AI Game Development Enters LiveOps

Hello, and welcome to the latest edition of the Techstrong AI Leadership Insight Series. I'm your host, Mike Vizard. Today, we're with Chris Han, who's the co-founder for Thinking AI, and we're having a little chat about the future of gaming development because while Meta showed us a tool that basically allows somebody to punch in a bunch of prompts and it created a game, which is going to maybe roil the entire industry, but we'll see where this goes.

Chris, welcome to the show. Thank you for having me, Mike. So what are the implications of all of this in terms of how we build games?

I know that there is a multi-billion dollar industry that revolves around this, but if all I have to do is type in a bunch of prompts to create a game, well, will that mean everybody will just create their own game? Yeah. I think it's nothing new, like in game industry and like people, audience, obviously everyone has the attention problem, attention span problem.

Everyone spends a lot of time on the TikTok, on the short drama, a lot of things. It's definitely nothing new. But in terms of the creation of the games, and especially with AI, everyone is a game developer right now.

But it really depends, because those successful games are the post-launch area or post-launch, the applications, operations, and the product iterates. So for me, Pocket, this kind of app is nothing new. I think for Meta, just create a new way to grab the people's attention and create a new kind of platform to help Meta to retain their users.

I think that's my understanding. So it is probable, though, that we'll have more games than ever, and won't that create a set of operational challenges for those companies? So how might they approach that?

Because whereas in the past I might have had, I don't know, a dozen active games that I was supporting, and I might soon have hundreds. How is that going to work? Oh, yeah.

We're going to have more and more games, abundance of the games, in terms of the content creation. We see a lot of things already happening to support those kind of games, the operation, the product iteration, everything. We need a better tool.

We need to-- A real-time feedback. We need to move faster. But I think overall, the logic is still the same.

The data, the first step for support all those kind of games is always the data collection, and especially in this AI era, the data collection. The data is always and is still the fundamental area, fundamental layer for everything to support all those kind of games to really understand what are your users doing and what are their preference and your users' habits inside of your game and on your platform. Those ones are still the basic, the fundamental for setting the basics for the applications.

The importance is rising for the data collection, especially for the agentic applications. We're talking about that a lot recently. So in order to let your agents or let your AI applications to realize more, to feel more, data collection is always the first step.

Yeah, I'll share more when more question comes in. Yeah. But right now that's it.

Do you think as we collect that data, am I just going to see more customized, personalized versions of games, or do you think we might see some new and interesting genres emerge? Because, well, some folks would say we're in a bit of a rut because all these games are variations of each other. Oh, that's a great question, Mike.

I think both ways. For example, on one game, if you have a huge app and giant app and millions of DAUs, and definitely we cannot just run those users at the same level or feed them the same content. There are a lot of cohorts, tags we need to differentiate them.

And because very obvious, different bunch of users, they have different habits. Sometimes they are waiting to see more weapons, sometimes they want to do more outfits, and sometimes they focus on more the skills or capability, the role carried. So in one platform, we need to figure more about that.

And so one gen platform as the game operating, and we need to feed different contents to different cohorts and users. But in terms of we're getting more data for developers, when you-- Yeah, recently we see a lot of niche, every genre, and we see casual games are booming, are rising these two years recently, especially in Turkey, in Vietnam. A lot of very creative ideas are generated in that area.

So it's just so low learning curve. And for new users, for those non-gamers, and it's very additive once you have that. I believe those great deal of ideas, some part are generated from the creation, innovation.

Of course, that's very important, but some part may generate it from the data collections according to the past user habits. So I believe both ways. On one giant platform, that's very good for creative feeding new content to different cohorts.

But at the same time, we can see the diversities and still growing in every genre, new categories. Yeah. Do you think, as this kind of evolves, is gaming still a profession that you would say people should join or get into?

Because it seemed like there was a flood of folks who went to work in gaming because, well, they grew up with gaming. But if AI is going to make everybody a gaming developer, is there such a thing as a professional gaming developer, or how is that going to all evolve? That's an interesting question.

So I thought about my son, my 10 years old son. Because my wife, his mom, is very strict with his screen time. That every time he cannot play the games.

So why not just I created some new game. I prompt to the cloud, "Go to Codex," and then create new game on his own, and he starts to play and he can sit there for a long time. And it's interesting.

But I think everyone, actually everyone, every people, every person on the Earth is kind of the content creator. A game is just one form of the content. So I believe everyone can just-- Is a game developer, for sure.

But for those very successful studios, developers, and they really know how to operate and how to deliver new things, how to iterate their new kind of the contents about the game. I too inspire people to join in this industry. This is a very fantastic industry for creative people, for innovative.

I treat the game as kind of art. And- I always can feel the power in it. That's why I really love the industry.

So I inspire people to join. Just the ways to create the game and the ways to operate and enter the game need to be changed, especially in this new era. Mm-hmm.

You talked about the data in gaming, and it's very intensive, that data, and there's huge volumes of it. What is the challenges with kind of managing that at scale, and how does that look, or how will that play out, do you think? Oh, that's a great question.

I've been working in this industry for more than a decade, especially for game analytics. It's a lot of users, millions users just online at the same time generate, no matter in terms of volume or in terms of varieties, a lot of data. So there are a lot of challenging end points we need to face.

I think the first one is always about what are the targets and what do you want to analyze? So in those kind of big data. So clarify your target.

And around those target, you should find a reality, behavior data, UA data, monetization data, any kind of data around it. And from there you may generate what we call several insights. But for gaming, that's a special point and a special stage because for a lot of legacy and industry, you just get a metric and you get a sense of the what is going on.

That's okay. And for the next step, you have a longer time window in other industries. But for games, the time window is so short, is people love it, people don't like it, and people will churn, and you won't have them anymore.

So the time window is so short. Once we got the insights, we need to act on those insights right away. And because there are too many users, just like I mentioned, you need to feed those different cohort, different texts with different strategies.

That makes game industry, the Liveops, the live service part, so challenging. So actually that's the pain point. And it's just thinking, I want to solve, and not just only get the insights, but also we can act on those insights and figure out the right strategy right away to retain the users for longer time and to give your users the best and the unique experience they want.

So I think all those points make the game analytics, game operations very challenging, much harder than the others. Yeah. There's a world of difference between using AI to create a game and actually putting AI in the game, and that's become somewhat of a controversial topic these days.

What's your feeling? Should we have AI inside these games, or does that create such an advantage that the humans can't win anyway? So I guess humans want to play humans.

Yeah. Of course. So putting AI in games, there are a lot of perspectives.

So you're talking about the NPC, I think. So people really want to play with real humans, for sure. But sometimes the AI becomes much better right now because the two teams matching is always challenging part for game experience.

If you cannot match the same level of skills with different users, you won't get their best experience. I believe some NPC or AI can do that much better than that. But putting AI in games, there are a lot perspectives.

For example, the creatives. Our people and the graphic people always want to generate some inspirations from AI. And for us, in terms of the analytics and in terms of the live operations, we can see much more atomic opportunities from AI because it can behave much faster.

So there are a lot of perspectives. Depends on how we put that into the application. But I do believe there should be the sequence and the logic.

And the starting points always start from human, which means we need to clarify. We need to define what the goals are, what the targets are. And in a part, we can put a lot of execution work into AI, into agents, try to then finish much faster.

But at the ending point is always come back to the human, and we need to define what is good, what is bad. And I still believe humans have a much better taste of the beauty and taste of what the good game is. So that's my current understanding right now.

But I do believe nobody knows what's going to happen next year or tomorrow. Yeah. So what is that one topic that we're not talking enough about these days that people aren't thinking through as it pertains to AI and gaming?

Is there something on your mind that maybe you wish the rest of us were thinking about a little bit more? Yeah. I think everybody right now in the game industry, every industry, we discuss a lot about the AI applications.

I've been thinking this question for very long. And to embrace a new technology better, we cannot just stay like the tools, and there should be a complete philosophy. I think tool is the first step, just like I mentioned.

We're trying to figure out, fill the chasm between the insights to action part using AI. But before AI, we already can figure that out. Use our software.

But with AI, I believe that closed loop can be much faster. But the tools is always the first step, and at the very, very beginning step, we need to think deeper. And I think the next step should be we need to rethink the process.

And just like I mentioned, the starting point, ending point, and the execution part. But let's come back to the analytics to Liveops. In the current process, because we put a lot of more agents into the process, how we redefine this process based on your current process, how we rethink.

I think that's another part. And finally, I think the most important is to rethink the culture we already built in the organization. Because before, as a manager, as boss, as operator, you're always trying to figure out the relationship between human and how human together collaborate each other.

But one step further is when you put more agents in your current process, how we can figure out the relationship between human and agents, and agents and the agents. So that will become way more complicated. So yeah.

That's what I believe. When every new tech, we want to apply them to industry, we need always have a more complete philosophy. So from my understanding, we need to rethink the tools we have, the process we already built, and the cultural and the management skills we already put into organization.

" All right, folks. Well, you heard it here. Hey, the one thing that is for certain is we're probably not going to have any shortage of games.

There'll probably be many, many more to choose from, and a lot of them will, well, there'll probably be a hero with their robot sidekick, but that's not necessarily a new idea either. Hey, Chris, thanks for being on the show. Thank you.

Thank you, Mike. AI Leadership Insight Series. You can find this episode and others on our website.

We invite you to check those out. Until then, we'll see you next time.