 Cool, all right So at lightspeed we love partnering with extraordinary founders including in gaming people like you We're now managing almost 30 billion dollars across 13 global offices including an increasing European presence in London in Berlin in Paris and in Tel Aviv Really excited to share the stage today with two gaming legends So Hilmar Michael, maybe talk a little bit about your background and the games you're working on right now Yes, I am Hilmar Werber-Pietzon. I'm the CEO of CCP. We're mostly known for even line Which is celebrating its 20 year anniversary today or this year And we were in September busy announcing our third decade plans and Yeah, that's me I'm Michael Chow. I'm the CEO of the believer company if you've played any games of mine before it would either be words with friends which is from my first company new toy or League of Legends in particular on mobile and console which was sort of the terminal point of my journey with my second company which I sold to riot and The believer company is my third company which makes it seem like I enjoy starting games companies It's pretty nice when you have a phenomenal investor like Moritz and lightspeed here, but oh my god But our focus is on making what we would call next-generation open-world games Awesome, and we want to talk today about the intersection of gaming and AI Michael, maybe we start with the most important people in this conversation the players So how do they feel about generative AI? What are they looking forward to? What are some of their biggest concerns? Yeah, so I mean AI so I think it's the right question first of all is like what do players care about to me some of the hype trains we've had in games in the past like web 3 and Yeah, sort of a bundle of probably throw the metaverse in there a bundle. Yeah a bundle of things around that had Sort of a Darth or an absence of players really caring about it AI is interesting in that players are pretty passionate. I think they're pro and pro and con The pro part is like players want big rich deep games that are really huge and they think that AI can help us do that and It's a good example like there's this fascinating debate this year about Baldur's Gate 3 where it's this really huge and rich game and Players loved it and then some developers online came out and they were like, please don't hold us to the Baldur's Gate expectation going forward which I Thought was kind of crazy but But it's sort of this players have this desire for really big and deep and rich games and Even indie games like Baldur's Gate is somehow an indie game at this point they want us to be able to sort of Do that more easily I think the downside the thing that that they don't want is they just don't want us to fuck it up They don't want us to make bad games by by using technology that makes the games more mediocre And that is the fear that players have The converse example I'd use is if anybody's been following The Silent Hill the new Silent Hill sort of interactive experience Where the writing is so terrible that the players think that it was written by AI and then the the CEO of the company had to come out and say publicly No, we didn't like only human beings wrote the story in the script But that's players expectations is that they can they can either be great or it can be terrible and Yeah, we have to our job is to sort of navigate it in between Awesome and Hilmar like kind of piggybacking a little bit on this like one of the firm beliefs at light speed that we have is that in a Platform shift like the introduction of AI and kind of drawing in an allergy to the emergence of the internet the introduction of the mobile phone We believe that the companies that will knock it out of the park are the ones that built completely novel Impossible experiences versus just incrementally improving, you know Dev tooling workflows, which is actually where we see most startups today is is in that space And we're kind of like trying to fish elsewhere So if we if we think bold like what do you think are some of these magical? Unprecedented experiences that players will now have with the help of AI so Maybe piggybacking a little bit on the on the conversation you were starting is that so we who make games we We talk a lot amongst ourselves about like The cost of them the risk of them the tools that can increase the order of success It's a lot of risk Mitigation because there's a part of making games, which is a bet on our creative vision Back done the ability to execute etc So and we have this tendency of then take our sausage making on States and talk about like now we're putting this one ingredient in this sausage and now it's AI we're putting AI in the sausage And I think generally gamers just don't really care. They just want the sausage They don't really want to see the sausage factory. It's like I'm sure the sausage factory is under control But please don't show it to me. So but we're always taking our sausage factory on stage and being shocked that now that people see the Ingredients they have like massive opinions about the ingredient. It shouldn't be AI. It shouldn't be blockchain. It shouldn't be Direct tax it shouldn't be a PlayStation like all these ingredient talk We sometimes do and then we're surprised that people don't like talking about it Because they really don't they just like to to play the game or eat the sausages They don't really care so much as we do how they are made exactly So we then take your question into it. Okay, what are people really looking for? I think most games are a form of power fantasies So I want to imbue players with some abilities They don't have in the real world and I when I play a game. I usually want to grow in Ability or most games are some kind of mastery journeys. I'm playing with a little sword I would like a big sword. I can jump a little distance I can jump a big distance and now I'm learning the game through mastery of the game. I gain more power So when we take something like AI and what it could potentially do It could and it's kind of fully final form Maybe provide ultimate power fantasies where it is really rearing from my input What I really want to do as a power fantasy I mean, I want to be a wizard. I Want to be a wizard who don't mean you're almost there. Yeah, I want to I want to make a fireball in my hand and Back to the sausage making Making that happen for people with the sort of sausage factories. We have today Is very expensive. It's very unforeseen But I think the promise is that if I as a player am Interfacing with something that can mold my universe in accordance to my imagination Which we kind of see in kind of the mid-journey loop of like when you're alone with mid-journey. It's a big Powerful experience and then you show somebody else what you did the mid-journey and they don't care about it But I took care about it. I like it a lot. This is my imagination But the rendering that in the format of a game where I really could just make a fireball and it happens exactly as I expected it to Because there was a magic AI reading my desires and rendering to me in real time in a unique way to me I mean looks like you want to react to that a little bit too and maybe I know you have strong perspectives on that Because you know one one thing we also always try and figure out is not just where will value be created but who does it accrue to and You know here as well. We try and draw analogies to previous platform shifts If we look at the interaction of the mobile phone for example a lot of the value ultimately in accrue to incumbents It also created new companies including local superstars like Supercell and and Rovio You can you can add into this category, but you know with with this like bold Risk-taking even in the in the you know back in the factory Who is best equipped to win? This AI battle is it the startups or is it the incumbents and maybe I'll just throw the startup Perspective to you and then maybe Hilmar you can make a case for the incumbents So the first thing I'd say is I'm not sure I agree so Obviously, I think it's the startups That's one of the reasons that we started believers an independent company rather than staying within riot for example So I think it's startups I'll explain why but I I don't wholly agree with the premise yet and it connects because I think the mobile revolution was very important for gaming I do think a lot of it accrued to the incumbents, but it was over the course of making big companies that sold You know to the incumbents. I mean, you know the Microsoft's acquisition of ABK is partly, you know a part of You know King being acquired by Activision Blizzard in the first place and the fact that they've built Pretty enduring mobile business there so anyway the reason I think there is strength in specifically for AI for startups is Structurally, it's mostly a pretty classic innovators dilemma problem I think if you want to get AI right in for game development in my opinion You need to have the right people you need to have the right Organizational structure the creative workflows that correspond to that or organizational structure the right financial structure and the right company culture I think to make it all work Well because the metaphor that we use often is Pixar Pixar was not just a movie company They were a serious technology company that had to invent new ways of making movies based on the technology that was available to them we have to kind of do the same thing in games and and the risk is With too much of the existing structure and in a lot of cases, maybe the wrong people I think it's really hard to take full advantage of the technology to make games the example I'd give is My last team in League of Legends was a 1,300 person team we had two senior technical artists on that team for 1,300 people at believer we have two Technical art directors for a 50 person team and that's kind of like that's what the balance needs to be for us Certainly at this point and so I think as a startup when you can start from a clean slate And you can build the right people into the culture you can build the culture around them the workflows and the organizational structure to support it Then you can take full advantage of how to do games in a completely new way Whereas I think it's a lot harder to do it as you know as a larger incumbent company That's my my perspective So I think there are two General activities in businesses. There is like the exploration of something finding something new and there is the exploitation of it Make an industry. That's how I look at life in general. Yeah, I mean this this is what it's about and and any Any cycle is a mixture of both? I mean you need to explore new things discover them and then you need to scale them up and deliver at mass You mentioned picture pictures are a great example. They explore this element. Where does he scale up? It scales up through Disney Supercell local rock star very similar journey So I think it is a mixture so I'm excited for both There is an aspect with AI, but it's such a broad term that okay that might be a platform It might be an ingredient. It might be many things and it is many things So there is a lot of refinement Capability in it. I already see it at my company. We really have to do much Our people at CCP are just taking various AI techniques and just Doing greater work in their current jobs And I think that is a very unusual aspect of of of technology Disruptions that this is led very much by just the people Individually learning how to do things whether it's cold pilot for code whether it's a mid-journey and staple diffusion for Inspiration and managing sort of ideation and all these things Chat dbt for a lot of just even just internal communication that just becomes better where you have kind of a Sort of a scale editor for everything you do So I think there is an aspect where a lot of the sort of value out of this is just captured through people just having better tools to do their current jobs and then kind of do Better work or more work Both things are happening But it is overall Leading to an increase in quality You mentioned co-pilots. I want to drill down on that We also see this bifurcation of Some tools focusing on design time others on runtime. It kind of plays a little bit into you know Some developers game designers being extremely excited about runtime AI and the ability to create Intelligent NPCs or emergent behaviors branching narrative But others wanting more creative control and ultimately a tighter grip on the final product like Do you this is a false dichotomy to say what will win like design time AI runtime AI? Do you do you think we'll have both like what's your take on that? It's a false dichotomy like the So if we take a look at for example our game even line Even line The star map of even line was made with a generative process it was a procedural system that Run a big bang simulation and through sort of Crystal accretion grew the whole world based on design primitives and it literally was sort of a the solar systems were walking around That generative structures to find their place in the universe. It was a generative algorithm It wasn't based on machine learning, but it was a generative algorithm So and that allowed us a small team in Iceland of 30 people to make 5,000 solar systems That were all astronomically accurate based on our based understanding of the world at the time So we used a generative process to do kind of creativity at scale Because if you were to have done it manually it actually in a way loses creativity because like manually placing 5,000 systems you at some point are just repeating yourself. Nobody has the creativity to make 5,000 unique systems So that's an example of like using generative methods whether they're a I must learn learning neural networks Whatever they are, but they are generating an output based on creative inputs And that is a good thing and that will just continue from even line to whatever means people are using today When you come to the runtime effect Then also games are obviously famously for using runtime methods Which aren't language models, but they they used to be called a I mpc's in a one line are Excipiting some degree of intelligence Certainly, there's a lot of intelligence attributed to them. Even if you look at how they actually behave Maybe not like that intelligence, but we have such an ability to attribute human traits onto many different things We say insects are angry and and things like that So I think I mean it's a false dichotomy both things are very valid and Shockingly both things are very familiar again in one line made to generate the process Currently uses very primitive AI methods to have mpc's both these things will get better through these methods. I Want to shift gears a little bit and talk not just about the technology But also the the culture and the leadership that's necessary to implement AI and Michael I know you thought about this like what's your take on AI culture Like what do you think is required for teams to really be successful in this new world? Yeah, so this is one of the reasons I think I think Insurgents have an advantage is that you get the time and the cycles to really crank on this part of it. It's really hard I will say that I Think most developers we still don't really have their heads wrapped around what the potential is We won't for a few years like in my opinion It was maybe last year that game developers as a community Got like a B or a B minus on mobile, right? so I think it's going to take us a really long time to get there with AI and and a lot of the culture crafting is done in this context where There's sort of like a there's a lot of ongoing kind of nastiness between Technologists who I think come off as really insensitive to creatives and creatives who come off as really Luddite to technologists and and so one of the things that we do to cultivate that at believer There's like two things in particular. The first is what we call hands-on curiosity So we we mandate that everybody spends 40 hours in the thing that you hate So if you're in a concept artist or you're an illustrator and you think mid-journey is terrible and it's evil And it will never be good for you by the way That's like we probably won't use mid-journey still You have to go and sit down and spend 40 hours cranking through it to really understand the depth of it It's not to not even in the ordinary course of your work But just to deeply understand what's what the capacity is and understand how the technology might help you So that's one thing is we we require that people get hands-on and they open their hearts and minds to things that they Otherwise think might be really bad The second thing is it's a general what we call intellectual pluralism. It's if you read the Brené Brown book Dare to lead it's kind of this the idea of of daring leadership, which is fundamentally it's vulnerable and it's open-hearted and open-minded and That is a core part of it too, and we actually sort of do some active practice around How do we get people being comfortable? Like debating with one another in a way that is very sharp, but is also very loving very respectful How do you learn how to change your mind? How do you recognize your when you have your own bias? Because to use a for Brené Brownism, we're not here to be right We're here to get it right and making sure that everybody has that mindset and we cultivate it and how we work together is Really important, but it's also really hard. Most people don't want to be they don't most people want to win arguments and they want to feel righteous and even when we have a really strict way of bringing people into the company and and Recruiting based on the right culture We're all human beings and we still have some of these same challenges So I love what you said recently about playing to win versus playing to not lose. Yes, talk to us about that Yeah, so I think that the in this space I Think that it's very very easy to play to avoid failure and It's it's easier for big companies to do that because it's a core part of what you have to do like the example that we use is Relevant to us is Grand Theft Auto 6 Cannot be very good much better than Grand Theft Auto 5. It can't be they have to they cannot lose with that franchise They cannot destroy that franchise There's a lot of player player value caught up in it And so they kind of have to play to not lose to some degree versus playing to win which is taking really huge risks and trying to create disruptive value and Like you know GTA 6 is you know It's been 10 years over 10 years since GTA 5 So it's really hard for them to to make it that much better And I it's a tough position for them to be in but But yeah, I like yeah, I don't envy it I don't envy that position Yeah, I mean we're early-stage investors because we share that yeah sentiment and hopefully come out on the right end of it Hilmar, what do you and your leadership team do personally to prepare yourselves and be best equipped for this new AI landscape? I think generally Training One self and the culture of the company in curiosity. I think curiosity Really is the path to all innovation and and To all changes if you're not curious about something different Then it's very hard to change. So I think Curiosity is kind of a winning strategy of like maintaining that as the company Becomes older and bigger and all those variables. How can you maintain curiosity and again to this element of AI usage at CCP? It's very much driven by curiosity From just everyone throughout the company It's not some big corporate strategy that we should all go and do some AI stuff it's just we can trust the company to be curious and And that in my experience has been a much more productive way To innovate it's just to lean into the curiosity of everyone at the company that kind of sort of filters in or it filters out Often when you see you in a company then you want to set strategies and directions and all that then of course There is a part of it But as they say culture eats strategy for breakfast So if you have a healthy amount of curiosity in your company generally you can trust it to sort of move to the to the to the Right direction that makes sense obviously within some boundaries and constraints But once you have set up then curiosity is a very powerful thing Then there is a general thing. I've seen both the gamers and game developers is that They have this beautiful ability to hold two opposing thoughts in their brains at the same time They don't like the status quo and the hate chains at the same time And and and and human beings can't do that People are always in contradiction with oneself. We have two logical impossibilities In our brain all the time and it happens all the time sometimes people are Critiquing each other about not being consistent, but we never are we can just really be in this state of like hating the status quo and Hating chains at the same time and often the path through that and we have learned this from decades of interacting with If players is more about showing rather than telling like every time you're trying to explain people about some very abstract Concept that's in the sausage factory somewhere here and how everything is going to be better if you just have that They just like I don't like that, but I also don't like exactly what I have right now I want that to change also and the best way to do that is just to live deliver things and Have people play with them rather than to be talking about them endlessly I don't know how it's going to be awesome once we have this new piece of ingredients Which now magically is going to get everything better. It's about like talking about chili versus just tasting it So yeah, I would bring that curiosity and people have the ability to both hate the status quo and Not liking chase at the same time You know while we're already segueing into more touchy feely topics and while we hold you captive on stage here for another five minutes or so I want to I want to go a bit more Vulnerable and authentic you even mentioned vulnerability if You could reflect on this almost past year of 2023 Where do you think you failed as a leader and and what are some resolutions for you going forward and I'll start with you Michael? Yeah, this one is sharp for me because I had feeling it maybe for a few months Like even answering some of the questions today you can feel I have like a little intensity to how How I feel about people who are sort of like not adopting this really like not being proactive Being entrenched in their own thoughts not being sufficiently curious not listening to other people and I've spent a few times this year being kind of like disappointed Asian dad to a lot of a lot of folks on the team who are struggling with it in that way I It's and it's not the place that I naturally come from But it's a little bit To be honest like a technique I acquired in my last role because of that 1300 person team about a thousand of those people were Chinese developers in the Tencent is ecosystem various sister studios and There's a book called culture map which I recommend to anybody if you're ever doing any cross-cultural co-development Strongly recommend this book. I love it and I hate it because one of the things I learned from it is in Eastern knowledge where cultures you you can't always use love and inspiration you have to use Disappointment and you have to sort of anchor the emotions around that and you have to use like very hierarchical mandates To get the job done and I think I kind of adopted that that's like a mode that I added to myself over the course of working on League of Legends and for like six years and And I kind of like I get caught in that mode a little bit And as it pertains to this work where I think we have a bright future And I see people dragging on actually realizing the future. I kind of get disappointed Asian dad mode And I can I need to work on that What about you Helma so? it's the downside of being CEO is that the mistakes you make Often don't really reveal themselves until much later. It would be Very easy to up if there was more immediate reinforcement And I do miss my development days when I was a game developer was making a 3d engine Especially making a 3d engine. It's a very close reinforcement loop when you make the right thing and the wrong thing is like visible in a span of days and weeks But when you're a CEO, it's a span of months and years So if I now especially in light of the fact that he was turning 20 years old Probably my biggest mistake in my 20 years Well, yeah 20 years of being CEO is not to have started the intern program sooner The we live on a tiny island cause Cole Iceland there's only 300,000 people there when we start There are no game developers now 20 years later. We have amazing game developers, which are natural born within CCP But if I were to have kept the intern program as it's almost like a talent pipeline over those 20 years the the the scale of the talent base and the footprint the human footprint of the company would be much bigger If we're running a company make sure your intern program is robust I think with that we're on time. Maybe one conclusion is that With AI or not even building games are still a people business and needs to be treated as such It is that thank you so much. Yeah, thank you. Thank you. Thanks you guys