NEST Roundtable 17: AI in GameFi

Participants:

Host:

Woody, NEST Protocol

Noah, WhaleCoinTalk

Guests:

Alice, Cofounder of SecondLive

Mark, CMO of RACA

Ray, Cofounder of YSight Capital

Pedro, CEO of Mouse Haunt

Frank, Builder at Desider Core

Grant, Head of Live-Ops at Yeeha Games

Alper, NFTb

Conference recording:

https://twitter.com/i/spaces/1lPKqBlekOnGb?s=20

Introduction

Mark: RACA is a metaverse crypto coin, and it’s one of the largest metaverses on BNB Chain. And we’re working on just providing a 3D environment for users to interact with each other.

Alice: SecondLive is also a very big metaverse on BNB Chain. We have users of around 1.5 million already. We build lots of 3D spaces and we developed lots of fun functions such as dancing.

Pedro: We have currently two games released under the Mouse Haunt IP, which is a Free to Play PVP game with unique mechanics, from battle arenas to fun party games.

Grant: Yeeha Games team is just a team of video game enthusiasts, and we just want to bring some more enthusiasts and really fun gaming experiences to web3, consisting of a lot of professionals from both that traditional gaming and the web 3 industry.

Frank: Desider is also a game, and we want to create something new and found in the Web3 area. Our game was played by 300 million people. We are very passionate about the new gaming area, and we want to bring virtual games into the real world.

Ray: YSight Capital is a Hong Kong and Singapore based token fund and incubator as well. We focus on web3 and GameFi, so we favor the ability of entrepreneurs to provide better solutions in the field and break the tradition.

Discussion

Can someone share us what is the current AI use cases in GameFi industry and how do you think of it?

Pedro: The copyright is the major point when we talk about AI-generated images, because the AI just searches the internet and glues the images, then you lose the copyright. So I think that’s a point that it needs to be solved at some point.

Mark: I’m sure you guys have seen those videos on like tik tok or YouTube where you ask an AI to draw something and it and it draws a pretty good version of what you ask it to do. AI is pretty transformative work. I think bots and fortnite where it’s like artificial intelligence that actually plays the game.

Noah: NPCs are part of the game, and I can tell if something’s a bot. For example, their patterns are very repetitive. But a bot powered with AI has something entirely different. It's behavior like a human. With AI where you won’t really be able to distinguish the difference between whether it’s a human or an AI bot playing. I never thought about the idea of like let’s say NPC’s powered by AI, this static part of the game act differently based on the player that’s interfacing with them.

Grant: Our team kind of came up with an AI champion or AI bot that was actually able to play essentially like perfectly and play to a point where there was no way we could defeat this bot. That was a pretty eye-opening thing for me. I think a big part of what AI provides is actually a little bit similar to what blockchain can provide to gaming in terms of immersion and incentive. We have experience before trying to use certain API integrations into NPCs to enhance the experience.

Alice: AI can be used as an assistant for NPCs and can help in PVP environments. In the pure metaverse, we do want AI for NPCs more. Let’s say we have a conference; we need NPCs to do an introduction, so AI would be very useful for now. In the future, we could explore more scenarios.

What do you think the current use cases in GameFi and how to think about the whole industry and what will be the focus of investors in this game industry in the next year?

Ray: Before we’re talking about the long-term trend of GameFi, we get into some basic things about machine learning or AI, so I have to mention that reinforcement learning because reinforcement learning is a subfield of machine learning that’s study how AI agents should take actions into an environment in order to optimize the concept of the cumulative rewards and efficiency.

Woody: How startups could use AI? Will this bring pressure to traditional games?

Ray: The tech giants have very strong R&D power and they got talents, but I don’t think they got enough qualifications for developing a brand-new popular game with only the technique things.

Frank: There’s less room for players to improve in traditional games, but AI could help you to suppress the real people playing. So this could improve the experience. There is a very strange phenomenon that all of the creatives are created by the team first. So they have rules etc. But in the later stage, maybe if you cannot beat NPCs, you can incorporate them or some ways. What I observe is that at the initial stage, creativity is still the major thing.

Who is the biggest player in the GameFi industry?

Ray: There are a lot of big guys coming in. They got a lot of dollars. Both in US and in China, they’ve got at least several hundreds of millions of dollars.

Alice: Do you guys ever think about maybe someday AI can help to design the game itself?

Ray: Quite like the AIGC.

Noah: I think it would eventually happen. It’s inevitable for AI to be able to write these scripts and be able to write out code and be able to basically build a game from the ground up. I don’t know if that’s going to happen anytime soon, but I do think it would happen. There are many people who have creative ideas in their heads but may not have access to funds or developers. It’s going to lower the barrier to entry for so many people.

Mark: Integrating AI into blockchain games is harder than normal games, maybe a few years behind.

Pedro: I don’t agree that much. I think the blockchain doesn’t necessarily make things harder. Blockchain actually solves a lot of problems for us. You just integrated with the game engine and just connect the player wallets and their NFTs to the assets in the game. It would be really cool if we start seeing some smart contracts that have some sort of algorithm and that this algorithm is AI and it can be an upgradable contract that can be rewritten and upgraded and evolved and the players can interact with that.

Grant: From a game developer’s point of view, a big goal is actually to only implement it where it makes sense and only implements it when it actually is beneficial and helpful to the game experience. Also, I kind of feel like a lot of ad companies, like Facebook, and Google, actually might even know me better than I know myself because of the machine learning algorithm. This could help in games.

Josh: Even if you opt-out twitter won’t share your data with any third parties to get like a targeted ad to you, but what they’ll do on their own platform is they’ll find other users that are really similar to you. AI invariably gets superhuman, it gets unbearable. It just does things you can’t even imagine but it’ll be interesting to see it teach people.

Woody: The professional chess player all train themselves by using AI to know if the actions they do is the same as AI to make a higher chance to win.

What are the challenges of using AI in blockchain?

Josh: Contents is now queen and data is king. AI needs data for training.

Ray: AI algorithms can focus the consequences of gamer decisions and things like weather and emotions to comfort in game complexities. The ultimate team in FIFA, the football or soccer game is a great example of this technology. The players’ personalities in the football club are used to calculate the team chemistry score by FIFA, and the teams mode varies from bad to wonderful based on the game outcomes. In this manner, teams with better players can lose against the weaker sides because of their morale.

Josh: AI is going to be incredibly useful, but it’s a completely different endeavor from something creative or something interactive like gaming.

Ray: For AI GameFi is very limited right now because most of the AI technologies are mainly focused on image enhancement, picture improvement, and attempts to use a deep learning method to transform 3d model pictures into realistic photos.

Josh: AR is the way that trying to interpret images to calculate the depth and put into project, 3D animations into your world.

Pedro: AI could help to prevent fraud on blockchain and I think that’s a really cool mission, congrats on that. But I want to share a little bit of something that we are doing. It’s not about fraud, but we are analyzing all the transactions that happen with our token, and then we have all our user database that plays the game, and then we have all the social media. So we created an AI that goes through every transaction that has ever happened to our token. It goes through every message that a user has sent in discord. It goes for everything that has happened in our project on our back in blockchain and the market. It assigns cohorts and kind of segmentation those users in different behaviors, and we even give them names.

Mark: In my opinion, AI makes it much harder to create a game. Not a lot of developers are too familiar with integrating blockchain into traditional games. Integrating AI, especially AI bots or AI NPCs, makes it that much harder for metaverse projects.

How about AI in play-to-earn and incentives to players and developers?

Josh: The aim of any builders that are trying to build the web 3 game should be to build the game that’s first and foremost fun to play.

Grant: There is too much emphasis on this idea of earning, but also people are not understanding the idea like earning means actually getting compensated for providing something of value. You’re not generating any value when you’re playing video games the traditional way. I think that’s a huge issue that is within the industry right now.

Hana: You leave your time here, so you increase the stickiness in the game. You have some incentive to actually be here longer to play and to have the artificial feeling, in kind of that perspective many games became popular. But are there any other incentives?

Josh: People are spending loads of dollars for no reward in games actually.

Mark: If you take something like VR chat which is one of the most popular metaverses right now. The reason why people keep coming back to VR chat is that they have friends and they like just hanging out.

Noah: Play to earn has been around since online games have been around. All we’re doing is we’re capturing the value that is currently on.