Panel: The State Of AI: Sustainable Or A Bubble Waiting To Burst?

Third panel is the state of AI sustainable sustainable or a bubble waiting to burst. Okay, so um here's the panel. It's a state of AI sustainable or a bubble waiting to burst from. We all want to know the answer to that. Um and I introduce you to RA from Raptor PR who will be moderating. And let's have a round of applause, PLEASE. [applause]
Good morning. Um awesome to be here. Um, I've I've had AI in my life probably for

quarter of a century because I'm just such a massive sci-fi fan. Anyway, it's here. Um, Rana, founder and CEO of Rat PR. We are a pure B2B tech agency for the games industry. Um, I'd love to introduce the panelists. Um, if you could just introduce yourself, what you do, um, and the company you work for, that'd be cool. Charlotte.
Yeah. Hi, everyone. Um, my name is Charlotte. I uh run my own consultancy business um which kind of sits at the intersection of sort of strategy product and then commercial viability um slreity. Um I've been in the games industry for 15 years. Um I've previously ran uh all the revenue delivery at team 17. I've worked at Sega Take 2 and I've also worked on the service side of the business. So I've kind of moved around all the different sort of areas of the business. Um and I'm currently working as a fractional um chief commercial officer for two businesses who have a an AI focus but in very different areas. One is on IP translation and sort of democratizing um licensing through IP. Um and then the

other is very much focused on sort of QA optimization um leveraging AI as well.
Awesome. Hi everyone, my name is Pal. I'm the founder of Pirate Blood Studios. We are an AI first gaming studio. We have built out some some some tech that sits at the core infrastructure and UX layer called pirate link and it enables real time voice to action. So basically what it does in very low latency is it converts your voice into any action in the game. So it takes probabilistic input which is the intent i.e. you talking and it turns it into a deterministic output which is an actual output in the game. Before that I led product in audio at Tik Tok in R&D. I've had a startup before which did quite well and recently I was director of tech innovation at Baller League which is a six-side football league and very much I've been across sport entertainment and media all my life and yeah I was thinking whether I should do a little rap on this microphone because it always

I always feel inspired to do so and hold this but I I I pulled out of that so I've spared you. Nice to meet you.
Hi. Hello. Very nice to meet you. My name is Orian. I'm the co-founder and COO of X and Immersion. We provide AI tech engineering for the video game industry, ranging from production to live playing features. Thanks for having me.
Cheers.
Hello, my name is O and I've been in AI since 2017. Did some different things like some machine learning, AI consulting. Uh I had a startup. We did UA automation, marketing, automation with machine learning. And then uh I was head of UFO on other AI like text to image and now I'm advisor to several AI startups and at transform gaming I'm chief AI officer like doing the AI strategy and in all different use cases in gaming for machine learning AI and nice to see you awesome thanks guys. So

um the title the state of AI sustainable or bubble weighted to burst. Um so I began my career in the dotcom boom um late '9s you know early naughties it was new you know there's huge amounts of investment huge opportunity but it was the worked world back then um so AI is here it's been around for a while but it's now you know it's mainstream in a business and consumer perspective so it's a diff it's a different landscape um so what does sustainable mean in reality with AI nowadays is it time saved, you know, money made, risk reduced, player player trust improved. Um, what what does that mean to you? Sustainable AI right now.
Well, uh, when you say I would say what I understand from sustainable is is it going to be used like two, three

years down the line, five years down the line and I think the answer is yes. But maybe there is some like we need to set expectations right. So
today I think is most useful for speeding up things. It's not a magic thing where you just sit back and things happen for you. It's more like uh you know let's say take an example of UA creatives marketing creatives. It will help you much faster.
Yeah much faster uh create creatives and try things. But you still need the to actually someone experience managing it. So I would say a a already is in uh life in the in gaming studios but
you know we need to kind of manage the expectations there. Okay, Orian.
Um, I think that for me it would be something like
always enhancing the player's

experience. Like everything that you do in your game, everything that you work on that you yeah that the creative vision I would say needs to serve one purpose and at the end is like is your player having fun or not. Um, so yeah, I think that sustainable is it's an interesting word because of course things are evolving super quickly. Um, but at the end you can try out, you can make some experimentations. It's going to be super interesting, but what you're going to keep is what your players are going to have fun with in the game. Um, so yeah, for me I would mostly say enhance the players experience
through like AI tools or whatever AI means because it's very very wide.
Yeah.
Yeah.
Does anyone have a perspective on like a macro investment level? Is this kind of incredible investment in AI sustainable? Do you think it's going to carry on? Is there going to be, you know, a downturn

then a recalibration? Is it going to follow like, you know, an emerging tech kind of curb?
Um, I think I'm trying, we're trying to fund raise at the minute and it's hard. Um, and we're an AI business and I think what you have is you've had a lot of investment in AI businesses. Um, and these VCs and investment investors are looking and sort of waiting to see what happens. M so I think there maybe there has been that investment bubble and now I was reading something recently that that explained with the dot bubble exactly that that you had all this you know going in and then actually it was the businesses that survived that and came out the other side that are the ones that we know today um the likes of Amazon you know and and the way obviously they've grown into these huge um organizations so I think we're in this kind of wait let's wait and see sit tight minute which isn't good for obviously you know the fact that obviously it's the advancements in technology um
the investments ments you could be making today arguably could be better than maybe the ones you made a couple of years ago. But I do think there's this sort of hesitance exactly that because there's this rhetoric around you know

waiting to see what happens in adoption. Um and I think the risk we run as well is that you know especially within the games world is studios adopting adopting adopting and then quietly quitting and um therefore those services having a huge turn in terms of users and tools internally and that that could sort of fuel um a bubble further. Sure. Pal, any perspectives to share?
So investment right now is hard.
That's undeniable. The the game or the playing field has changed. So,
you know, a few years ago, I remember when we were in that web 3 sort of boom,
you could literally raise capital with a good idea in a nice pitch deck. Whereas now it's changed to the point where you know with all these actual sort of tangential AI tools like let you know high level example lovable
you know those sort of things anyone can build a prototype now so VCs themselves

are actually looking for different things now so like if you want to raise preede they want to actually see that you've got you've got that distribution you've got a customer base that's already there you've got you've lined up letters of intent so if I you know I can't just go and try and raise capital now and be like, I've got this great idea. And even like before you could have said like, oh, I've got a prototype. Well, everyone's got a prototype now, right? So, how many people have you met in your line of business? How many users are there that actually want to use this? Is this a problem that people want solving? So, you know, it's actually made there's more work for you to do as someone raising than there was previously. Now, the good thing, you know, it's good and it's bad. You can build a prototype now and you can get it into people's hands, but you've got to do more work. you know, there's you've got to you got to like think about a bigger sort of chain or or funnel that you've got to accommodate with all of these new tools to try and raise capital as well. So, you know, it's given us opportunity, but it's also given us more work as well in terms of showing traction, I would say.

Just had a thought, um, I go to a lot of conferences around the world from console and PC to mobile. We're obviously at a mobile gaming conference. Do you guys feel like AI is more embraced by mobile gaming than say console and PC? Or is it a chasm? Mobile tends to adopt emerging tech quicker, one would argue. But I'd love to get your you guys opinions on this.
Yeah, I agree actually because uh I think mobile teams are usually smaller.
Uhhuh. And especially the new startups, I mean there are less I would say there are much many more mobile gaming new companies from startups anyway. And when they do start now, they want to be as efficient and as fast as possible. So they adopt AI already as much as possible versus most many of the PC console are traditional companies that have to worry about IP lawyers like bigger uh bigger institutions that needs like

are you saying that mobile tends to be more innovative in that respect in terms of wanting to experiment with AI?
Yes, I guess there they move faster. Absolutely. Yeah.
What do you guys think?
I've got a view on that.
Okay.
So, I think right if you think again of of the stack as a funnel
Yeah.
Prefunnel you've got user acquisition, right? So, you AI is undeniably helpful there, right? So, like you've got you can look at profiles of people who will be playing your game. Like I think Google and Meta have got this tech
where it looks at the player profile. So they can tell like, you know, from the from the first day you play a game, you can see like by this player's behavior in that game,
they're going to be in that top top 10% of customer lifetime value. These are the people that we should sort of overindex on. And then at the end of that funnel, you've got the c the customer support. So obviously AI really

helps with that with the sort of generic answers and FAQs. Now when we talk about AI inside of a game, that's where it gets difficult. We saw all those AI NPCs that you talk to and there's a like 5-second delay and it just feels so wrong, right? So, the AI in the game is where the magic needs to happen. So, you got the you got the front of the funnel, which is fine, and the back of the funnel, but inside the game generating AI content in a game.
You know what players want is they want reliability, speed, and trust. AI as an LLM, it can hallucinate. it can come up with different responses. There can be a delay when it comes to a response as well. So, you know, you don't want to be sitting in a game waiting for something to happen, which is why the adoption of AI has currently been quite slow.
Now, sorry to circle back to pirate blood, but you know, unashamed pitch. This is where, you know, this is why we've built a [snorts] hybrid logic stack, right? So, in our game with our

tech, you know, we've got a machine learning layer which takes intent detection. So it understands what you're saying or it makes an inference because there's a number of ways that you can say something move forward go forwards you know from that intent we take we map that to an action graph where there's a deterministic output so that output in the game is always the same right which is forward right so I think you know we are going to get to the AI is going to get better but it needs needs to work in a way which is which just feels like magic or is seamless to a user. It doesn't feel like it's there. So once we get to that point, the adoption is going to going to go much higher. But I think right now we we need sort of hybrid approaches to make the feel of it within a game feel seamless.
Yeah, I agree. I think I think the stakes maybe feel a bit higher in the PC

console world. And I think um purely also because you you can't iterate as quickly. you know, it takes much longer to do updates, patches. Um, you're not doing them daily, you know, or weekly. Um, where the mobile space thing can move much much more quickly. And I think, you know, once you've lost player trust, that takes a long, you know, sometimes you can never get it back. And other times it it can take months, cost significant amount of money, a lots of pivoting um to get back to that point. So, I think exactly that. I think until the the adoption of AI can seem frictionless to a player, I think it's going to be slower to be adopted by sort of PC console um teams just because the risks I think the stakes are higher if it if it goes wrong and and is is obvious um and disrupts the players experience. Yeah, I I would add I completely agree and I would add that I think the difference between mobile and PC would also be like the business model
because when you're a mobile um gaming uh production, you're like iterating of course very fast and then you generate the money and then if like the tool or

the AI or whatever you've been using is not relevant, you just drop it for the next production. um for the PC console because production timelines are really bigger, you want to make sure that you use something that is really relevant for the game. And I think because of AI right now is kind of finding its place regarding the business model. Like we don't really know what the um the long-term business model is going to be. uh I think that people are it's easier maybe to adopt tools uh for the mobile whereas for the PC console you want to make sure like do I use your local inference do I use a third party but that I have a deal with I mean it's I think everything in the end revolves also to what's what's it going to cost to me like what's the price and of course prices like AI is very like um hard not hard but like uh It's it takes money to run AI obviously. So if you

don't know how much money your game is going to make in the end, you don't really know how much money you want to invest.
So yeah, I think the like the costs pressure is also very very big for all the AI service providers. And so
it kind of loops in with your last question of but what where where are we at for like the gaming investment? Well, if investors don't really know what the final business model is going to be, well, they are going like maybe to invest lower because they want to know, okay, what where's the money in it, you know? So, yeah,
I think one thing to add is I think the use case like in mobile it's more about like creating the game, marketing, optimization, library, those things in PC. This actually the topic before us was about moderation community all these things they already are using it there so it's the adaption is a bit slightly different
and I think personally I think UGC type stuff customization those things will start to use AI [clears throat] tools

I mean they already using it but adoption will go up
interesting um so if the bubble does pop um what do you guys think it would look like and do you see any early warning triggers? But it' be good to yeah hear your thoughts on it.
I very quickly I'm not sure it's going to pop like crash, but I think it might what we could see is like a gradual like people stop using AI less and less. And I think that really one of the the things that we see in our line of work is the the cost of of AI and the cost pressure of AI. So if loads of service providers are like raising their prices and then at some point it's going to be completely unaffordable for studios to use AI tool. So the studios are going to be like okay so what's the most relevant for me to use? um it's going to be something that has really meaningful

impact for like gameplay mechanics or features or whatever. And I think that studios will need to rethink about what's the most relevant and important thing for me. And I think it might be the end of like cheap and easy AI,
right?
So that people are going to think about okay, what is really meaningful? how can we like uh really work around the resources that we have? So maybe not a crash but like a shift of okay cheap easy is done we've done experimenting now let's do something that's scalable.
Yeah
I think as as pal was saying I think the investment landscape is probably going to mean that the b the bubble won't be there in the way that maybe we've seen sort of other trends
into the industry raise load of money and then disappear. Um, I don't think the landscape at the minute is allowing that to happen. Um, I think we're kind of at like an Excel spreadsheet moment, like everybody has it, but it's how the

how you're using it and competitors are using it maybe better than you and it's it's kind of that evening out. And I think it's as people scale up and we get to that kind of level where adoption is pretty universal and and um I think you'll see some consolidation from service providers because of that cost element. And it is it's incredibly hard to um package up and price AI tools at the minute because it doesn't necessarily follow standard SAS models. So I think there's so many unknowns at the minute. Um
I don't think we'll have a big yeah pop moment, but I think we'll probably see some consolidation.
How it's a good question. you know, I don't, you know, as as others have said, I think like we're getting to this point where the cost of AI is is going to increase. Like I saw an article recently where I think Jeff Bezos was talking about you know people won't be able to afford their own sort of machines and they'll

have to rent them which kind of happens with AWS already but like just from a personal perspective like we're building our own sort of GPU server based servers and the cost of RAM and those other components has gone up so much and I think it's going to keep on rising where it looks like it's going to keep on rising. So, you know, I think like models are still basically they're being baked in as we speak and and that sort of whole infrastructure layer is is forming and and in relation to that I think the models that businesses use, how they monetize, they're still being formed as well. So, I think you know from that perspective there will be a leveling out. I don't think there's going to be a sort of pop per se, but it's going to
it's like anything, isn't it? Is there's going to be a massive cull. Loads of loads of startups and companies won't make it and a few will and like I you know, I don't think it's about that. I don't personally like I don't even like

looking at it from a bubble. I think like what you should be looking at, right, is is the problem you're solving for the user big enough, right? Are you hyperfocused on that? do you know that area so well and you're providing so much value
all the other stuff can then work out from that you know because the AI bubble again it's it's a tool you know but if if whatever it whatever it is you're building and I see it all the time because I I advise for different startups as well it's like
the value proposition of what people are trying to pitch me and tell me is just not there so I think like going back a step I think just think deeply about the problem you're solving and and and that's that should be the focus. I mean I the bubble, you know, I work in PR. I've been hyping brands for 25 years. Um it is like a potential high point in terms of your valuations,

investments, but I'd like to hear get a reality check from you guys because you guys are the technical experts of like a real AI win that's made a measurable impact from your own personal expertise. Um so Ariana love to hear positive AI cassette success story
feedback.
Yeah.
Um thank you.
Yeah so obviously we work a lot in production like we provide uh tech AI tech services and engineering for the studios. So I'm going to be very like production focused. Um, I've met this studio who um was trying to build sorry [clears throat] was trying to build a game um that's like a goofy horror co-op game where there is someone in the van that's moderating a chat and there's someone exploring on the ground and they need to like talk to each other and the moderator in the van has like a AI

dynamically generated Twitch but where there is actual ual narrative information for the person that is exploring on the ground. So this is like the pitch game. Um they've actually so the Twitch chat is a core gameplay mechanic of the game and it's very cool and um they've been trying to build it with chat GPT just to see if it works and the answer the answer is yes but then the whole business model around it would not have been possible to be the like the game because it would not have been scalable to use it that way. Um, so I think that AI, yes, uh, is a very positive thing when you have the creative vision. So they had this amazing creative vision of the game that one of the core gameplay mechanics is used by AI.
Yeah.
But they needed someone to do like a

local inference model of this LLM of this like in-game Twitch chat. And I think that it's it's good because they've been like, you know, playing with it like like like children been like, "Oh, it's so nice. It's working so well." And I think this is great because without AI, there would not have been here in this particular game. Yeah.
Like the like the the immersive interaction that the the players got with the chat.
I think that's what it's all about. It's like using new technology to create new experiences
which couldn't be done without it. It's still very much a nent stage of development, but you know, as a as a Gen X gamer, I find, you know, new experiences kind of what it's all about. Um, guys, I'd love to hear about some success cases from your sides. Charlotte,
mine's a bit of a shameless plug as well, but um all right, go for it. Uh, so Infinite Realms where I'm I'm working currently. Um, two two things. And so one is um working with a a major IP holder um that through the use of the

technology has decided to open up to smaller uh indie teams. So traditionally this would have sat behind a a huge payw wall you know large MGS every year just to have access to the IP but by working with us um they feel safe and confident that they can put it in the hands of of smaller teams which I think for me having spent the last two and a half years working at the sort of forefront of indie publishing um is really exciting. And so that sort of democratization of large IP because you can create a tool that an indie team can interact with and become experts on the IP to stay cannon compliant and true to the to the law of that that world. Um, the other one is another major IP, a literary estate we're working with, um, who has had a vision to bring these characters to life, um, and a trans media strategy that they've really wanted to put out in the world, especially in gaming, but haven't really ever, it's almost been a bit scared to take that leap because the production times, you know, think it could take years. um we've been able to accelerate that for them um to sort of start looking at bringing true game concepts

uh to to life and working with teams to to bring them sort of reducing that that solving the blank page problem effectively. Um so it is more on the the creative side and those are two things that in the last sort of three months we've been working on which are are really exciting.
Awesome.
Yeah. So from my point of view I think it's you know what what is exciting for me is userenerated content. I think like being able to create your own worlds, remix things, you know, like, you know, if we take audio as an example, I think being able to put, you know, we I was at Pixel Links before, which was a web 3 company where, you know, we worked with with really big artists and you could take songs and we basically leveraged the Niantic Light Ship SDK, Niantic powers Pokémon Go. And what we had was a was a mix of real world in real world in real life IRL and sort of augmented experiences. So you could go to a different concert and you could find different pieces of audio from artists

in in the real world and then sort of remix those and stuff. And
I think it was that sort of amalgamation of gaming real real the real world and and bringing assets in and being able to remix them and put them in in your own sort of way. Yeah. And we we see that in in like lots of games now and people love it. Like you look at Roblox like just bu making stuff in there or Minecraft and stuff. I think being able to put your own impact or imprint within a game through your creativity is where
people resonate towards you know like that's what my kids love building stuff in in like Minecraft and and they you know being able to create within a creative medium is using AI as well to be able to do that is really exciting and being able to you know the world is a remix that's what it's becoming and and people love remixing things Right. And that's what I like doing. But, you know, you take a song, being able to split a song up into its constituent stems and play with them in a game somehow is is is very interesting to me.

And that's happening with with video and and other mediums. And there's this whole convergence happening as well as a result of that between previously isolated stacks. So, we had like gaming, sport, you know, they were they were quite siloed. And I think there's a convergence happening with them because of AI. And like classic example of that is you look at creators now. People love watching people like KSI and and those sort of creators and like you it's gone to this point where like a musician can't just be someone behind a desk anymore that you don't see. They they kind of have to have a a a presence of some sort. And again like AI can sort of help with that and and and put you into different mediums. So it's an exciting time. It's a creative time. It is it's it's a massively polarizing subject um for obvious reasons. But do you based on you know the conversation um do you think this is a great time of

democratization in terms of content creation for the games industry where you the barriers to entry are what I hear is much lower to create gaming experiences. Do you see what's what's your viewpoint in terms of you know the rise of more small creators to create gaming content?
It's a great question. You know for me it's there's one word it's a one word answer to that. It's like the there's with everything there's a positive and a negative. Right. The great thing is now everyone can create a baseline very good level of anything.
Yeah.
Right. And so everyone's creating stuff that sounds pretty good. Like you could be using any AI to create something that sounds good. The problem there then is distribution. That's the word I was going to say. Like how do you stand out from everyone else who's making really good stuff and have your your personality? How does that come out now? I think that is that's where the

magic happens, right? So, like, and I think that's where we where we need to be digging is like with all of this stuff that can create quite homogenized output,
how do you make something that's super creative or or really wacky? Like some of the best music is sound as humans, we like stuff that sounds wrong, you know, it's been proven, you know, that the recordings that sound a bit crap, like there's some magic to them, right? with everything sound or every output so homogenized and so clean and beautiful like where does the where do the errors come in? Where does where does that human element come in? And I think that that linked with distribution is is the interesting spot here for me.
Interesting. Um got to wrap it up. Any any questions from the audience for the assembled panelists?
Time for questions.
Sorry. Okay. So, so if um so if the audience if you've got any questions, please find the panel uh about the show. Um yes, sorry we're running.
All good.
Time is a premium,

guys. Thank you so much. It's been absolutely fascinating.
Thanks.
Thanks.
Cheers, guys. [applause]
Thank you so much.
Thank you.
[music]