squeeze an extra cent out of the creative process by removing –

one person who's doing a concept art? How are these sets of people using it to create new and better things and to take more creative risk in the industry? Sometimes it's reduced cost to be able to take more creative risk, but still it is this world where the output is not just extractive. Anyway, and I'd say they're like, that is growing. It's not growing as I want it to grow. It's not growing as fast as I'd love it to happen. There is innovation happening there. And I did see

one just legitimately funny demo for an AI game that was just great and came from a creative place and was still early and broken in 15 ways and blah, blah, blah. But like, I don't think I walked away from a single AI demo in games last year with a smile on my face. And I did this year. So that's good. Funny, like funny with intention. Absolutely. Yeah.

Funny with intention. What a teaser. Yeah. And then I have another idea, which is I want to, I've been trying to think about what would be the right way for an AI studio to get started. Cause like these guys, the issue is they can't hire researchers. Not really. Yeah. And you'll probably hear me repeat this on Monday, but it strikes me that we, what the industry could really use is a handful of people ambitious enough that they're trying to rewrite the rules of the games industry versus playing by the current rules. The current rules are like,

I'm just going to make content packs for Fortnite, or I'm going to take eight to 10 years to ship GTA. And that works, but that works for like five companies, right? Yeah. And, or I'm just completely beholden to Steam, or I'm completely beholden to the App Store. And the completely beholden to the App Store is kind of like asymptoting to...

match three casino game. Like it's, it's somehow found some like, you know, survival of the fittest thing, which is like a boiled essence of junk is exactly what, what the app store is turning into for games because it's the only thing that can pay the cat, right? Like it, it has to pay the cat to get a customer. It's a casino. It just needs to hit this thing so hard. So I give you the $5 so I can pay the cat can get another user. Yeah.

that without Apple really doing something, that's where it's headed. That's so sad. I was just going to say. It's really sad. It's so sad. Yeah.

And that's got to be a failure of the platform owner, like because that can't be healthy long term. It's not. Yeah, it's not healthy long term. It reminds me a little bit of what Facebook's kind of treatment of games on the games platform. I also just think they treated, you know, like they just didn't treat them as good citizens. They didn't understand that.

They didn't understand those users, and so they never really understood games. They didn't understand how to get the best out of them. I will give Apple credit that they are trying to do this subscription model for games with the idea that it will give up harbor for premium games and interesting scripted content games inside of there. But speaking with devs, obviously there's not enough money going into it.

Apple Arcade is just rehash games that existed before. Please take Fruit Ninja and build Super Fruit Ninja, which is fine, but compare that to their TV strategy where they really believe and you see the difference. And then Steam is this weird situation. I don't know. Do you know much about Steam? You don't play many games, right? No. Steam is the PC app store for games in a business case where –

Everybody likes Valve and everybody thinks – because they're indie and they're non-corporate and they never raise venture and they're not a public company and it's a bunch of kooky dudes up in Seattle. They give off good vibes. But realistically, they don't allow any real ads on the platform. So you can't really acquire users. No ads has bad secondary effects. So there are these really –

crazy dark patterns happening now. One example, they prioritize the promotion of your game based on how many things are added to the wishlist. Okay. Like you can add it, you can go to a game and be like, oh, this is an interesting game that's going to come out in three months. I'm going to add you to the wishlist. Okay. Instead of a world on Apple where you're paying for installs,

there are now game companies paying for you to just add to the wishlist. Just to be a wishlist. Yeah. It's this kind of like a communist world where you're like, if you just don't accept the ideas of capitalism and market dynamics in a society,

capitalism will find its way into your world in a much worse way than a design system. Yeah. Water is going to find every little crease and crack. Yeah. Right. And then work gets there. Exactly. Happens to be wishlist for

And so there's just like completely, it's just so weirdly shaped. Valve still takes 30% for nothing and restricts most innovation. And then you also still don't own your customer and so on and so forth. So the whole ecosystem is kind of broken and that obviously affects the games that get released. Netflix and Discord have shots to change this, but that's a longer conversation.

Anyway, I ranted for a really long time. Thanks for listening to me for four minutes. That's why we want to do this. I'm still trying to process what the startup opportunities are in games, you know?

Like, what is the venture role in games in 2025, 2026? That's a question that I don't have an answer for, but I'm going to give you a version of an answer that has been on my mind. And that is, I think it's hard to ignore what's going on with the vibe coding games over the past couple of weeks.

And I know that we've talked about this before, and I know that you like rightfully pointed out that it's like as much of the game making is the entertainment for that levels guy. That is flight sim is, is what it is.

I feel like it's more than that. There's this joyous sense of community that even the early internet had where you're all kind of like piling in and you're playing this stupid game and then you got hacked, but the hack itself was kind of funny. And there's a lot of joy going on. And I was thinking about my interest, stick with me on this, on my curiosity and like,

intellectual interest in web sim. And I remember making games with my kids and the joy of making disposable games that we played. And like my daughter goofed on her, her brother by making a game where like he was the villain and we played it for an hour and we've never played with it again. The levels flight sim thing. And then what we did on web sim are like two

two pieces of the same puzzle. Yes. And there's going to be something that feels like a games platform, but isn't a games platform. And it's easily written off because it's going to look like lame flight sim with horrible graphics. That's not really playable. And you play it for 10 minutes. Yes. And then you don't play it again. Yep. And that's it. Painful.

patience and discipline and first principles thinking pay off at some point for us because we will see the thing that we've been circling around with curiosity for some time. Yeah. And it will be like presented into the world in a way that actually makes sense and resonates and the world's ready for it. That's just that was on my mind last night. Yep.

It is unsurprising that the last couple of major breakthroughs, you know, are things like Roblox and Minecraft, which are now old, to be clear. They're not new. Really old. Or they are things like Monopoly Go. They are on either end of the complete spectrum. They are these like...

Crazy, weird, almost feels uncommercial at first. Right. Playgrounds that don't even feel like they fit in the games business, to be honest. They're not part and parcel of the games business, to be real. And then the other end is just like late stage capitalism. Like hyper optimized. Like every little edge is perfect.

polished, you know, human experience. And not to say that people didn't get fun out of Monopoly Go. And I met with a couple of friends that I know that work on the team and saw the rest of that stuff. Like it's a thing. It takes a lot of skill to make that right. I am not saying it's sloppy. I'm actually saying it's, you know, it's like playing pro basketball. Like it is a very, very high level sport to make a game like that and make it work. The only other thing that has worked and it kind of feeds directly into this is like

couple of folks making something that's almost non-commercial and it hits. Ballatro was the biggest talk of the last year in games, which basically poker meets a deck building roguelike game. You're just trying to make hands of poker. You could play it and you'd love it. And it was made by like a dude, you know? And like, is that going to be a $10 billion company? It is not going to be a $10 billion company, but it's like the thing that everybody got joy out of playing and it felt fun and so on and so forth.

Should that thing have been vibe coded in 2026? Like, of course. Right. Like, of course. Right. Is there a world where you have 10,000 more of those? Yes. Is Roblox too hard to build for? Yes. I agree. It sits over in the land of play more than it sits over in the land of like, let's do a PE roll up of game studios and help to give them more creative direction to build a more, even to build more indie hits or something like that. Another thing.

VCs like to back game veterans, but maybe vets are not the way through to more risk. There are worlds where the thing is so new and is so outside of the norm, new market creation, that it is better to have a new person to that market than it is the vet. Like the vet's so burdened with what has to be that they can't get there. Right? Yeah.

And there's a world where you walk yourself into, you know, what is the web sim of games or what are these platforms? And you'd say it needs to be that. It needs to be an outsider coming into the system. But I'm not sure if that's true. I think at the end of the day, the reason that would be a problem is that a consumer still comes in with expectations.

The truth is, Fraser, if you popped open the next vibe crafting game creator, you would not create a new genre of games in the first 30 seconds. No. You would do what people are doing on Twitter. You would make a flight sim. That's right. You would make a version of...

of the thing. The cheap version. That's right. That was like your childhood, like love. Like you'd be like, I want it to be. I'd make like a Double Dragon 2. There you go. You'd be like, I want Double Dragon 2, but I want it to be like something closer to an IP that I like now, but not actually that. And then I also want it to be, you know, mashed up with some other thing that I like. And the problem is that each of these genres is,

comes with an awful lot of know-how to make it fun. And look, having shipped a game, it's an incredibly painful process. And the last 20% takes 80%. And just even understanding what infrastructure you would want on the insides feels like something...

That is not an amateur sport. So it leads me to think that the right founding team is going to be somebody who has actually shipped games before, who has made, you know, if they're going to make a thing that's so good at making RPGs, then they need to have not just played RPGs. Like they kind of need to understand how RPGs are built at a fundamental level.

Because what you want here is it's not, it shouldn't be clawed code writing every single line of code, obviously. There's like base sets of libraries. How do you do multiplayer server work, you know, with latency that works? How can I manage the stat system for combat? Like there are base libraries that you can, you know, kind of like function calling in LLMs. I don't imagine this being clawed code with some graphics. It is more like,

you know, function calling. It is more like MCPs. It is like a set of libraries that are actually kind of probably relatively fixed because we're talking about genres. And then it just understands like, oh, you said RPG. RPG probably means I need to call these nine libraries and pull them in and then we'll work from there. Or, oh, you said first person shooter. Well, then clearly I need like these sets of libraries. Let's pull those in and let's get going. It feels like it's much more like that. And if you've never shipped a first person shooter before,

I think you're starting from very far behind on getting something that actually feels more than just loose prototype-y, you know, that anyone would play for more than two minutes. But maybe I'm being old. I was just about to find a... You could be straight with me, dude. A way to ask if you're thinking about it from the perspective of somebody who's been in the industry for a long time. Like, undoubtedly, everything that you just said is true if you're trying to build a great game. Mm-hmm.

Right. But what happens if you slide the slider even further into the realm of play and creativity? What about if you're just trying to build a toy? Yeah. And like the games that you play are like, what if they're like five minutes and then you move on to the next one? And maybe the most you ever do is 10 before you then go and try and build and fork and make a derivative of something else that somebody else plays for five minutes. And then where I went on that is like, what does that end up with? Because...

That feels fairly shallow and unfulfilling over time. That's my problem. Like for all our conversation about TikTok and it's only, you know, I'm watching a video for two minutes and blah, blah, blah. And like, this is not a movie. There's an unbelievable level of good execution on TikTok. Yeah. Like phenomenal, good execution on TikTok.

And my worry is if it's really, really only shallow, you don't get to excellence. And that's what I'm saying. Even in a three-minute experience that you're having with somebody, there is a thing that excellence feels like. And

We can't confuse short with great. Way too many of the early experiments with like, "Hey, game makers." Way pre-LLMs, there's been all kinds of like, "Whizzywig make a game stuff," and like, "Little game makers on mobile," and things like that. The real question is like, can you build anything compelling with it that you would actually want to play? Right. That 10 people will want to play twice? I don't think you have to play it as long as you play Grand Theft Auto online.

But you certainly have to get a sense of joy from playing it that feels interesting and unique and it's not just skinning. The other thread I would pull on that would be possible is not that the playing is the best part, but that the making is the best part. Yeah. Two threads I would pull is that you could pull a thread on the playing is the best part, in which case you need to have strong sets of libraries and it needs to be able to like

spit out an FPS fairly quickly. Now, there's actually lots of libraries online. So there's a way to do this without it being like a five-year endeavor, to be honest. I suspect that's what consumers want because when consumers say they want to make Double Dragon, they actually really mean Double Dragon. If I then give them really, really crappy ASDF, they're unhappy. It's like they're not satisfied. They will not come back to that platform. And I think we have lots of examples of that, quite frankly.

That's the consumer side. The second version would be that the coding is the fun itself. Like that's actually the point. And it's the journey of the coding is the fun. I said a thing like maybe, I don't remember, two months ago or three months ago, which is like the consumer AI startup that I yearn for is not like a vertical or a market. It's that feeling of

you know, nine minutes before bed, you're super tired, your brain is fried and you would instinctively reach for something that like is like in flow. It just like doesn't challenge you too much. It's just joyful and you can just use it. It's Netflix, it's a book, it's Instagram. For many, it's like one last Fortnite run or one last League of Legends run before I go to bed, blah, blah, blah. Like that should be vibe coding.

And that's revide coding a game. And the issue is it cannot test you too much. It needs to actually be in flow and fun the whole time because it's the thing you're doing 90 minutes before bed. Your brain is fried. There's a world where you're optimizing like everything about the gaming platform is for that feeling. My kids have discovered on their own Canva and...

I thought that they would use it as a Photoshop replacement. Like, I'm going to make an invite for my party to send to my friends. They use it for what you just said. They like go down to the computer. It's messing around. And they just are. It's crafts. It's creating. And they're like. This is amazing. This is their scrapbooking. And they're. Yeah. Yeah. And they're like, Dad, you got to see this. I've made a movie. I'm like, you made a movie? Yeah.

The movie is ephemeral, but they've spent like, you know, 30 minutes putting transitions and slides and music, and it is clearly scrapbooking for them. It is like you in mid journey and it is a delightful outlet for them right now. If I really internalize it, like the joy that I had

Building these, I hesitate to even call them games, on WebSim with them was, again, the creation process. Like, no, you got your brother's face moving around and you're the tank that blows them up. And the playability of it was like the small slice of what. So the thing that's missing from that then in everything I've seen, like if you're going to optimize for that,

Then it needs to feel closer to Figma and it needs to feel multiplayer. And I don't mean the output. I mean the input. It needs to be like me and you are going to hop into a code base with an AI. And it's like the three of us, me, you and the AI are messing around.

The multiplayer game is the making of the game. Yep, that's cool. Is another way of wording it. I can get my head around that. And if I go back to like... That doesn't exist. I haven't come across anybody who's thinking that way. I have come across people that are trying to do JS, who are trying to do vibe coding either at the JavaScript level or at the deeper level, forking some open source code or something like that and trying to make...

Right. Deeper level code. Right. I don't think any of them are treating the authoring itself like a game, at least not in a way that's manifesting itself yet. Here's one that I've been been working through and I gave you a version of it in our meeting with the team earlier this week is we are like on the exponential of just coming across like raw capabilities that we've never had before.

Like the first time that I use Dolly and you type in text and outcomes, it's everything over the past five years. Like, right. It's like you do this and they're like, oh, okay.

Man, that's magical. That is that's something I've I've never I've never been able to make music. I've never been able to make a game or software or anything else like that. You start with the capability. Yeah. And and there's now a growing number of people that's non-trivial who just want to try the capability. They want to experience the magic. And if you look at that, it can be very exciting. Yeah.

Because you now have millions of users or millions of dollars of revenue if people are then paying to try it. But it feels like there's a big gap between the capabilities and a durable product. And I know that like that's simplistic, but it feels like there's like the start of a way to think about what we're seeing in the world. Yeah. Like if you bring a capability to market early, when are you in a good position to

also become the group who turns that into a durable product versus like you're just the first person who brought the capability into the market. And there's like a little bit of heat, but it's more a reflection of the fact that you're giving people a taste of the future. Getting a million users to have your product is necessary, but not sufficient for success today. There are just a million users that seem to be running around

trying to try AI products on any given week. And you may catch the zeitgeist one week and have the million come through. The issue is the million don't necessarily stay because they have another AI product to try next week. And that's certainly a behavior that we're seeing across startups right now. And I guess what you're trying to say is,

What are the indications? If you were a founder and you have the million flow through, you're trying to figure it out. Like, hey, how do I now turn this lightning in a bottle into something that lasts? And how can I be the one who's introducing a capability that wows you, but I still want to be the guy in a year or two years? Like, I want to turn this into not just...

I don't know, a funding round or a couple of hires this week or whatever it is. But I want to be the lasting change. I think this is why you're getting so many conversations right now amongst the builder class about pace and about speed.

Because I think the truth is that you don't have a durable edge right now. And so the only thing you can do is stack wins on wins on wins. Right. And at some point, it becomes durable. At some point, it becomes habit forming. At some point, we get out of this window that we talked about before. We're like, I'm no longer comparing toothpastes. I'm just going to use the thing I wake up to use every day. And then you don't change toothpaste for 10 years or ever. Ever. Yeah.

I use the deodorant my dad used and my son uses the same one.

That's right. Like it was just, why? Because there was no reason to try another one. It was like, I don't know, it seems fine. I didn't need to relitigate my laundry discharge and it was tied. And then, you know, we're a tied family somehow. That's a weird thing. These things will metastasize. But the awesome thing about technology and that reason we like this stuff is that there's a little, you know, you shake the trees, a new innovation comes out and that makes everybody think about where they might move next. Right.

But in the meantime, it's just stacking wins, and those wins cannot all be revolutionary. But some of them can just be about pulling your attention back two weeks, three weeks later. I had this conversation with a founder last week. The surface area of your product changing a little bit, adding another value to somebody in a way that a user can see, not just a random bug fix or infrastructure improvement, but something that a user can see.

Right.

And so the thing that's happening when you have high pace is the first thing, which is the obvious first order effect is like, Ooh, shiny penny, new thing to try out, new thing to promote. Great. Right. Sure. Also maybe something that somebody else has to play catch up with against you. So you're saying ahead. Awesome. That's the first order effect. The second order effect is that I think,

get tired quickly. They just want the job to be done. But more importantly, one of the jobs that they want to be done in this age is, is this a company that will keep me up to date with all the changes that are happening? And so by having Pace and by stacking three, four, five wins, you also start to represent in like a brand for your users that is, hey, this is one of the innovators. Like,

So if I'm using this photography software or I'm using this coding software or I'm using this – I feel this way about Windsurf and Cursor right now. They're pacing so quick and close to each other that I feel comfortable using either. Right.

Because they both represent a brand for me of innovator. Like I know if some new thing comes out and it's awesome in AI coding, maybe Cursor got there a week ahead of time. Maybe Windsurf got there a week ahead of time. But I know the other guy is going to cover it a week later. So they both have this like represented innovator brand in mind. And so I don't have to relitigate which one I'm using every week. I pick one. Never mind which one I use, but like I pick one.

And like, I feel comfortable. And so it means I don't go through the process constantly. And that's what you want as a brand. You want the consumer to feel like, yeah, don't worry. We're like, we're your AI steward. We're going to carry you through the next 15 model changes and crazy things that are happening. And we'll be like, just come to us. We'll cover it. Yep, yep, yep. So then from our perspective, trying to make sense of the opportunities, I really like the concept of stacking wins rather than pace. And... Why? Yeah.

Well, win has connotations of like delivering on the vector that matters. Whereas like pace just means like I'm moving very fast and doing a lot of stuff. You don't like the empty calories kind of feeling the hamster on the wheel. Like, of course not. But like, I think like if you're first to deliver a new capability or to like help people experience a new capability, stacking wins means delivering what users want from that moment on.

And like the win, I think in that case is like connotation for like delivering what users value and what they want and what they care about. Yeah. But like you can imagine a place where somebody delivers the new capability and they move exceptionally fast down the capability curve, but that has diminishing value for what people actually want in the product. If that makes sense. Yep.

Right. And so like my like very tenuous thought here is that like,

A lot of times we get excited about a new capability because it feels like magic. And often, more often than not, that new capability is likely going to make its way into a product experience that like solves jobs to be done for us, but usually requires like good product work to get there. Right. Like stacking of the wins is the good product work to get there, if you will, rather than just like banging on the capability and making it ever more capable.

as like a simplistic example. Yeah. In other words, like the execution, you do need to keep thinking orthogonally and delivering new value to a user. It's not just that the thing got 0.03% faster this week. You still want a little surprise and delight. I think so. The reason that I've been like wrestling with this one is to try to sharpen my thought for when we do see something like this

What is it where we would be more inclined to lean in? Even if we just went through with that, you either do it the day before or three months later. And I just come back to what you said around stacking wins in the meeting and that the direction of the win that they're stacking is the thing that matters. The moment that there's an inflection point and something feels like it has product market fit, but you're seven days in, it's probably the worst time to invest. You should have done it right before it launched or you should have done it

Three months from now, six months from now, when you have evidence of retention and engagement and importantly, that they can continue to have pace and stack wins. Right? So if you play that back, what are the few times when that moment of inflection and the hype cycle is high and so the price is at its highest, what are the times when you think,

screw it. We should be involved now anyway. And it's when you feel like you have an insight to the other bits we just talked about, right? It's when you can reach a confidence threshold that is like, no, I actually think they're going to stack wins. I haven't seen it yet, but we have a conversation about, this is why we have a conversation about products, why we're product oriented. Because I think the kind of conversation about being founder oriented is all wonderful, but founder oriented only means one of two things for me. It either means you have a fancy resume and

And so, and there are VC firms that act this way. It's like, I don't know that person was a senior executive at Microsoft or Google or OpenAI, whatever their thing is. And so like, I'm really investing in resume because you don't know these people. Yep.

And so it's either one, you're investing in resume. That's what founder is. Two is you maybe do know them for a really long time and you do have a really strong intuition, but that's another way of saying you think they're going to make good product decisions and good strategy decisions, which does happen. And I respect those people. Or the third is...

you know, the good founder thing is PT Barnum. Like, I don't know, they make a good pitch. Like I was swept away by their reality distortion. Well, it turns out that like lots of PT Barnum people can't execute with crap. They just have a good pitch and that's all they have. It's interesting is Brendan at Sesame had all three, at least from my experience, right? Like

We have history where he has made good product decisions. So we have confidence there. It looks like he has made good product, like thoughtful decisions and like projects into the future. Good, thoughtful product decisions. And there's like a little bit of PT to Barnum. I am not saying you don't have to be a wonderful pitcher to be a good CEO. Like, look, there is marketing to all of this. And yes, Brendan, yes.

For us, Brandon at Sesame is, we've known him for a long time. We can clearly see that he executes. He's tapped into something huge. But also, there's also more than one win there internally. Yeah, yeah. Like that business, it might not be launched, but there's more than one win already happening there. And you can feel pace. Yeah. Yeah.

So that's a good qualifier. But on the far edge of risk where you're like not really sure what you're trying to get there, I like our way of processing things, which is just figure out whether you think they're going to stack wins into the future. It does make it harder.

we have a harder time doing the, "Hey, this is an amazing thing to play with right now, but the founders have absolutely no idea what's going to happen in the future and there's no durable moat yet." And like hard to reach confidence in those people. We just met the person two weeks ago. Right. They seem great. Demo's kind of cool.

No idea how they're going to navigate into the future because they have no idea what they're going to build next. I think the only ones we can get really comfortable there are areas where there's a lot less competition. So actually, ironically, we could say on the same subject, when Brendan and the team came in with Oculus, like, that's a good example. We don't quite know what it's going to turn into. We don't know where the wins are going to be a year from now. But the demo is amazing. And you're like, nobody else is building this. Like, like, like...

Like, come on. Like, VR has been a dead market for 20 years. Like, come on. Right.

So that's an area where maybe part of the heuristic we're having in AI right now is that if somebody releases, is taking advantage of some new model capability in AI, the thing that you know is that there's going to be five YC companies in three months tackling the same thing. It's going to be 500 people doing this. And then you're going to see two guys spin out from Google that are trying to raise $100 million doing the same thing. It is the nature of this kind of feeding frenzy thing.

that creates a different decision matrix for both the company and investors, right? Right. That's a great call out. Is that like, what's the nature of the competitive market? Yep. Is it obvious and everybody's going to do it in three months is very different than like, oh, nobody in their right mind would ever do this. Yeah.

And I will say, to loop back to it, the craziest thing about the games business right now is that, you know, we teased about one thing that I think we both had not around, which is vibe coding in games gets you outside of the normal game cycle. It will be fun. It will be a thing. It should happen. And there's a really sizable company to build there. Awesome. But I will say...

There's nobody really doing much of the riding the model capabilities inside of games right now. Like if I'm trying to ride the edge of reasoning and model capabilities and I'm trying to make a B2B law startup, like it's ants on cheese. Like it is just a complete feeding frenzy, you know, but yeah.

But if I did literally the exact same thing in games, I'd have to figure out the form factor. But I don't think there's anybody doing that. There's no framework for a play space where I'm the author. So forget the kind of people making their own games. I'm the author of that game experience. And I am just letting you play my game, but in a mod framework with whatever new awesome model capability that happens. And I'm glossing over the nature of building games and finding the fun and

A little bit of the things there on purpose. But I will just say, if you do this in games, like, it's pretty green-filled, man. Like, I just came out of GDC. I just had, like, 35,000 meetings. Like, there are not that many people doing this. I may regret this question. Why not? Like, what's your thinking on why this isn't happening? So I think the best way to describe this is to describe GDC properly.

So GDC is a five-day event. It's huge. I don't know how many people come there, 50,000. Monday and Tuesday historically have been the summit days. The summits are two things that the first thing they do is they have a sound summit, an audio summit. They have a technical artist summit. So basically they have small niche things for specific parts of the gaming world. And then the other thing they do is whatever the new cool thing is, they run a summit for it. So

There was a virtual world summit when virtual worlds were big, right? There was a VR summit when the VR summit was big. Like these kinds of things were a real thing in the day. Not surprisingly, there was a game AI summit this year and there was a machine learning summit this year. 10% of people go to the summits on Monday and Tuesday.

And then Wednesday, GDC kicks off and like 90% of people show up. I see. And it's all console and PC. It's all PlayStation and Xbox and the next GTA and Call of Duty and all the rest of that stuff. And so that should tell you a lot about what the games industry is like and what it feels like, which is most people want to make cooler looking guns.

And they love games. They grew up playing games. They play Call of Duty all night long and they wake up in the morning and they want to make the next Call of Duty or something close to it or World of Warcraft or whatever it is that they grew up with, which is to say that it's super crazy, interesting and satisfying to build something that's only 2% or 8% different. Right. It's just fun to build games. It's too fun to build games. I see. Frasier. And so people don't,

innovate enough. They don't take enough risk on the creation side. That is so counterintuitive. It's the same thing as movies, right? You're not trying to make the weirdest movies in the world because you just love the craft of making a movie, right? I don't need to reinvent whittling wood or playing the guitar because playing the guitar is fun. It's great. I love it. I'll do it for uneconomic reasons, like supply of people wanting to build games far outstrips demand.

But AI feels like we're going to give you a new whammy bar on your guitar. Like I'm trying to make the analogy. Like it feels like it would just add to the, it should add to the experience of creation.

Yes, it does add to the creation experience, but it's hard and different. I see. And for most people, they're quite... It's like, what's the satisfaction bar of creation before you feel satisfied internally? Mm-hmm, mm-hmm. That's what happens when you're making something, when you're building. Yeah. And in games, like taking...

and just putting Bart Simpson's head on the demon is like pretty satisfying as a 14 year old. Yeah, yeah. I get it. And that's just a mod, right? And so I think for a lot of people, they're happy to build genre games. That's the way of putting it, right? It's very satisfying to build a genre game. They get in the flow. They have the tools. It's familiar. They're...

They know the audience is going to be there already. It's yeah. Yeah. All right. I had a, I had a founder yesterday give me a very common piece of advice that he said it was really hard to learn, which is the plus one model of gaming, which is you do a game that is exactly like another game that you like. You had one small thing.

I have another founder friend that I've known for 20 years. I will not reveal the way he talks about the way he wants to do games. He's released a lot of really successful games over time. And like, he has the like one thing, the same one thing, mildly different and one thing wildly different. And like that puts him in the far edge of innovators. I see. Interesting. Interesting. The game world is like,

You would assume it would be very similar to like software engineering and like software products, but it's very different. Yeah, it's very different. Anyway, let's be done now. Is there anything else we should you want to talk about with me that's off topic or whatever things you're processing fun for the week? I'm in town on Thursday and the NCAA tournament is in town on Thursday. Do you want to go to an NCAA game? Let's just say yes. Well, let me know and then I'll chase down tickets of a sort. Let's let's check back and forth and figure out what we want to be. Yep.