Subscribe
Share
Share
Embed
The Future of Gaming DAO or FOGDAO is a decentralized, tokenized community exploring the future of the gaming industry.
Nico Vereecke:
GM friends and welcome to the Future of Gaming podcast. Um, my name is Nico, as most of you will know, investor at BitCraft, we've got with us Sebastian Park, co-founder at Infinite Cavus and venture partner also with me at BitCraft. And then we have Devin Becker who is gaming lead at NAMI. Um, what we're going to be talking about will be found in the title and in the description because I tend to say this a lot, but we don't really know where we end up, but Um, yeah, let's dive in. I think a few weeks ago, I think two weeks ago, um, Sebastian company and Sebastian's company, infinite canvas announced a raise, um, saying that they would focus on AI powered or building an AI, AI powered discord game. And so I think it might be a good time for you, Seb, to, um, to talk a bit more about that and then
Sebastian Park:
Yeah,
Nico Vereecke:
we can
Sebastian Park:
I
Nico Vereecke:
dig
Sebastian Park:
mean...
Nico Vereecke:
into, you know, why you're excited about AI and what you think it can do for, uh, for games.
Sebastian Park:
Yeah, for sure. I will be clear, we built a bunch of these games. I think that's a very important distinction that's worth mentioning. I really coming from eSports dislike announcements of announcements. They, they, they earth me. And so, yeah, Infinite Canvas was fortunate enough to announce this race a few weeks ago, where we raised $6 million to focus on AI powered games. Just to be clear, one thing that's really worth digging into here. before we talk more about AI and the awesome things about how we change it up, is that there's been a lot of focus in the zeitgeist for press you're reading about taking the technology forward, right? The idea of, Hey, like, can we discover room temperature superconductors is like top of mind for people. There's also a bunch of stuff happening around specifically, is it possible for us to generate 3d assets? Like Nico, you and I have talked a bunch about 3d asset generation. what Infinite Canvas works on is we actually are not a research company. We do zero research. We're a game design and game development company. We're a studio publisher. And so our tact has been just to ignore a lot of what could be possible in the future. In exchange for trying to figure out what is possible today and how that informs game design and how that informs the types of games that are possible. It's been a lot of fun. We announced two games. The first one I think that got more to press is Creature Craft. Creature Craft is a game that's built natively inside of Discord. We can talk a little more about the social algorithmic reasons as to why we wanted to launch in Discord. But effectively, the game loop in and of itself, be it the RPG loops or the idle RPG loops or some of the collection loops, are all powered by off-the-shelf AI tooling. So we don't have anything more custom than tuning our own Lambda. We don't have anything more custom than switching between GPT-4 versus using Dali or Vertex for image creation. It really is playing with the idea of making every player a creator and how can we use existing tools to do so. So that's what Infinite Gams has been working on for the past, I'll say like nine months or so. And so it's been a lot of fun to finally announce it, especially because the actual game that people are seeing right now is, you know, took about two to three weeks to get to production.
Nico Vereecke:
So before we dive in, Seb, I, um, it's really interesting what you said that infinite canvas essentially uses the tools that are built by others is, doesn't see themselves as a technology company. Um, are you aware of any, uh, data around technology adoption curves and the point at which startups start adopting them, like where the world is on the curve and like success and, and where. most value gets captured or at what point most value gets captured. You have any
Sebastian Park:
Yeah,
Nico Vereecke:
idea?
Sebastian Park:
yeah, yeah. So a few things to unpack there. Number one, we are definitely a technology company. We're just not a research company. Like I want to be clear about that one. Like there's a difference between research and capture because there's that famous concept of crossing the chasm where you have your early adopters. You have this like very large chasm before you have mainstream adaptation or adoption by the large groups of folks. Typically speaking, When you have new technology introduced, there are two chords that can be applied to this technology. Either it's A, a very specialized technology, at which point capturing the actual underlying research is going to generate the vast majority of the value. Or B, it's such a transformative, disruptive technology that being a technology that's almost middleware, that's almost playing with the tools itself captures a ton of the value. So let me give you an example of that. On the, on the like, Hey, this technology is super boring. You got to just capture all of it in order for it to work. A good example of that in the modern technological sort of zeitgeist so far is VR, where a lot of the macro value in VR is still being captured. Necessarily by the folks who are either making the headsets or doing very specific things there. VR hasn't pervaded the zeitgeist or been cross the chasm to the point. where there are people who are the specific VR subsegments that are making a large amount of value. The jury is still out for blockchain. I think about 18 months ago, people were pretty convinced that it was going to be the latter category. Now there's a good chance that there's a bunch of companies that are in the former category, which isn't to say that blockchain is dead. It just means that it may require more generalized capture in order for it to accrue value. quintessential example of a technology that's so disruptive and so large that basically everything that's built on top of it is interesting is the internet. The internet is a great example of a technology that's so large. An example of this would be Amazon in its initial purview was simply the internet applied to books and bookstores. It was a wrapper company. It just so happened that the wrapper was internet and the internet was pretty amazing and large. And so... depending on where the technology lies, it determines whether or not the captures, the value is captured by the middleware part, by the platform part, or by some combination of the two. Clearly, I think AI, and especially a lot of the stuff we're seeing in generative AI, is on the latter versus the former.
Devin Becker:
Sounds to me like a lot of that's on whether or not there's enough people using it on the user side, right? Like when you're talking about VR, talking about blockchain, a lot of it is on like, is there enough people here using this? Like the internet example with Amazon, where it's like Amazon really became successful when there was a lot of people actually finally onboarded onto the internet. Like, cause there was that whole period where everyone was like, I don't know about this internet thing, like could this internet thing take off? And like Amazon wasn't really that successful at that point, right? It was one, like everyone was there, like that sort of post AOL period. So I, like, I would say, I think, At least from my perception, that seems like a big part of that. Uh, like with the VR side, like I've been using VR for a very long time, but I'm also like in a minority and that's why it's like not take, take it off. Right. Like, but it also does sound like sometimes the middleware is like a big part of that in terms of like the smoothness of the user experience. So like, uh, for example, like browsers, even on the internet side, right? Like they kind of sucked were clunky. Like once those kind of got up to speed or I guess once Microsoft kind of stole the whole market for awhile, uh, or You know, built it into the, the OS or those kinds of things in the blockchain. I think was a good example of that, where it's like, you know, it, just the accessibility of it, but I, but going back to the AI one, like chat GPT, like that's when, like the AI was there, like GPT three was pretty good and it was there, but no one could really get to it. Like there wasn't consumers there. Like no one at the watering hole kind of thing. Like, and then once chat GPT came out, like that was the moment where it's bomb saving like Uber or any of those other ones, right? Where it was like accessibility. to like the masses and then the masses coming there, like that kind of set it off. That's my perception though, like on that stuff.
Sebastian Park:
It's really fun. A lot of folks think that, especially when you look at history and look backwards, that things are fairly binary, that things just happen. And it's really funny as the older you get, the more you've seen the transition yourself and you realize it's more of a spectrum than a binary constraint.
Devin Becker:
Yeah, half of it. They didn't see the prequel that happened. They just see the hit.
Sebastian Park:
Exactly. People think about Amazon as this monolithic Goliath of a business that everyone uses always when, at least in the three of our lifetimes, we remember the internet was a bit sketchy to buy things on the internet. There were
Devin Becker:
Yeah
Sebastian Park:
risks and concerns about sending money over these digital lines, these
Devin Becker:
When
Sebastian Park:
digital
Devin Becker:
people
Sebastian Park:
virtual
Devin Becker:
forget how much
Sebastian Park:
tubes.
Devin Becker:
PayPal changed that, right? Due to like eBay and PayPal and how much like that had an impact on that sort of
Sebastian Park:
Mm-hmm.
Devin Becker:
thing. Like
Sebastian Park:
Yeah.
Devin Becker:
it's, yeah, it's crazy.
Sebastian Park:
Exact same thing here. I will preface or not preface. I'll caveat this by saying we're talking only about consumer technology. And so by definition, consumer
Devin Becker:
True.
Sebastian Park:
does have that issue where it's like, yeah, it's purely number of users. It's I'm sure different for folks who are in the more SaaS or B2B space. Like this is one of the funnier things that you see in self-driving cars for those of you who are following where. There's the consumer use of self-driving cars and the requirements that have. And then there's also like trucking self-driving cars, like self-driving trucks and like the shipping channels of trucks. That has much lower bar because it's just such a hard thing for people to do anyway and it's not interacting with consumers very much. It's not interacting with, like you will not see a giant big rig truck, usually on the streets of San Francisco. That's just not happening. unless it's like the last mile delivery at which we can have a human do that. A lot of trucking happens over the course of thousands of kilometers or miles. And so that happens all the time in this world. But I think the most important thing, especially with the GPT style stuff, especially chat GPT, is that it really sparked the imagination of the public to believing that a lot of the AI stuff could do everything. And the answer is it's cat. It's still in the research phase for a lot of it. but it can do a lot more than people knew and a lot more than people think. And so now it informs the product design side of either of consumer products, but in our case of games as well. And like what kind of games are created as well as published by a company that works on the intersection of user generated gaming.
Nico Vereecke:
So let's take a step back then and walk me through your process when you look at technologies like chat, GPT, um, when you look at technologies like mid journey, um, from a first principles point of view, you know, you've been building games and working in games for, for a long time now. Um, how, how do you think about that and how did your thought process evolve? And, and, you know, w in what direction are you building right now?
Sebastian Park:
Yeah, so I think the thing that gives me a lot of color and context is that I was fortunate enough to start my career right when Facebook switched over to mobile and we were trying to figure out what games made sense back then. A story I tell often is that Sword and Sorcery is an awesome game and it's a great game that's just really not maximally using mobile as a platform because no one had done it before. Games that really started to use mobile as a platform really well were games like Rami's Ridiculous Fishing, games like Threes or 2048, the clone that took off and doesn't completely another story. Games like Dots and Two Dots. These are games that really leveraged mobile well because it used the form factor, it used the fun, it used the fact that tapping and placing things is important. Like use haptic feedback. These are things that are part of the device for a number of years, but weren't used. And so from a first principle perspective, it is always important to use the first principle perspective. It's, hey, what game loops don't work? What game loops do work? I'll give you a couple examples of things that happened early in Infinite Canvas' iteration. Because we launched CreatureCraft, but CreatureCraft's actually, I don't even know, like dozens, maybe almost game a hundred of our iterative loops before we decide a game that was decent enough and even then, you know, up and down as to whether to launch. Good example number one, we were a huge fan of the idea of a lot of these gotcha games. There's Han Kai Starrail, there's Genshin Impact. Can you do a gotcha style game with Gen.AI? Our conclusion strangely is that no, you cannot. Whereas places like blockchain use technology that by definition encourage a scarcity. A lot of the underlying tooling of Gen. AI encourages creativity without scarcity as a limiter. That is actually the value prop. And if that's a value prop, you cannot have a gotcha style game or a game where the core loop is based on scarcity or exclusivity. It's just a different style of game. So that's a really straightforward and easy game loop where we thought we would do something like that and then we're just wrong. It's not possible to take. So you start with things you know and then you... Say no to them.
Nico Vereecke:
By the way, before you continue, because I really love, you know, the way you're approaching this, I think you said that blockchain is built around scarcity as a, as by design or like fundamentally. And I actually think that it's like, that's the way it has evolved until now. Uh, but I don't think it's, it's necessarily like blockchain doesn't equal scarce and I think we'll,
Devin Becker:
No.
Nico Vereecke:
we'll likely start seeing, um, blockchain be used by game studios in terms of like, like Mormon became physical things where you know, you buy things and then it exists
Sebastian Park:
Mm?
Nico Vereecke:
versus necessarily like having like 10,000 of
Devin Becker:
Mint on
Nico Vereecke:
one
Devin Becker:
demand
Nico Vereecke:
item.
Devin Becker:
instead, yeah.
Sebastian Park:
Yeah.
Devin Becker:
I think
Sebastian Park:
Actually
Devin Becker:
so too, but
Sebastian Park:
that's
Devin Becker:
it's funny
Sebastian Park:
a really
Devin Becker:
because
Sebastian Park:
good...
Devin Becker:
it's like, go ahead.
Sebastian Park:
Oh, go ahead David.
Devin Becker:
I was just saying that's one of the things blockchain was able to enable was that scarcity, and that's why people are doing it. Like it's just, it's a new toy to play with.
Sebastian Park:
Oh, for sure. I think that's a great example of this, which is you can enable gotcha style games with AI. You can use the tools to generate more images, to generate more
Devin Becker:
Wait,
Sebastian Park:
items.
Devin Becker:
what about Aeromancer? I'm curious, like your thoughts on that, because that
Sebastian Park:
Yeah.
Devin Becker:
is a Godgesal game that uses generative AI stuff.
Sebastian Park:
For sure. I think that the folks like Corey and the folks at Spellbrush are amazing. I think they're absolute geniuses. I just don't think that is what the future of AI gaming is going to look like. And I think that's very similar to the folks who are using, who may be doing some interesting stuff on blockchain. It's just different.
Devin Becker:
It's like what
Sebastian Park:
I
Devin Becker:
you
Sebastian Park:
just
Devin Becker:
said
Sebastian Park:
don't
Devin Becker:
about initial
Sebastian Park:
agree
Devin Becker:
mobile
Sebastian Park:
with the...
Devin Becker:
games, right? Like it's just,
Sebastian Park:
Yeah, exactly.
Devin Becker:
it's like baby steps.
Sebastian Park:
It just, you have to go through it. The flip side, by the way, in addition to taking loops that you know, and then seeing if they work or don't work. is the opposite, which is it's actually really fun to take loops that you knew worked or knew didn't work and then retest them. And I think that's been a lot of fun with AI as well. So here's a good example. There are games where the actual coding of the game is impossibly annoying, like impossibly annoying. A great example of this is Magic the Gathering. Magic the Gathering, if you think about how that system works.
Devin Becker:
I coded one of those. I worked on a digital TCG and I was about to say exactly that. That is a spaghetti nest of madness because like timing rules alone, like that people have to resolve in person, are madness on a computer. I totally agree 100% from the actual experience coding that. Like imagining the coders who had to do magic the gathering, like that's just thinking about it's masochism.
Sebastian Park:
Yeah, it's crazy because it's exponentially annoying. Let's say you have a 200 card trading card game. If you have a 200 card trading card game and you have 40,000 interactions, because it's 200 times 200, each card interacts with each other. If you add 100 cards, you don't add another 20,000, it becomes 90,000 interactions. It gets very exponential very quickly. What's really cool is that The heuristic around this before was, hell no, I'm not doing this. That is, unless this game is like the, unless this game's literally Magic Gathering or Hearthstone, it is not worth the engineering time to do that. With a lot of the AI tooling we have, even that's in existence today, you can actually just basically create a framework in which it writes out the remaining 50,000 interactions once you have 200 interactions already trained against. you can ride out the hundreds of thousands of your actions. So suddenly the carrying cost of growing a card game set where it used to be insanely difficult, now is totally fine. Super interesting.
Devin Becker:
Especially for QA, because that's a nightmare for QA on that. I had to really rely on the QA people for fixing engineering stuff, like because they would be the only ones really able to like explain the weird quirks that you wouldn't get from looking at the code.
Sebastian Park:
Right. And people, by the way, have no idea how impossible a task task. Like I think one of the things that people don't understand, especially with like, Hey, why is it that you launched this new patch or feature in one of these League of Legends style games and you didn't solve for every bug? And the answer is actually you as the audience playing millions of games will see far more than any QA team possibly can because it is impossible. Or not impossible. It's cost prohibitive to test all 50,000 interactions. That sounds ridiculous. Why would you do that?
Devin Becker:
I'm excited
Sebastian Park:
And.
Devin Becker:
for QA IQA like that's gonna be because like I see whenever I see like bad Ratings on like Google Play and I imagine same on iOS So many of the of the one stars and stuff like that are because of a bug or a crash And it's like if you solve that alone like the ratings on the app store would go up by like a whole number Like not even a decimal just like a whole number because of that I think generally so like
Sebastian Park:
Yeah.
Devin Becker:
five point Oh would be like the mandatory to get someone to download your game at that point
Sebastian Park:
Yeah. And that's actually a great point, Devin, because that I think is the lane where we're straddling an infinite canvas. There are a lot of people already, and it's just not talked about enough, which are using, who are using AI to improve their game loops from a QA perspective, from a tooling perspective, from an efficiency perspective. We're doing that too. That stuff's really cool. That stuff, I think is less fun. There's a bunch of people who are pushing the boundaries of AI, boundaries of 3D asset generation. That's really cool. That's a research problem. I would say that like 8 out of 10 people who are working on that problem aren't technically qualified enough to really push that boundary. We are very clear. We don't push boundaries in that way. We're not focused on creating a tooling company that solves your QA or your other efficiency needs. We just think this is a really cool space to design games in. And that for the first time in a long time, this could be a very revolutionary way that changes paradigms around game design. Because the end user, by the way, does not care, like you said, Devin, as to whether or not there are 50,000 additional interactions you need QA for. They just care that the game is good. And so suddenly the imagination of players who previously we would encourage to make games, work in the game industry for a number of years to be told what they can't do, now can just do it. which is always a really fun place to be. And you're gonna see, start seeing a lot of really strange, esoteric games.
Nico Vereecke:
Talking about strange as so theory games. Tell me more about yours.
Sebastian Park:
Yeah. So, so creature craft real straightforward loop. We've realized that gotcha doesn't work as well, but we'd like collecting and we'd like the idea of creating affinity to your items. The two things that AI does incredibly well and Gen AI in particular, that's super well is affinity from the perspective of, Hey, I like something. I have made it. It is mine now. Therefore I provided more value. It is very good about that. It, you see someone else do something interesting. It's not that you want their thing. It's in fact that you want to see your envisioned idea come to life. So that's really great. It creates a lot of lock-in. It creates a lot of people interested in your product and wanting to try it out. The second type of affinity, which was a term coined by our head of engineering, Bruce Hill, is that AI is very good at taking unrelated, seemingly unrelated topics and figuring out what relates them. Like really good at creating that interaction. So we rely heavily on that bit. If you imagine cup, pen, rock, you may think there's nothing that puts those three things together. Someone may, in AI can say, hey, there are forms of pottery that use, takes the rock, turns into ceramic and uses a pen in order to engrave it. Right. This is what we use inside of our games. And so a lot of our game modes, and you know, if you check out Creaturecraft.gg or the official Discord servers, you'll see a bunch of these games and we'll certainly be posting more examples of this. effectively is just pulling against your creature, Nico, versus Devin's creature, in an environment or a new arena or a pinata match or a race, and then pulling from the information it has from that in order to determine a fairly deterministic outcome as to what will happen when those two things clash with each other. And that's a large part of the cool affinity-based game loops that we see. Some of it comes up in RPG loops. My favorite one right now is that if you take your creature and you fight the ice golem in one of our like RPG loops, it asks the question from a prompt standpoint, how well does your creature do against ice and golems? And if the answer is poorly in both regards, it just does poorly period. If the answer is well in one regard, it's normal. If it does well in both regards, it's super effective. So it's effectively recreating the Pokemon like weakness strengths. resistances, benchmarks, without having to worry at all about us pre-programming every interaction,
Devin Becker:
It's handing
Sebastian Park:
which is a
Devin Becker:
you
Sebastian Park:
ton
Devin Becker:
an
Sebastian Park:
of
Devin Becker:
extra
Sebastian Park:
fun.
Devin Becker:
complicated version of rock paper scissors for you.
Sebastian Park:
Yeah, exactly. I think Rock Paper Scissors is a great example. We call it internally Rock Paper Everything in
Devin Becker:
Thanks for watching!
Sebastian Park:
that it's, what if you had a game where it's just Rock Paper Anything and immediately it's adjudicated for you.
Nico Vereecke:
It reminds me of the show called Big Bang Theory, where they have like a rock paper scissors that's like 10 things long and it is like this incredible web
Devin Becker:
There's
Nico Vereecke:
of
Devin Becker:
rps
Nico Vereecke:
different...
Devin Becker:
25 too which is just gets nuts
Nico Vereecke:
Yeah, really, really cool, Sebastian. And so talk to us about, or tell us more about the form, like how do people play? It's inside Discord, like what's the loop? How do you get into this game?
Sebastian Park:
Yeah, so this is something that we really believe at InfantInk Canvas. And my co-founder, Tal, is writing a piece about this, which hopefully it will be out eventually, I don't know, I'll say in the next week or two. But one of the things that's really interesting is that if you believe our thesis that gaming is becoming democratized, if you believe our thesis that you can turn players into consumers, players and consumers into creators themselves, what ends up happening is that there's a lot more of it. There's a lot more game than there is. people's attention spans for discovery of games. And so what we push for is we look at a lot of platforms that have built-in discovery and built-in algorithmic distribution. Places like Roblox and Fornite are quintessential examples historically. Now what we're looking at, which I think is a ton of fun and something that I've been thinking about a lot about, is Discord. So Discord is great for a community. Almost everyone who's listening to this is on Discord by definition, if not just by fun. And it's a bot. We... have created a bot that you can install on any server across the Discord verse and it'll live in your server. It can be your own version and your own pool of cards and characters that are interacting in your server. And you go forth and conquer. And that's how you interact with the game and interact with the bot. You install the server, you click play, it'll bug you and then go from there. And that's how we've been able to generate as much distribution as we've, we had so quickly.
Devin Becker:
smart to make it so people are basically like spreading it themselves by putting it on their own servers to entertain their own communities. Like it reminds me of like some of those bots for like even twitch channels where people would run for like different games and stuff like that.
Sebastian Park:
Yeah, our inspiration to do another callback was IRC. If anyone remembers IRC on this?
Devin Becker:
Egg Drop bots and everything, yep.
Sebastian Park:
Yeah, yeah, so
Devin Becker:
Or even
Sebastian Park:
IRC.
Devin Becker:
BBS door games, like where you drop those in. Like a, like a, BBS was coming, like an old school discord in a way.
Sebastian Park:
100% instant relay chat, really big in the 90s and 2000s. But it is always fun because you can pull from techniques and gameplay and art forms from decades ago now. That's why gaming is so fun now because it's new tech and things that we've learned decades ago.
Nico Vereecke:
So what's the distribution of players, or yeah, players inside of your Discord server and outside in their own Discord server?
Sebastian Park:
1 to 99, maybe 1 to 999. Like, I think it's like, it's like our, our data dashboards seem to be very clear that the people who are in our official discourse is a small fraction of people who are exposed to this.
Nico Vereecke:
Fascinating. What, you know, having been working in games for such a long time and seeing the metrics of new games, is there any specific takeaways, any specific data points that made your eyebrows go up where you're like, holy shit, I did not expect this to turn out the way it did?
Sebastian Park:
Yeah, I didn't expect us to run into rate limits as quickly as we did. I would say that, that I think like, you don't have to do sharding until you get to a certain scale and size, we, we ended up having to spend a few days doing sharding, the pod was a bit unstable for a couple of days, a couple of weeks in. We didn't expect that. I think the other things that we didn't expect is we didn't expect sort of the power user behavior, right? We thought, we thought this quarter as a platform that's more casual. And so. Our expected game behavior was people come out and they'll check it out, they play for a couple times a day. It'll be a second screen experience that's somewhere in this periphery of your non-main screen. It ended up being some people just love the game and they just keep playing it. And that's been surprising to us for sure.
Devin Becker:
I mean, it's like IRC or discord tends to be dominated by a certain group of users or forums or any social things tends to be like the power users, whether it be the power users is in the power talkers essentially, or that be people that are just like super into the game. There always seems to be like a core thing. So I guess that's not terribly surprising in a way. But but it's interesting that like, there's people I wonder if there's people that end up core to the game, but don't talk much in the actual discord. That would be interesting to see.
Sebastian Park:
Yeah, that would be interesting to see. We have seen Discord engagement, like server engagement, go up 20 to 30% across servers where it's installed in, which is pretty crazy. So there's a world in which this is no longer a game, just for game's sake. It may be just be a useful bot if you want to have people engage in your server more.
Nico Vereecke:
How
Devin Becker:
Is there any
Nico Vereecke:
customizable
Devin Becker:
metrics on how- that's what I was gonna ask
Nico Vereecke:
is
Devin Becker:
about.
Nico Vereecke:
the experience? Like what you're saying around increasing engagement, but do you see a world where the game, like you can turn it in such a way that it becomes, like you could build it around an existing IP and then the game exists within that IP?
Sebastian Park:
Nico, that's a great question because it's actually sort of interesting how our bot works. So there are versions, you can, there's like a tab you can turn on or off in our bot where you can effectively make all the creatures generated on your server exclusive to your server. And that it will still show up in the main server, but it won't pull from our like library of a hundred thousand plus pre-created creatures. It will only do servers that are in your server. And we have seen one of the coolest emerging behaviors from this, where there are quite literally servers because everyone on that server loves a specific IP or loves a specific community of things. Everything is related to that thing. And so all the interactions are related to that IP thought. Now, this makes for some really hilarious takeaways. We're on a specific Pokemon server, like Discord server, where they are huge fans of one Pokemon. and they're having a hell
Nico Vereecke:
Which
Sebastian Park:
of a...
Nico Vereecke:
one?
Sebastian Park:
It's a fire type Pokemon, Charmander.
Nico Vereecke:
I need,
Sebastian Park:
They
Nico Vereecke:
sorry,
Sebastian Park:
have a
Nico Vereecke:
okay.
Sebastian Park:
hell of a time with the water world. Like they just made a bunch of fire type things. So it's just really hard for them to beat up the water thing, right? And that's such cool emergent behavior.
Devin Becker:
I'm
Nico Vereecke:
That's really cool.
Devin Becker:
interested in how many, like if you have metrics on how many people are actually customizing versus like using it more off the shelf and I don't know the process to know how easy or hard that is, but like, do you have some kind of rough metrics on how many people are customizing it?
Sebastian Park:
I don't, it's just a tab, it's a toggle. We don't actually even track it for the most part. What's strange is just emerging behavior where we didn't expect people to customize it and they're not customizing in that it's not a server owner customizing it for their audience, it's really the audience customizing it for the server. Which
Devin Becker:
More like a
Sebastian Park:
again,
Devin Becker:
mud kind of thing in a way, I guess.
Sebastian Park:
yeah, I mean, it's, it's very much again, hearkens back to DND in some ways where it's you, the user. are informing your world in this way that's different. And the way we see it, it happens over and over and over again, where hey, if you are a cyberpunk server, it's more likely that people are creating cyberpunk-themed cards and creatures for you, just because of your shared interest and your shared signaling to your peers that we're cyberpunk people. Well, as it turns out, as a result, like, your instance of creature craft just looks different than my personal server instance of Creature Craft, which has everything enabled.
Nico Vereecke:
So
Devin Becker:
When are we gonna load this
Nico Vereecke:
do
Devin Becker:
on
Nico Vereecke:
you...
Devin Becker:
Fogdow, Nico? When
Nico Vereecke:
Sorry?
Devin Becker:
are we gonna load this up on Fogdow, Discord?
Nico Vereecke:
That is a good question, Devin. Nothing's stopping you, you know?
Devin Becker:
Alright.
Nico Vereecke:
You know Seb? You have access to the Discord, so you can set it up.
Devin Becker:
There we go. What
Nico Vereecke:
Seb.
Devin Becker:
have you done?
Sebastian Park:
So I will say you need to be a server admin or owner to set it up in order to prevent people from adding things to servers
Nico Vereecke:
Yeah,
Sebastian Park:
maliciously.
Nico Vereecke:
we've disabled that. Um, so tell us about how you're thinking about the future. Is this like, do like, what are your thoughts on discord as a platform to start scaling, because I get that it makes a ton of sense to distribute your game and get people engaged. Um, but part of me struggles to see something like, let's put it differently. Like I could see something really cool. come out of this concept and this idea. But that feels like it should be out of this court. What's your thinking there?
Sebastian Park:
Yeah, it's interesting. The answer is I don't know. And I love when I don't know things. Like it's one of my favorite feelings, especially when it comes to company building. It is a great unknown. I love Discord. I think Discord is an amazing platform because it's an underserved place where we can get a lot of people to play our games. And if you are a game designer, a game creator, a studio or a publisher, having someone play your game. is like 95 to 99 percent of battle. Devin mentioned the one star reviews earlier. I always found it amusing when people gave one star reviews because, you know, it's bad for the business, it's bad for the game, but you see a bunch of reviews. That means people are playing this game. It's pretty crazy how many people
Devin Becker:
Yeah,
Sebastian Park:
are playing
Devin Becker:
you have
Sebastian Park:
this
Devin Becker:
to at
Sebastian Park:
game.
Devin Becker:
least download a right to give it to leave a review.
Sebastian Park:
Exactly. Yeah. And so one of the things that I see specifically for discord is that we have an early leg to just figure out what loops work. We're launching a new game mode or a new game, depending on how you think about games in the server twice a week, three times a week, we have entire ideation sessions where we're just like, Hey, is this ridiculous? Should we do this? Sure. Let's
Devin Becker:
That's
Sebastian Park:
do
Devin Becker:
the
Sebastian Park:
it. And
Devin Becker:
AI.
Sebastian Park:
we have it and we built, we built it up and we, we launch it. And I gotta be honest with you, most of the stuff just not compelling at all. Like our, our usage, we track usage rates on each mode. Most of the modes that people are exposed to are rarely ever used except by like the super, super power users. Shout out to like this one guy who really happens to enjoy this one mode that no one else uses, right? Like good for them. That is often what we see, but we want to be testing this. That said, it's not that discord doesn't have problems, right? UI. And UX on Discord is really hard from a game design perspective. And it's really hard in a way that you have to work around in very strange and esoteric manners. So a good example is that it's very tech space because it's Discord. And so you end up having to use a lot of custom animation, right? How do you generate animation? 2D animation, 3D animation, like that's stuff that's annoying, but useful. And, and some of the, we are working through.
Devin Becker:
Could you put in GIFs?
Sebastian Park:
We have, yeah, that was a big innovation.
Nico Vereecke:
What's a GIF?
Sebastian Park:
It's like an animated 2D image.
Nico Vereecke:
Yeah, like a gif, right?
Sebastian Park:
Yeah,
Devin Becker:
No,
Nico Vereecke:
That
Devin Becker:
it's
Sebastian Park:
GIF,
Nico Vereecke:
thing.
Devin Becker:
actually pronounced
Sebastian Park:
Jeff.
Devin Becker:
Jif according to the creator, sorry.
Nico Vereecke:
Okay, good.
Sebastian Park:
But again, and I think this will harken back to our initial point, Devin, it doesn't matter what the creators say, right? That's like the
Devin Becker:
True,
Sebastian Park:
beauty
Devin Becker:
true.
Sebastian Park:
of this new world.
Devin Becker:
I'm curious, do you guys have any interest like going kind of what Nico was saying in a way about opening up like, and maybe you already do, like an API, so let's say I've got this running on my Discord server, but I also wanna open up some elements of it to things outside of Discord. Like let's say people wanna be interacting in some way with my copy of the game that they don't wanna do through Discord for whatever reason. Say they're on mobile, mobile Discord doesn't work well for them, or they wanna do it through their web browser for some reason. Is there like... I mean, obviously, you could open up to the general thing, but to individual games, like individual Discord servers, implementations or instances with an API outside of Discord itself.
Sebastian Park:
100% that is certainly on the roadmap. I just don't know why. Like we, we want to make sure we understand the loops first. I mean, we're, we're a studio publisher in that the reason you're always a studio publisher is because you are a publisher in the true sense of the word. We want to work with creators and help their games get to life. We have to first figure out what games make sense and how to do that. And then once we do that, we open up to everyone.
Nico Vereecke:
good. It feels like it's too late to start a new topic. Any final remarks, Sebastian? Any final learnings that you want to share?
Sebastian Park:
Oh, I am a huge believer in shipping. I think that's really important. I'm not a fan of folks who are in the cool indie, like indie, like office for 10 years, working on something and never comes to daylight because it wants to be perfect. I will say that the technology has changed so dramatically, even in terms of how we understand it and also technology in and of itself in the last six months or last six weeks even that it's so important to be getting things out to the public. We're just big believers in it. And. It's obviously hard. People will yell at you. People will tell you you're awful. Like that's part of the gig. It's part of everyone's gig in a lot of ways. So just take a deep breath, know that you're doing the right thing and then go from there.
Devin Becker:
We need to set up some AI filters that somehow can like process that feedback, right?
Sebastian Park:
Yeah, exactly. I think that's been really important for us. And that's something that's been super resonant.
Nico Vereecke:
Awesome. Good. Yeah, as I said, I think it's good. I think it's perfect to call it here. So this was really interesting, Sebastian. I'm super excited to see you progress. It feels like you're at the early stages of innovating on these axes, and I'm very curious where this AI-helped-created-driven game ends up.
Sebastian Park:
You and me both. Thanks for having me.
Nico Vereecke:
Awesome. Sepp, Devin, thanks. Listener, thank you as well. If you enjoyed it, let us know. Join the Discord if you haven't already. Maybe we'll play. Well, Devin will get the Sepp's game in there. And so we can play that together. And with that, we are out and look forward to speaking to you in the next episode. Ciao.
Devin Becker:
Yeah, I'll definitely, is it easy to set up to just try out on like one of the Discord servers I already have?