The good, bad and baneful

It seems like it was only yesterday when conversations in the video game world were all about how “next-gen” Red Dead Redemption 2’s NPCs were or how the dice rolled with each decision in Disco Elysium, carving the story ahead. Little did we know that these belied a hotter topic that’s now evolved beyond recognition. Today, our video game speak is riddled with two letters – AI. Like it or not, AI is here in video games and doing more things each day – both good and evil. It’s the little guy on your shoulder cosplaying as both, the angel and the devil. Here’s looking at just how AI has impacted and is impacting the world of video games:

AI/ML application in gaming: When non-player characters and CPU players made in-game ‘decisions’

There’s a clear differentiator when we speak about AI today versus when it was mentioned in games a decade ago. Years back, AI often referred to the CPU player that you play against in a single player or a multiplayer game or Non-Playable Characters (NPCs). Today, the conversations are around using generative AI to build video games or getting man-made programs to do your programming of games for you.

AI in-game, in recent years, has been a crucial part of the way we’ve enjoyed games. Autonomous decisions taken by NPCs when you interact with them have been vital to building an immersive world. You whisk away the rake towards the lake in Untitled Goose Game and the NPC gardener gives up everything he’s doing and dedicates his life to bring it back. These kinds of simple AI mechanics are a mainstay of gaming.

This is the kind of AI that can be further developed to really put you in a world that’s far different from yours yet so familiar.

This machine has a read on you: Moves that are only possible with machine learning; and procedurally generated levels or dungeons

The Last Of Us Part II boasts some ridiculous AI programming where NPCs feel almost too real. You shoot a dog that’s about to attack you, it whimpers, and you hear its NPC owner in the distance cry out the dog’s name, making you feel like the worst human being in the world. The way the NPCs track you in the same game is almost too unreal, either on scent or from the way your last victim fell to the ground – this is the brilliant kind of machine learning (ML) capabilities that are needed to evolve games.

Procedurally generated levels or dungeons are some of the finest applications of AI with algorithms. No Man’s Sky, Hades, The Diablo Series are all well in the spectrum of this kind of programming that brings an unprecedented level of replayability to these kinds of games simply because no level is ever the same. The levels shift based on the AI and the algorithm, and a new day brings a new dungeon perennially.

AI/ML to prevent cheating in videogames

Anti-cheat software has been one of the biggest applications in the large-scale world of gaming. Esports tournaments and players often rely on AI detection of cheaters and ensure that no one gains an unfair advantage over the other player, levelling the playing field for teams or individual players.

Machine learning helps understand movements, paths, shots and impact times to estimate whether a player is cheating or not, and take the appropriate course of action. Denuvo is one of the best examples, albeit plagued with controversy regarding its digital rights management and exposing those pirating games as well. It also affects developers with its staggeringly high subscription costs for licences, but at a core functionality level, it’s doing what is expected of a player or a dev.

Doomsday: When Souls-Like Bosses turn on us

Perhaps the worst-best scenario of AI/ML shifting the way we game, could be someday seen against traditional boss fights. Souls-Like bosses already scare me, it’s horrifying to think that someday machine learning could be attributed to these programmed bosses. This might enable the bosses to figure the player’s patterns and play the reverse uno card, making us the boss to the game. I hope that day doesn’t come anytime soon. All of the above point to the fact that AI can make gaming a beautiful place and really up the satisfaction levels of players and what they expect from video games.

(For the uninitiated, Souls-like or soulslike is a sub-genre of gaming marked by very difficult play and highly atmospheric world-building).

Is AI coming for gaming jobs? And what will this mean for players?

AI undoubtedly enables developers to let the computer replicate complex algorithms rather than strenuously hard-coding each one, making for a better gameplay experience. But what happens when generative AI enters the picture at the development stage? You get a lazy developer, resource cost cuts, lay-offs, corporate greed and most of all, games lacking that much-needed human touch.

Sure, you could create vast open worlds beyond what is possible today and keep them ever expanding, but the elements that make that world interesting need to reside within human judgement. Generative AI software is being used to write NPC speak and craft art. Take-Two’s CEO, Strauss Zelnick, has already spoken countless times on the benefits of AI for NPCs in the Grand Theft Auto series but even those ‘barks’ and ‘chatter’ could seem less human and more robotic with the unchecked use of AI.

Furthermore, you’re bringing down the potential of the NPC, which could have been tapped by an excellent narrative writer – a trait you often want in large open world games. A recent survey by Game Developers Conference asked thousands of devs about the industry, at a time when there have been layoffs at big companies such as Epic, Twitch, Microsoft and more. More than 3,000 participants suggested that there were more layoffs to come and nearly 49 percent of the respondents mentioned that their companies were using generative AI in some form. Studios growing too quickly is mainly attributed to this, however it’s not easy to deny that the ethics of AI usage too will play a strong role in this.

There’s a very muddy area of copyright infringements in AI usage. Studios fear loose adaptations of iconic characters that can be generated by AI and sold in shoddy clones of games. It’s also scary to imagine the role AI might play in monetization strategies of games. The loot-box world is already under the watchful eyes of those who loathe it and those who love it. Fairness flies right out the window when a robot does your evil deeds.

Do we let the bots through the door?

There’s no denying that most folks have tried their hands at AI in some form, whether it’s ChatGPT or some form of art creators. The potential benefits are almost instantly clear the moment your eyes widen to that first generated response. But it’s up to game developers to use these powerful tools mindfully. Should they use AI to make the basics and let human-touch fine-tune it? Sure. Should they allow games to work their way through most part of the game development process, absolutely not. Should AI art take over traditional artists? Definitely no. Strict ethical norms need to be cemented before we use AI in the world of video games. It’s the biggest boss battle that’s going to take game developers and consumers alike; the spoils of which make gaming enjoyable for everyone – both as a passion and as a profession.