 Trenau. Di garaiba. Lampu извimerao. arbitrary kato cal Fame I Tu Lia i mahada a Rianasha, as panel members and got Chris Cagnar as our moderator. Today's topic is incorporating AI and gaming and startups for competitive advantage. Now, this is first of a many part series on regenerative AI. Chris and I have been having quite a few roundtables with fellows. And what could the fellowship collectively create as thought leaders and users and decided to broaden this conversation out to include the New Zealand ecosystem. So during the series, the fellows want to cover like risk and reward to New Zealand, security and regulations, well-being right through to AI being used as a tool positively in business. So that's like today's session. It's about how can we use AI positively. A little bit about EHF for the first time as to one of our conversations. So Edmund Hillary Fellowship is a collective of over 500 entrepreneurs, scientists, storytellers, creatives and investor change makers who want to make an impact globally from Aotearoa, New Zealand. These live sessions are informal conversations with the fellows so you get to know them, what they bring to New Zealand and often the start point of an ongoing conversation that takes through to action. We'll be having about 45 minutes of conversation with the panel and then moving into Q&A and discussion with all of you during the next 90 minutes. Now a little bit about our moderator Chris. He's from Cohort4 here and now lives in New Zealand and is a master in mentorship, coaching, agile training and mindful instruction. He does a lot behind the scenes in some of New Zealand's biggest organisations that you're probably not aware of and I'm excited to have him moderating this AI series over the coming months. So as I said before, this session is recorded and it will be up on our website. Just stay muted until you are going to ask a question or you can put your questions in the chat. And it is, if some of you need to leave, that is okay as well. Over to you Chris. Thank you Michelle. Kia koutou. Really nice to see everyone. Good morning. I would love to introduce our panelists today. It's just a delight and a pleasure to have some extraordinary panel with us. I first like to introduce by introducing Justin Kahn. He's the co-founder of Fractal, a marketplace to discover, buy and sell durable game assets in Hamiltonies, as well as Rai, an API to power e-commerce anywhere. Justin also co-founded Twitch, which was sold to Amazon in 2014 for 970 million. Ahmad Mahajan is the co-founder and CEO of Proof of Play, a company on a mission to create compelling and fun on-chain games that facilitate an inclusive and creative web-3 ecosystem. He also co-created MyMiniLife in Farmville and he founded the CEO of Toro, which was acquired by Google. Brianna Shaw works within Binance and Technology. She has worked on deals such as gaming company Unity Software's acquisition of IronSource and think-tech company Affirms Public Listing. She is the founder of ISTA acquired an education technology venture that redesigned schools for the 21st century. Brianna was a venture capital fellow, a pair of ventures capital, where she advised and sourced early-stage startups at MIT and Harvard and was a fellow at the Harvard Innovation Lab. So I'll be inviting our panelists with some questions and we'll start up with Ahmad and I will invite Justin and Brianna as well to follow on. And so Ahmad, the question for you. How do you see AI changing the landscape of gaming in the next couple of years? It's something we think about daily. So first things, to give some context, with gaming companies, so much of the cost of creating games is creating content. And there's different stages of content. There's concept art and you go from concept art to prototype art and then final assets and so on. And depending on the type of game, there's a lot of time that's put into the various stages of that. So the simplest way is that the idea of a concept artist has changed. It goes from potentially having to go into these painterly styles to typing a few words into an AI-generative system and be like, OK, I want to do a spaceport that looks like it's kind of like on Mars but with three sons, right? Like a producer or game designer would tell an artist to do that and then they would go and do it and then maybe a few days later you'd get that back. But now they can just go straight to the source and type it in. And the idea being is that when you're building something like a game, you have an entire team of people that are trying to roll in the same direction and someone is trying to set the vision and get people aligned and kind of go in the same direction. The point of concept art is to kind of help people do that. So the first thing is that the idea of getting a team kind of centered and seeing the same vision for what the game is going to be is a lot easier now. So from our perspective, when we're thinking about new features in our game or we're thinking about trying something else out, like we can start with actual imagery. The second piece of that is then when we think about actually going to final assets and Justin has an investment in the industry and we've messed around with them. There's a bunch of these companies now where you can basically go and say, OK, I need like an icon for a sword or I need a model for some item. And now you can actually go from going to like an art store, like a digital art store where you're downloading different asset packs to actually going and getting a unique asset for your game. And so now even when it comes to prototyping the game, you're just very quickly getting closer to the final product. And for developing games, one of the key things is that you don't really know if a game is fun until you start to play it. And so you try to get to like a playable slice of a game really fast. And so this thing, this all allows you to get to a playable size really fast. So so long story short with games, it is completely changing the way that I think games are made. And it just allows you to get to the heart of making a game first, which is finding the fun. And so and this is just the surface level of it. But I'll stop there and see if I'll just have something to add to like it's it's really revolutionary. And I think it's and more importantly, it's going to bring down the cost, right? The time and the cost of making games is going to go down straight up. And so I don't know, Justin, I don't know if you guys have anything to add to that. But yeah, I think that's a really great summary. The investment we were talking about earlier was this company called scenario, which is this generative AI platform for games. And so like I mentioned, you know, like something like a third, I think of the cost of developing a game is building the assets. And, you know, the games can be like hundreds of millions of dollars or billion dollars to produce now. And so this is like a tremendous part of the industry's cost is building assets. And there's a bunch of companies, one of them I invested in called scenario that are working on creating that pipeline so that people can prototype faster. Like I was saying, but also production eyes, the stuff that they generate. So like a lot of, you know, kind of like the V zero of these generative AI platforms like mid journey and Dolly like one of the big problems with making them production ready is like the outputs are consistent. Or you might want to like create a bunch of things in the same style, right? But like, people did not do that before. And so now a lot of these kind of like new tools, including like runway is another one that's doing more of like image and video generation is about like creating consistent output and letting you like more finely control your output and create So you can like make more easily put things in production versus just like the prototyping and storyboarding portion of it. That's, I mean, it's pretty powerful because, you know, these companies, companies that are doing this right now, it's like, on the side, like an artist has to go produce this content. If you want to like create a, you know, new assets for a game, and you could do it in seconds now as a user of one of these apps. And so like scenario, like had 40,000 companies signed up in the last six months that they started, you know, and so I think it's, you know, there's really high demand in the market for these types of tools because they allow you to move so much more quickly than before. It's kind of like Photoshop like before when, you know, did, when before Photoshop was around people like did all the things in Photoshop like airbrush, you know, touch ups and airbrushing and stuff like that, but they did it like manually. And Photoshop enabled a much broader audience, a set of people to like participate and do it much to this, you know, photo retouching much more quickly. But I think it's going to be super powerful for the industry, you know, yeah, go ahead. I was going to say to build on what Amit and Justin are saying. So there's of course the like production side of gaming where AI is very heavily being deployed and it's like being able to in like a beautiful world right like being able to have these like open player game or open player games where people can kind of play and it's like things get created as folks are playing like that would be that's a really interesting part of a like future vision of gaming but then there's also the sort of the nuts and bolts of like how do we enable creators to create a future and so when we're thinking about the gaming landscape, there's a folks who are like the eight like the triple eight plus studios right that are producing these like games that are worth like millions and millions that take millions of dollars to produce and then there's like the like developer the individual developer was like producing games and it's kind of like the fun thing that they're doing like on the side. And so for so it really I think AI is really helpful for for all developers but especially for that like individual developer right like who's kind of doing this on the side who doesn't have millions and millions of dollars to deploy into a game. So how do you produce, how do you produce something quickly, and then not just the production piece but also the monetization piece so how do we then take that, and how do we actually monetize that game quickly and it also and AI helps that a lot with the like matching engine of like putting a game in front of the right people and that helping them stay on the game, etc. And if we're able to monetize effectively it just helps more game developers stay within the space, especially if you're a smaller studio and aren't already super well known and so lots of different places we can go around the like visual media piece of it, but there's also just a lot more like content around it, a lot more around the like development piece. And then I think one other area that neither of us have touched on where it where AI makes a huge difference is around accessibility. So there are these like games that you produce and you can make them and being able to sort of automatically do certain things makes it a lot easier to make the games accessible to a lot more people because right now it's you have to program every line right it's just very very expensive to do that, but being able to bring down the costs also then helps your console or your gaming platform understand what kind of gamer it's like oh you can't see in your left eye like okay like let's figure out a way to have this stuff be in a certain place so you can actually see it or your deaf and you can't hear the explosion it's like let's generate captions really quickly to be able to tell you what's happening in the game. And so there's just a lot of different things that you can do with AI in gaming and think the universe is quite broad and applicable in many different places. I'll turn off. Yeah, I'll add one more thing. So like, you know, when we think about AI we kind of split into two different buckets, it's like acceleration. Right so the idea like you know you get to assets faster and bring down some costs you don't need as many artists all those things right. Which is just like is the ability to create new capabilities it's like enabling. And so, you know, like I, my background as an engineer I can I have very little artistic ability. And for my entire life I've been creating programmer art, right, and it's just, you know, programmer arts called programmer art because I think you can imagine what programmer art looks like, right. So now, you know, it's possible for people to become like, you know, one person shops right they're able to maybe have a skill in one area and I augments you in the areas that you know you don't have the skills. And so, you know, we're seeing that across Oregon other ways but like now like our community manager can go and like type some stuff into mid journey and be able to create stuff in the game. And so one of the things that we've been thinking about is how it's also affects user generated content so you know someone in Roblox or in fortnight or hopefully in our universe as well has the ability to go and basically type a few words into an AI generator and now create content that's within the style of the game they're creating content for. But you actually enable people, especially, you know, maybe even children to be able to go and like create stuff just by talking to the computer and be like I want to see X, which, you know, also creates a whole new category of games there's a lot of folks that are starting to you're starting to see companies pop up around creating games that are specifically the game itself is you describing what the game world is, and the computer generating it for you. I'm sure we've all had ideas want to be cool if we get a game that did where you could do X, right fill in X with whatever your words are and now the games are being given back to you by the AI. So there's like all these different capabilities and we're only just scratching the surface of it. It's just like it's like trying to imagine like what we could have done with the iPhone back in 2009 when it first launched and then trying to imagine tick tock like it's just this too many levels of successive generational leaps in terms of capabilities to actually know how it's going to affect the game industry broadly. But like, I think we can see a little bit ahead, but I'm, I think, you know, five years from now it's going to be like, you know, oh wow we need to think of that. So, you know, one cool story that I have, like somebody using generative eyes, or like, I guess London, it's a, they were using chat GPT, one of my friends, who is not a programmer he's a marketer has this idea for a word game and like and so he asked, he basically described it to chat GPT and then asked it to start outputting code, you know, and like it built this game like he kind of iterated in programming the game for him. And this is someone kind of like Rihanna and I'm with we're saying is, you know, this is expanding the market and people can create games because it's someone who like actually doesn't know anything about code. So, you know, I thought that was pretty cool because it was like a cool concept per game like it probably would cost him a couple thousand dollars to have someone else build it. And that's like, you know, he's going to launch it actually and see if people a lot of prototypes game and see if people want to use it. I don't want to play it. Yeah, I want to try it. It is I won't give away the idea it was you taught me that could describe the idea but I will, you know, someone else that won't give it away because it kind of told me confidence so we'll see what happens. I'll send it to you after it's done. It's like a game once you know the idea you could program it with chat GPT and you know, a couple hours I think. So one of the things that I'm hearing is that generative AI is going to make creating the type of work we do now much faster is going to accelerate. But it's also going to enable entirely new use cases. And Riana, I'd like to kind of see you with this question to start. How do you think generative AI can help some of the smaller game studios such as those found in New Zealand differentiate themselves in a larger in a highly competitive market against some of these triple A plus game studios for that matter. And how might in the present independent game producers, they relevant when anybody can create an indie game just as Justin is saying about a lot of skill or background. Yeah, I mean that's an interesting idea right it kind of is almost kind of like how do you create something that someone wants to use, and that's the fundamental question here right it's like, I think the thing that generative AI does that's helpful is that you don't need to have a ton of resources to pour into something to actually produce something that's of quality right, and of course it's not, we can't just like, you can't just put crap it I mean it's like generative AI will like give you trash sometimes right and you have to really be thoughtful about it and so I think what is really going to distinguish what games get played versus not get played, it's really going to come down to quality and what you're thinking about your, it's like fundamentals of like creating anything right like do you have a target audience in mind, what is it that you're enabling them to do through this game. How thoughtful have you been and actually creating it and producing it, and I think the thinking the generally I can't really do the thinking for you, it can help enable you and it can help kind of turn your thoughts into something tangible sort of like the game that Justin's friend created, but the ideas are really going to have to be be more fundamental but what generative AI does do is it democratizes this like I have an idea of a game. I don't, I don't have the time and I don't need to learn like the like 1000 different steps I need to take in order to produce something of quality. So what it does enable me to do is like have the idea and then be able to use a like, use like generative AI to be able to generate this idea and turn it into something that is usable and then there's lots of platforms that can help you publish that game and just make it more accessible and it just it's a lot faster and it's easier to test it's easier to pilot and as a result is easier to iterate and just make changes to it. I think that the, it's, I can give you a lot of stuff but it turns, you know, what's that what's that quote like movies are made on the cutting room floor, right the adding room floor. I think the same thing is, it's like doubly true in the world of AI where you can give unlimited generate generative, you know, capability and so it comes down to like, you know, what's your taste. Right, like what are you choosing to include the, the thing about the one thing I'll say with games in particular is that so much a game design is this idea of like game feel. Right, like when you play Mario like just the physics they just feel a certain way right like there's a there's a weightiness to it. And, you know, AI can help you program and get get a version of that going but like, does it feel right does it does it like hit that dopamine center that's a fundamentally human thing right, which is, it's good for our jobs as game creators I think for the time being and maybe eventually there's a, you know, AI can figure out what what what good game feel is to but I don't think we're there yet. And so I do think there's like that umami to it right that indescribable quality to it makes a good game, and makes something like the at least the interaction design feel good and so, you know, I think that that that ultimately, you know, comes down to, you know, are you are you thinking I think for the opportunity for New Zealand game studios is probably the same as for everybody everywhere in the world which is like every technology innovation cycle. There there's like a chance to reset and you know kind of do more with less and so you know you become I've become more competitive with like a greater set of people as you create more with like fewer resources, you know, and so I think the internet was like that for you know all sorts of developers of software and mobile apps and everything and the tools are just like their creative tools right they don't really. I think, you know, kind of like these other two are saying it's like they don't replace taste. And we already live in a world where there's like infinite content right like there's more games and you can play more YouTube videos and you can want more music things and so the problem is not like there's it's not I don't people concerned about like there's you know as going to flood out human creativity because there's like too much content I think that's like kind of fake you know that's not real like. I think that the cool thing about these tools are they allow people to be more creative and more expressive and up level their skill sets and create more, you know, do they maybe fill in the gaps where they don't have. They don't have all the skills and so it makes actually creation more accessible to a broader number of people. And so that both makes the market more competitive actually but it also makes it's an opportunity for people to who maybe we're otherwise able to compete to compete so like I think the opportunity for for startups and for New Zealand companies or game companies it's like that you can maybe like bring something better to market more quickly now you know. That's actually a plus one that and say that like for country like New Zealand that may not have, you know, access to talent like that company and like LA would have, for example, right for for creative talent or whatever. It actually levels the playing field. And so someone has a unique idea like that that like that I that concept of like the gap between idea and in market is just shrinking right. So it's the same what the same thing the internet did for ideas. Right or the app store did it's allowed you know flat people was massive it was some random dude in Vietnam who like came up with this idea and that game blew up. And so it's basically democratising the best ideas. And I think this is huge I think this is like actually huge for for for a country like New Zealand because it's just, you know, it enables. It's, I mean the point of EHF is like, you know, out is it's like New Zealand solutions at scale, like global scale right like am I capturing it the right way. I mean this this actually enables that right and it makes that much more practical pragmatic reality that you can that someone can do that. And I think at the end of the day, generative AI doesn't really replace like human creativity or storytelling right a good story is a good story and maybe it can help. Maybe we can produce something kind of generic, but it can't really I mean it's like the games that are really thoughtful games that have a like beginning and an end like all of those games take a lot of thought. And so at the end of the day comes down to who's going to put time into it and actual thought into it. I'm reminded of the in buck threes on the future so bright I got to wear shades. We've opened up this Pandora's box of content creation. And it's going to be less about creation and more about curation and filtering or sunglasses are going to be more important. Curious, Ameth, what are some of the ethical considerations when using AI and gaming, and how might you ensure responsible use. Yeah. So, you know, we employ artists, right, and and folks who have put a lot of work into their craft and to create, you know, create differentiate skill set. And these models are trained off of other people's art, right to the point where like, I think you can actually get the shutter stock, like, I mean there's a way to get the shutter stock like tag to show up if you like enter the right prompt in some of the in some models. So the question is, is like, you know, who's like yes these things are generating new art but it's actually built off the collective art of you know, millions of people right. So there's a few things that we think about one is do we go straight from AI to in game, or do we use AI as a as kind of a helper to help us do mock ups and prototype work and so on and then it's like, you know, there's there's a human kind of in the loop before it gets to our customers right. So that's been kind of like because there's there's like this idea of copyright, there's the idea of. Copyright just are you building off of something that you created and take credit for someone else's work and marketing someone else's work right and so. So our solution to that is like this doing that that may not be for everyone I think I'm sure that's the overtone window on this stuff is going to shift in the next few years and people are going to be like it does what it is. But like even just for like delivering a proper thing to our users like we want to say hey this is created by us and and it's okay for us to charge for it or whatever and and that's just like our company values. And so that's the first thing the second thing is just, you know, when we think about this from like a jobs perspective, right, like was was AI mean for folks who had, you know, like, like, like, like an art like concept artists like you know it's I think, you know the stuff's not quite there yet for like game design and some of those things but for like a concept artists you can argue that that is not a job anymore. Right, like AI is that this is a solution that like in fact it's going to be better at it. And so part of what we're also thinking through is like, how can we leverage, at least internally how can we leverage people skills in different ways or use AI to make them more effective at their, at their job in some way and and I'm actually not really super worried about my completely replacing people jobs I like, I think that, you know, like we've talked about this there's so it's like, there's so much still around taste and quality control and all these other things that like, I actually think it's going to end up being like, I like to think that like, you know, I use I use ideally I'm a programmer I use co pilot, you know, we're, we're writing, you know, and basically the way that I describe it for me at least and that, and that kind of in that environment is before I was on foot and now I have a bike. Right, it feels I'm four before I had a bike and I have an e bike, right, and it feels great. Like, I don't have to write as much, you know, like, I don't have to figure out syntax. Now, my, the way I program has changed, like I type a few characters and auto complete stuff for me it's amazing. It's like, it's unreal. And so I don't know if that's how artists feel when they're doing it but I like to think that that, you know, it's probably similar. So, I get to focus on the fun parts of my job which are the creative thinking and the high level architecture problems and so on. So actually it's made the work more more enriching. Yeah, I mean that's how we're tackling it but you know I'm sure it's different if like your company like scenario is generating art for other people and how you're training your models and then they have maybe more ethical considerations I'm curious just if you have any thoughts on this. I think there's, like, there's a probably short term tactical question or legal question which is like how do you do licensing like I think licensing will get figured out, because there are a huge repositories and user data, whether it's like the text textual on reddit or shutter stock for images or other, you know, image repositories that are, and what practically I think will happen is like the major, you know, creators of the models will do deals with these data sources and pay them and those data sources will maybe pay the people who are putting their content up there or maybe just make that part of their end user agreement and then those artists will like, you know, maybe they'll get compensated but maybe not and but it will be legal, I guess. And but I think like the, you know, the ethical question is maybe a little bit like more interesting. I mean there's going to be creative destruction of jobs for sure, like I think that, and sometimes people will like it and be empowered to do something different. Or sometimes they won't like there was that viral post about like this reddit artist right who's like, I can't remember what kind of artists but you're posting that like their job. They thought their job got a lot worse because they were before they were like getting to design the characters themselves and that that's the part they liked and then now they just like come up with a prompt and like put it, you know, I think they were concept artists and then they like generate the thing it takes like 10% of the time or 5% of the time but it was like they didn't feel like that was fulfilling to them anymore. And I think that's like realistically that's going to open to people but like, you know, generally through the creative destruction process like people get value in society and like more people are able to, you know, for every artist who had spent a lot of time developing a certain skill set that was difficult and well compensated. There's like nine other people who like don't have that skill set but would want to work in that industry that the creative tools are empowering them. So it's not like, I don't know what the balance of like ethics or fairness is but like that's just going to happen. And I think, you know, usually what happens is like there's more creation, like I would bet on like they're being greater creation and more choice in the art that you get to consume in games and that'll be the kind of, that'll be the ultimate outcome of it, you know. Yeah, I think, yeah, I mean it's interesting that Justin, I mean you're both talking about sort of the like legal implications of it and some of the copywriting plus job destruction or job creation type of things. I think one other angle that we haven't fully talked about is also what are these models producing like if you are putting in, if you put in a prompt saying hey I want a CEO right like is it giving you a middle-aged white man because that's what the data set is right so it's like where is the data coming from and what is it that it's producing and how much thought are we actually putting into something after it gets produced is also super important like does it mean most games are going to feature certain male white characters because that's what a lot of games feature right now and what do we do about that and how do we kind of help people think through that. And on the other hand, I mean maybe it's possible that it's easier to create characters that we haven't quite thought about or aren't as in the in the in the fray because we haven't programmed them as much and so there is a little bit of the representation angle also that we should be thinking about as we're thinking about ethics in gaming and AI. I think I think it'll be the opposite now you can have your whole world you can customize every world the force to expensive to create assets for everyone but now if I want to I'll live in an entirely Asian populated gaming world where there will be no people who don't look like me. I don't know if that's any better but that could be my choice, you know. I'm almost reminded of the promise of no man's sky and the idea of this vast, procedurally generated universe and it sounds like that would really be much more possible today than it used to be. And what I'm hearing from this Kerero is that AI is not yet a replacement for human discernment, judgment of taste. Yet AI will be likely disruptive for creators and you're already seeing signs of this. Pending a little bit. I'd like to take it back to Justin and shift some transition a little bit from the gaming industry to the world of startups. Justin I'm curious. The question that we asked Hamlet about gaming I'll ask this to view the startup ecosystem. How do you see AI changing the landscape of the startup ecosystem the next couple of years? I think it's probably like similar just like different applications where, you know, there's going to be all these tools that reduce the cost of like thought work, you know, like there's so all these different types of things that you do in your startup that take, you know, that you would have a person for, you know, you'd have to hire someone to do, you'll be able to do and like much cheaper, much more quickly. And so like one example is one of my friends is starting a company that does lead data, they're like adding data to leads like sale leads, right. So they, you know, previously you would need to have a someone do research and say like, you know, like basically this thing, you know, takes in all these data sources, you know, public company data or like the company's blogs and stuff like that. And then like you can ask it questions about like, you know, whether basically they're relevant to like you selling your product into that kind of organization, right. Like, do they have these certain kinds of problems, you know, things like that, or like do they, you know, what are the, like, for example, I mean it's kind of complicated but they're basically like, you know, if you're selling, you know, more extreme services to this company than other companies and you can like kind of read all the public files and see if they have those kinds of problems, right. And that's something that you would have to do the research for before but now like AI can do it in like an hour, right and like for your entire thousands of leads list, like very simply. And so, you know, that's it. I think there will be a lot of examples of things like that that just reduce the cost and so it enables like what used to be a 10 person company, you could do that with five people now or you know, and you can do much more quickly. And so, I think every part of the, like every department inside a company will have things like that that will like allow less fewer people to do more. And so, you know, it's probably that happened before, you know, multiple times in the history of the history of startups, right, like the internet and then mobile enabled this like much more broad distribution for companies that were much smaller, you know, and so that's cost to come down to like the cost of starting a company come down like AWS and all these like cloud computing services are similar like when I started 20 years ago we like racked our own servers and built our own server infrastructure and that was like a cost, you know, and then now and then you could just like spin up AWS for like $10 you can like spin up, you know, host or your, your web app or whatever. And I think this is another thing that will like increase speed market and reduce costs for companies. I don't think we're at the point yet where you can be like, give me a startup idea and then use auto GPT to like actually execute it or anything like that so you know, use a founder, the founders themselves are probably safe for now. Haven't commoditized founding founding things just yet, what for close maybe. I think one of the interesting things around just a generative AI landscape is I think where startups have a ton of potential is really around the verticalization of AI so it's really at the application layer where because some of these language models like AI is not new, it's been going on for a long time, but there are certain things that we can spend a lot more time on where you can personalize all the more stuff so for example and like health care like there's tons of different applications that you can build on top of a lot of these. A lot of the technologies that generative AI has enabled like whether it be recording audio during patient visits transcribing them quickly and actually doing it well, and then using that to actually make give more context and produce a more rich history or more rich health history that then goes with you from doctor to doctor or from just throughout your your life cycle. So there's a little bit of that that is getting built and there's just a ton of different things we can do around that. And then there's of course there's lots of different software sectors where we're seeing development whether it be in productivity whether it be of course in gaming, whether it be legal industrials finance so there's just a lot of different applications that exist and I think based on some of the language models that are built. There's a ton of potential and there are a lot of startups getting getting built in this space. Yeah, what one of the things that is kind of interesting though is like the flexibility of all this stuff. Right, like, I don't know how many like when open AI came out with plugins how many YC companies they ended up killing like half the batch. It's like some ridiculous number of YC companies that were in development were like well we guess we got a pivot now open AI just like implemented the purpose of our entire business. It's interesting, like, definitely like, you know, when cloud computing came out, like, being like a server hosting company became less lucrative but being a cloud server management company became way more lucrative. Right, like, you know, at Zingo, we were like one AWS's biggest customers. In fact, you know, it was like us and Netflix were like two biggest AWS customers back like, you know, this like 10 years ago, but we didn't use AWS directly. We use something called right scale, which was sitting on top of AWS and allowing us to do essentially a bunch of like higher order functions. So what's what's interesting is I think like, you know, we're in this conversation right now, and anyone who's listening to this, you're listening to us because you're curious about AI and you're looking for like, you know, help on like maybe how to get started or business ideas or whatever. It turns out that building companies that help people do what this talk is trying to do is actually valuable at this moment, right, because a bunch of people they they see the potential they keep hearing about it. And, you know, I'm in blockchain and there was a bunch of there was a hype cycle on this before, and you get all these like fake influencers who were trying to like help people get into blockchain. But like, you know, in AI, it's like a real thing. It's complicated. It's like hard technology. You can go and use it, but then like how to go from this like using chat GPT to like creating, you know, like actually usable things with this. I think people need help with that. And then when you start to look at it from enterprise level, you even need more help with that. And so I'll make a strong statement and it will be on this call will be recorded forever, but I actually think that open AI doesn't have a defensible mode as a company, at least in the consumer side. And I so I think what you're going to see is that there's going to be more and more of these these language models getting created and. And so for them, I think their business is almost going to be entirely focused on serving the enterprise. And I think, you know, if there's one common theme is that AI changes literally every part of a business right it touches everything. And knowing that that it's a huge amount of opportunity to help companies and one or two specific niches in the enterprise, right? How do you apply AI to, you know, generate your marketing copy? How do you use it to generate your sales leads? How do you use it to, you know, generate game assets? Right. There's like a very clear thing which is like maybe the AI model is commoditized, but the interface to that model isn't right. You know, like if you look at there's this old graphic for Craigslist where there was like, you know, a dozen links to Craigslist. It was like, you know, buy and selling and like renting apartments. And then someone went and like kind of connected each of those links to a different startup that took that piece of Craigslist and like, you know, created an entire company based around it. So for like, you know, renting rooms that was Airbnb for buying and selling that was eBay. So the question is, is like for all the things you can do in chat GPT today, how do you create an app that's a specific version of that thing that makes it easier to like do the prompt engineering under the covers? I mean, scenarios doing this, all they're doing is like taking whatever, you know, inputs are giving it and then they're constructing a prompt and helping you kind of refine that prompt to get the game asset you want. And so there's just a huge new number of, you know, startup opportunities have been created by this because the type of application you need to build to use AI is fundamentally different than what's come before. So I think it's a pretty cool time to be an entrepreneur, especially if you don't have an idea yet. Some really, really cool stuff there. Kind of agree, but I think, you know, like, the question is, like, do these LLMs become good enough or like, do they approach, do they accelerate faster than you, these like kind of individual use cases? Like, do we approach AGI and you can just kind of ask it anything and ask it to do anything? Or do the enterprise specific use cases? Do they build enough kind of like of a mode around their like specific use case? I don't really know what's going to happen. I think though I kind of agree with you that like all these startups building AI, I think I actually think it's like not a great time to be an entrepreneur necessarily. I think it's hard to figure out an idea that's like that has long term defensibility. I think the whole creation of these AI features is like pretty deflationary. Basically, it's a cost that like, you know, kind of like what I mean is like Bing ads, you know, barred or sorry, no, the Bing ads like Bing AI or whatever, and then Google will add barred and like they're now they're kind of both forced to provide these customers, their customers this AI features, and the real winner is like the people lower down in the AI in the stack, you know, like Nvidia or whatever he's telling them the GPUs. And I don't know that they're like making any more money because they have these AI features and I think that's going to happen across the stack like Intercom, which was this like, you know, like Intercom, the customer support interface, they like very quickly added these like AI summarisation and like rewriting for their agents. And I was like, oh, that's pretty cool, but I'm sure there are like, you know, there are other stars that were like, we're going to be AI, I'm going to allow agents to like, you know, replace themselves with AI or whatever, but like Intercom can handle features faster than you can build the customer base and the distribution. I think it's really hard to think of like I've been trying to think of like what's an AI startup that I would actually want to start and I think it's really hard to think of something where the incumbents are not able to, you know, very quickly add AI features. I think it's a lot different than like, you know, 15 years ago when like mobile was starting and like nobody had a mobile app. Now there's a lot of tech companies with distribution to customers and they can like very quickly add these AI, the AI APIs are available to them as well, you know. And so, you know, I think it's got to be something where you are because of AI, it's either like a second order company, like something the world's going to change in some way because of AI and then like what does that enable or alternatively it's something where you can like aggregate a new customer base because of AI, you know, and I don't really have any good ideas but that's how I think about it. I do think there are two interesting places in the sense that I think the place where AI can produce a lot of value is or where you can create a startup that does have some form of defensible mode is through partnerships and so I think some of the stuff that we see with some of the larger companies partnering with some of these startups I think there is some value to that and I feel like if you end up embedding yourself enough within a customer's ecosystem like if you become sort of not the AGI part but if you are sort of more context based and you sit within an organization and you're able to kind of read all of the information within that organization and produce value in that specific large organization I think there's a lot of value there, which is pretty interesting. I don't know who builds it and I don't know how you actually get it done and actually like tying all of the threads of the super large very complicated organizations but I can see a lot of really interesting value from being able to do some of that knowledge management and being able to sort of more specifically create specific AI use cases within context. Very robust. Thank you. Thank you. I have a couple of questions last and then we'll turn it over to the audience. Just that I'm curious, you know, we've been talking about how companies can use generative AI to accelerate and augment their capabilities and spend some discussion around differentiation. I'm curious to hear some real world examples. Can you share some examples of how your investments have utilized AI to drive growth and innovation? I think I probably covered my best ones, you know, like with scenario and like this kind of example of this like lead school data thought up. I think there's, you know, kind of reaction when I say I think there's probably like every slice of every department in a company like if you do anything that requires kind of like basic, you know, repeatable thought applied to it to like get the job done like eventually there'll be an AI tool that does that, you know. Great. And yeah, I would describe it. I mean the way we describe internally is everyone has an intern that's always working 24 seven, but they're like clearly an intern. And so you have to like delegate but verify. I love that. I love that. Fiona, I'd like to leverage some of your considerable experience and slightly pivot a little bit. How do you think AI is going to change the landscape of public companies and public company mergers and acquisitions. Yeah, all views here expressed are my own do not belong to any company I work with but I think from a public company perspective and you see this are this like AI arms race right it's like obviously there's sort of some of the stuff going on within where there's a ton of like who's going to build the better search engine based on based on AI utilizing AI who's going to make it available to more people who's going to make it more accessible. And then how do you actually make money from it and our ads the answer to continue to make money from it by by providing this better search service. So there's that but I think when you just think about the various the various tech giants, they're they're spending a lot of time just on key partnerships right so there's a lot of end user devices that can utilize some of these partnerships and make them better so some of these home devices that can talk to you it's like those can get better and so there's there are certain things that the monetization of it is still under question but I think it's I mean there's a public companies that are building out sort of the like user facing stack that are spending a lot of time on figuring out where they can, where they can optimize what they're already doing through AI. And then there's at the bottom of the stack there's of course the companies that are based on compute and so the companies that are building the chips that are going to enable enough compute power to be able to do AI faster and do it cheaper. So if you look at it's pretty it's pretty expensive right it's like chips are not necessarily built to generate the amount of data that we're we're generating right now through AI. And so, as we start to optimize chips as we start to optimize some of the compute power, it just makes it a lot. It creates this foundation that allows more AI and more data to be able to generate it on top of it, and used on top of it is going to be different types of public companies that are going to be at different parts of the of the ladder when it comes to AI, both from a data perspective and from a powering the data perspective. Excellent, thank you. I'd like to turn it over to our audience and audience I invite you to raise your hand just click this reactions button at the bottom and raise your hand. Thank you to ask your question. Michelle. Yeah just what others are thinking of a question. During just if you can pepper through and maybe put them into the chat, any resources that will sort of like best reads kind of thing so that because there's obviously so much information out there. What are some sites or books or think areas or companies that we should be following or watching that are really good in the space and utilizing the space just so that we're honing in our thinking and our reading. I have a great recommendation for this. I keep up to date on the stuff. It's called Ben's bites have any of you heard of this. It's literally like you get an email every day, which is just like a summary of a summary of all of the like kind of AI happenings. I think I think I missed on the free one but yeah, this is like it's the reality is it's moving so fast that you might as well just read whatever is just the newest information because who knows like if you go back and read something from three months probably outdated already. Yes, the Ben's bite stuff is pretty good. And they have a like a good mix of like technical stuff some like regulatory stuff like the one yesterday it was about Sam Alton and like going to talk to Congress. You know some applications that are interesting so it's it's it's a good mix of stuff I'd highly recommend this. Just signed up. I think I heard about it before but I didn't I didn't actually put my email address so now you just tip me over. The problem is I'll have no takes for things like this if everyone's reading if everyone has the same upstream sources just got a new first you on new first and just say it and then you know you look at the looks more like jeopardy if the like buzz in we've got a question and chat from our random whale and then James and I'll get to you. So this from random whale. Do you think there's sense in provenance for specifically trained or biased AI models, especially since homomorphic encryption and ZK is coming to live on chain. So you can hide and pay while your AI models on chain. For instance Grime shared her approval to use her voice to generate AI songs. Wouldn't it be cool to have the model for her voice deployed on chain. And from there, it will be verifiable provenance, programmable royalties for artists, derivative works, etc. Go ahead. You want me to take this one all right. I mean I think this is where we're going right. You know the blockchain side of it aside I do think like you know there's two there's two sides to this there's the. What does it mean to be an artist anymore right like what does it mean to be a likeness. And so there is like we do have in the legal world a set of like kind of laws around this already this idea of like likeness rights and all of other stuff and our legal framework is actually pretty well equipped to deal with those things. So yes you can use AI to generate a song using Grimes or Kanye after those that don't know there's this like, there's like a bunch of songs that came out recently for like the weekend and Drake and they were like good songs, right and like. So then Grimes came out and she was like hey like you know what like, use, feel free to use my my likeness to generate AI work which is actually a super. I think it's like almost like a jiu jitsu move of this entire thing. And so, but in terms of monetizing those things. You know the copyright and likeness laws still apply. And so, you know, and just maybe have an opinion here as a as a creator, I think you probably deal with some of the stuff but like, but ultimately like, I think that this is like, if I were an artist that had a likeness that was worth monetizing. I would absolutely go do with Grimes just did because it's it's it's basically all you're doing is you're getting leverage on your on your existence. Right. Now, if there's 1000 Grimes songs and one of them blows up that's like really good for you as Grimes. So I think it's like super smart to do this. And then the on chain stuff like you know there's a there's a company called audience, which I think you know Justin you may you're like someone involved with it's a story by my living room. Yeah, right and they're basically doing programmable on chain royalties and like they're doing this for music and I mean plugging this in would make a lot of sense. Once it once like likeness starts to become like something that you can monetize more efficiently to react. I think the difficulty I mean I'm you know audiences really cool, cool technology and their idea was like to create this programmable world where you kind of automatically get paid if you contribute music into this platform. And it handles all the royalties and stuff like that but the problem is like the alternative music distributions is or the main music kind of distribution system is set up in a way that they don't really want transparency to the thing like that you have these intermediary labels are collecting all the, you know, like royalties and think rights and plays from the, from the major like streaming platforms and they're not really providing transparency back to the artists and it's pretty hard to break that monopoly. I think I'm not sure if like this will practically you know the AI generative AI will be the thing that like changes that, to be honest, maybe that's depressing. And like, it's, it's pretty cool that you can like make a Kanye song, you know, I think a lot of that will be like promotion like right now what the first things that will happen are people will use this to create like, you know, like David get a made like a m&m vocal tag and like they made an ID for like a set, you know, and it was like a viral moment on his Instagram, but he's not going to monetize that song, you know, James and you had a question. Yeah, just posted in the chat chat. Do the panelists see any likely convergence between advances in the broadening consumer use of AI and the trend of capturing value from digital exhaust and user thought or action influence by surveillance capitalism. Any impacts on personal privacy enterprise economics wealth disparity. And when that comes to mind is I think there will be better and better propaganda that is like, you know, you will be able to tailor the messaging, you know, the or saw that kind of like starting in 2016 where there's these armies of bots that are trying to like influence what people think on Twitter by just posting, you know, extreme views or like supporting some candidates or whatever. And then I think, you know, that will like explode, the people will create more and more realistic like chat GBT four is like, you know, as smart as people on the internet, or more smarter. And so, you know, I think there will be more and more tailored messaging to that is like tailored to individual people's psychographic profiles, people will collect. You know, you know, you know, you can create like a whole multimedia landscape to like drive home certain messages whether that texts and tweets and all the way to videos, you know, like that show up in your tiktok feed, you know, and so I think that's like inevitable, I guess, and like kind of, you know, that's a that's a huge problem. And even if we like legislate that in the US or the West, there's like other places on the internet or like other, you know, foreign governments who that will not be subject to those, you know, and I think, well, in that restrictions and we're not like super equipped. I don't think we're even really equipped to deal with like the kind of like ongoing sigh that foreign governments are running on Americans on like through our own like public internet and so I think that like this is the next level and we have like no we have no capability to like to do anything about it. Yeah, I think that's why you should move to New Zealand. I mean, while I think the other, the other side of that to some extent is thinking about education and thinking about how we're utilizing it like just AI in education and how we're thinking about an education and how we're prompting critical thinking so there's I mean one of the biggest use cases of AI so far has been in education right like people writing papers writing poetry writing a lot of things using AI. And so some of it is like, how do we, how do we work with our education systems to help people think critically help people think about quality and help people distinguish between analysis versus description, which is a lot of what AI works on. And so I think there is, I mean it's like, like any other other time in history as stuff gets more personalized, and as stuff is able to kind of be directed at you in a certain way to pursue it in certain ways some of what Justin is talking about. Say how do we also empower people to think critically and think about what they're reading on the internet and how that's actually impacting their life. I think that's, that's another great point it's like, you know, AI has the potential, it's already being used right like we're using it models and like social media sites to customize what you see based on your revealed preferences, like what you click on and watch right. And that's so we're all, but we have now the capability of creating like models and consumer tools that, you know, do that in all different areas of your life, including as I think education is a really great one because you know you have this ability now to create customized education tools for people, I think a much more tutor to people much more interesting degree than they have been in the past. And so it's not all negative, you know, there's like a lot of opportunity like up level humanity with AI tools. Like the internet I think it's like the internet might have been is it was a good or bad like there's probably some pretty bad consequences and some pretty good consequences. Thank you. Ian, would you like to jump in? Yeah, I'm just checking you can hear my audio. Just great. Thank you. Awesome. Hi, I'm Sung Jae. I work as a project manager in health New Zealand. So it's the public health care system in New Zealand. I'm really interested because it seems like health care is a very viable place for innovation gamification and AI, especially because there's limited public funding constant loss and poor retention of staffing, especially nursing resource and we're losing a lot of resources. But that's in theory and I guess a lot of the. Difficulty is the ethics, the AI governance that's being set up. The data governance health privacy or data quality in the system and then the old and divided tech infrastructure around the country for health care system and whether the audience is ready for innovation. So I guess my question to the panel and everyone else is what would you do if you are the mediator in the public care system? What are the applications that just make sense to bring in that are ready ready? And then as a potential founder, what are some of the ways to mitigate through a health care landscape from your expertise? I can share this. I can share an anecdote here just because you know it's so there's these stories about AI correctly diagnosing like like rare conditions. So I minimum just like it's an extra pair of eyes, right? But I think the whole like keep a human in the loop thing makes a lot of sense. It allows you to get that extra pair of eyes without having to potentially upend the entire system because I think going from. Okay, like you have a radiologist like looking at your at your x-ray going from like someone just you know today, which is someone goes in there, they take a look at it, maybe miss something. You know it's easy, you know humans are fallible to uploading the x-ray to a machine that then gives you your diagnosis. I mean that's that's two ends of a spectrum, right? And so I think it's important to remember this thing is a spectrum of how much you want to integrate. And I think it's actually it'd be like I'd love to hear a counterargument to this, but I think it'd be foolish not to use AI today to actually add as an extra pair of eyes, right? Like, I think the danger there is the same thing you see with self-driving cars, which is people just get like to rely on the system and then are not awake at the wheel. And when something goes wrong, they take control back. So I think it's important to like, you know, put maybe checks of analysis in of like, actually like, okay, did you view this yourself? Check, did you have the AI look at it? Also check, and then you kind of go through like that and ensure that there's always a human in the loop. But yeah, I mean, I'm excited for that. I think it'll bring down the cost of care. It'll make it more likely that rare conditions can diagnosed. I think it's just huge amounts of upside for this. And that's not even getting into like research, right? Like the ability for AI to help us develop new vaccines to and this is the utopia argument for AI is that like, you know, if we actually managed to pull it off in a way that doesn't end up being, you know, complete disastrous for humanity that, you know, they'll potentially solve a bunch of our problems starting with, you know, discovering new drugs, helping us cure rare diseases, just helping us scale healthcare to everyone, that kind of stuff. There's also a lot of really interesting personalization that can happen right with AI and healthcare. And I think that's super expensive to do right now. It's like right now you need to have a concierge doctor pay $40,000 copay per month to get that level of service. And if we're able to make that more accessible, it's like we're able to, I mean, we have like, we have the genetics, we can kind of predict a lot of things about people. We can predict a lot of things based on history as well and based on your genetics. And so if we're able to, I mean, we have enough data at this point to be able to actually really effectively predict certain things. It's like, but the hard part is how do you take that from the lab and from like a research model of like where like our random startup is working on it to actually being able to integrate it within a healthcare system that is fairly antiquated to some extent, right? It's like we're in a really interesting conversation yesterday with someone who is sort of the like godfather of like genomics to some extent, and we were chatting about how essentially it's like right now our system is built at being able to diagnose disease, but it doesn't really optimize for health. So it's like how do you optimize human performance rather than, oh, somebody got sick, they're going to go to the doctor, we're going to prescribe you medication that we're going to see what happens. But how do you, if there are ways to ensure that people don't get sick in the first place and are able to expand lifespan and expand sort of the ability to live a good healthy life. What are the things that we need to put into that? And I feel like AI is a really valuable way that we can actually really customize a lot of this data and understand people more deeply. I guess I actually am like a little bit of a probably a bear case on both of those thoughts, which is like, I don't know if you're like I'm kind of negative about AI, but I think that like it's not going to reduce healthcare costs because most of the cost is not a doctor, at least in the American healthcare system. You know what I know and most of it is like cost disease from having this like ballooning infrastructures and the administrative overhead and then also the fact that people are living longer and they're old. And you know, that's the number. Yeah, but for functioning democracies with universal healthcare, it sounds like you can actually get much, increase the average floor quality of healthcare in those other countries. I don't know, maybe so. I don't know what the cost disease in healthcare looks like for in New Zealand. And then I forgot what I was going to say about Rihanna's point, so never mind. You know, I don't know, we'll see what happens. Oh, the other thing I was going to say, sorry, what you're saying is like people need to get more personalized healthcare, which I think is true. But I'm a little bit like skeptical on whether personalization is the problem in healthcare. I think for most people will power. We know what like most people that there's like certain types of conditions that are very, very prevalent in the West and you know that are the major problems that people have like cardiovascular disease, diabetes, and we know what causes like their lifestyle, you know, there's a lot of like lifestyle things that you don't, I'm not, I guess, and I'm not a doctor so I, you know what do I know but like it just seems to me that like mostly people. Many people know what they need to do, but it's a willpower problem. It's not like a personalization problem. I don't know if they maybe if they have a better AI like doctor that will help them. I don't know, maybe Rihanna convinced me otherwise. I mean, I think some of what you're also talking about is, I don't know as much as it's a willpower problem as much as it's an infrastructure problem so like if we talk about obesity for example right like if you live in a place where you don't have a grocery store and you don't have access to fresh produce like it's not like you don't have the willpower to eat well. It's that you don't have access to stuff that isn't dessert being considered a snack or lunch or dinner. And so I think it's a that's like a greater like democracy or like a greater like America problem to some extent. But I do think like I feel like there's almost like various segments of the population right like there's, there's segments of the population that need more like basic shit like access to produce and there there's segments of the population that already have access to a lot of that stuff but it's like if there are certain things in your lifestyle that make you more likely to get diabetes or cancer and if there are ways to alert you earlier on like maybe there is a part of the population that does have the power to take that to take that indication and actually make changes to their lifestyle. Like, do I want to be motivated to exercise more like maybe there's a way to motivate me right I don't think it works for everybody and I don't think it solves the like fundamental problem of infrastructure, but I think at the same time, it is more, I mean I'm more hopeful for it than I am bearish on it. I think there's a bit of both I think one is the will power aspect by infrastructure if something is already near you and that's a good option by design, then you don't need as much will power to do the right thing. So, in New Zealand what happens is a lot of the new fast food chains come into areas of New Zealand where there's already pre-existing obesity problems, and that in itself by design increases the risks in that community so I think there's by design. And I think there's definitely a power of how can we make will power more accessible more easier to actually initiate an action and maybe gamification is the next way to utilize the tech savvy new generation of workforce. So, yeah, I think it's all interesting. Thank you. Thank you very much. So, this question comes from Murdoch. What strategies could gaming companies use to combat cheating and games, for instance, aim bots for multiplayer first combat or first person combat games such as Call of Duty, or AI learning to mimic pro gamers, and then being exploited in competitions where people can enter their AI doppelganger into competitive tournaments. I'll take this one. We saw this play out with chess, right, where, you know, years ago deep mind started to beat human players and if you go on chess.com there's like, you know, potentially a playing as boss potentially. So, I think there's two ways to approach it. One is that obviously for, you know, for high level competitions, you, it actually brings people to have those competitions in the real world, right? Like if you have whatever the world championships of League of Legends or something like, you know, you actually do that in real world you have ads verified for online competitions. It's a little bit more tricky and I think AI is also the solve this problem a little bit though that hasn't, I will caveat this was saying that has not yet been true. So, you know, we're seeing this in, in for example, in people doing like, like written content for like, for example, I'm writing an essay from my college class. And there's something called zero GP GP T zero, which is supposedly used to detect when something is written by an AI, but it doesn't work. And so, so, but the theory is that AI can use the tech AI, and it's going to be like a cat and mouse game on the stuff. You know, in terms of, you know, a learning to make pro gamers, like actually you lean into it, you actually have competitions, which are, you know, it's like, yes, we're all going to enter our battlebots into the competition and let them go at it and see who can use the best AI coder I think that's a different type of game. Right. But yeah, in terms of combating cheating. I think the real the mitigating factors that are using it to help protect it, and then also just doing something person. Well, it's, that's where my head's currently at on it and then leaning into it when it makes sense, you know, to actually, I think they're doing this right now with chess to they have like AI, AI focused competitions to see even build the best chess chess engine and so on. And then humans use that to skill up. It's actually pretty cool. It's like you have a training partner to play with. We used to watch the gladiators and now we watch battlebots. Justin, would you like to chime in for Rihanna? You know, I think I'm kind of covered, covered it. Yeah, I don't know if I have anything intelligent to add. The floor is free. The microphone is free. What questions do you have? We could go back to the second question before which was the kind of dystopian or the panic convergence between consumer use of AI and the trend of capturing value from digital exhaust and user thought action influences. I thought that was kind of interesting. We didn't talk about like wealth disparity or personal privacy or enterprise economics. I thought that was an interesting question. Let's go for it. What would you like to kick us off with Justin? What would you like to explore there? I don't know. I don't know if I have a fall off with that. What are the things that are going to happen? What are the impacts of AI going to be on privacy? And we've got a tease from random whale. You have even more dystopian questions. No, I don't want to hear that. I don't want to hear more dystopian. That's the best kind. Yeah, exactly. Yeah, I want it. Yeah. Yeah, I think there's going to be more, there are, you know, I think there will be. AI is like a critical kind of domestic capability, you know, like, and I think it's. We're fortunate in America. Anyways, let me have like the lead on AI and kind of these large language models and different kinds of models. And I think it's like we should be, you know, it's like it's, it's kind of important in creating our own ensuring domestic security. You know, so I feel like it's kind of like a huge frontier. I think that there's like, there's kind of like the obvious applications, which are maybe more like the types of stuff that Palantir is doing or like, you know, like information, like a processing of large, large amounts of information. But I think the one that I'm more interested in is like our national security around like, particularly around, you know, how like we enable foreign governments to communicate directly into the minds of Americans like at scale with like customized messaging. I think that's like the thing that we're like the things that are like more like traditional theaters of combat. I think like America, like America's military system is like pretty good at like under, you know, seeing what's coming and upgrading our capabilities. And that I think is like a new fear that we don't like really like know what to do about, you know, because like freedom of communication is like kind of a cornerstone of our democracy. And so like we, you know, it's hard to like, you know, the lines between like, I mean, when you're online, like, like it's like really hard to tell whether people are just talking, you know, there's like random Americans talking like saying stuff, or if it's like foreign agents boosting that message messaging, you know. I think that's the biggest, you know, to me that's like one of the biggest questions and it imperils our democracy to allow, you know, kind of like foreign influence over like what people think I think that's like, I think that's pretty, you know, that's like our, it's an existential question in my mind. I think I'll go one, I'll go let's go pragmatic on that kind of bring make it real today. It's possible to impersonate people's voices, right, potentially they're, you know, potentially videos of them. Like the amount of spearfishing that's going to happen in the next like few years is like insane to think about right like, so you get a call it sounds like you know your sister or mother or daughter or whatever right and like, you know, your family member and like, and, you know, hey I'm in trouble I need X. Like, how are you going to differentiate that actually texted my, my family I was like you ever going to call for me, like, demanding something, hang up and call me back. Right. Call me directly because it's it's AI has gotten to the point where you can basically impersonate people and. You know, even if it's not happening to you directly like someone's calling up your phone company and and and doing the same thing. I don't know if like seems like face ID is going to get cracked soon. Right, I can't I can't imagine that you know the Apple stuff is actually better than whatever it is able to generate. And so, so I think like the, and there's two things it's the you know personalization yes and all the things amazing. It also means that like you being exploited can also be personalized. Right like someone can go figure out exactly you know kind of riffing off of Justin said around like change the way the hearts and minds people like someone go figure exactly you know what you've been tweeting what you've been reading and essentially customized argument specifically direct towards you to intercept you with an idea. Now you know it's, I think it's more farfetch than some of the other things I mentioned but like, it's it's it's there right and so. And so I, you know, I think that dystopian stuff is. I don't think that's farfetched. People believe some dumb shit, people believe really insane ideas like QAnon and stuff like that in America and that's not even personalized to people. Yeah, yeah. You should, everybody should create a code word like with their family, you know, minus purple. So if you ever try to improve, you know, you can call my parents and you say purple they're going to send you a Western Union. Immediately. Question though that. Yeah, yeah, just going to segue into that we've got a couple minutes left and to flip the coin. What anti fragile alternatives to dystopian directions might be possible. I mean, I think that's like the question of the world, or one of them, which is like, because I think there's a lot of dystopian directions of AI, you know you see foreign governments using it to master veil their like score them other things that are like very 1984 and like terrible. And I think, you know, the antidote to that is like, unfortunately, like belief in liberalism everywhere is falling like in America too. And, you know, we need to like reground ourselves in the values of liberty. Like that's important. If you like living in places where that, you know, the consequences of that are, you know, free thought. And I think that's like super important and we need to like, you know, like that used to be really important to Americans and people in the West and I think that's like, it needs to be like re-encultured, you know. And that's to me, that's that's the antidote. Like it's kind of like everybody said duty to propagate and protect those values. I think some of our dystopian concerns. Sorry, I have like a gardener is literally right outside my door. I don't know if you can hear me, but I think I mean some of these concerns have always been around like with every new technology with every new technology frontier. There are things that are good and things that are not. And I think all we can really do is build and create and see what happens and think critical thinking is helpful and crucial and being able to fix our educational systems is super important. But I don't think we can slow down technological progress or technological direction because we're worried about all of the bad things that can happen. That's certainly the approach we're taking now. It's just like blind optimism, full force, lean downhill, go into it, right? And the unfortunate thing is because of the prisoners to love on this stuff, this that's like kind of the only real option, right? Like if we don't do it, then it's going to be a, you know, some other organization or state and potentially one that's got less of, you know, more of this is tilt towards making a positive outcome. So, you know, when we think about like, you know, what's the like, what's the antidote to this stuff? I think that when we think about like the risks of AI, it's really about like timing, right? So like with fake news and with like social media, like, you know, we as a culture kind of learned how to like start, you know, not like I don't know if you guys remember like when the internet first came out, like someone would like email you and be like, hey, can I get your credit card number? And then people would like respond to that. They would like give you the credit card number because they're like the idea of safety wasn't like a thing. People believed what they read. And over time, people got more and more skeptical fishing attacks became harder and harder and harder and social media came out and then we had this idea of fake news and this idea that your feed was like kind of generated to specific to you. And so like my belief is that like, you know, what we're eventually going to need is more like automated tools that essentially help you disseminate between truth and non-truth. You know, the stuff that Twitter is doing right now with like the notes. If you guys see these things, they show up on tweets, which is like below it's like, you know, the context. And stuff super important needs to be auto generated. Right. And so I think we'll start to see things like that where like information is not just parents, not just the information itself, it's paired with a lot more context. And I think that's potentially one way to at least slowing this down. But again, it's the arms race, right. Like, can the ability to generate, you know, the dystopian version future be it's it's like the cheating bot. Right. Is there any such thing as it's like that but like with our lives instead. So same problem. Nice. That's amazing. Thank you. What I'm loving about this conversation is that it's actually starting to morph into further into the series that we're going to be running. So we're actually going to be running more conversations where this has started to go. We just wanted to start it with like we have today gaming, AI and gaming and about into startups. But we've actually got a lot of fellows that are and we actually got one on the call actually in regulation around AI. So that's an area that we're more fun to but the next session that we're going to run. I'm just throwing it in there is in June, June the, what have I got there June the 7th. And that's kind of about the impact of AI in New Zealand. So that's sort of the human risk side of it. And Chris will be moderating that panel with some more about fellows but right now I just want to thank Justin, Rihanna and Ameth for an amazing conversation. It's been really cool. And Chris, great moderation. So if you didn't get that, that's Ben's bites is a really cool sort of link to have. But otherwise, thank you everyone for tuning up and hopefully see you on June the 7th for the next conversation. Ka kia ti ano. Goodbye. Thank you. See you all. Thank you.