Machine learning as a game mechanic. - TL;DR: A video game where the AI progressively learns to react to the player's behavior: good idea yes or no?

[Croaking Spider]

I feel like this type of thing would be used in a military application long before it became a novelty in video games. The United States especially has always been trying to find the next best thing in training simulation. Some of the earliest known graphical rendering techniques were being used by the Air Force for flight simulators in the 70s and 80s. Kinda like how DARPANET was the start of the "internet" as we know it.

 

[The Ghost of Kviv]

I'd like to see this utilized in an open world, GTA-esque + sims + simulator game where the world itself reacts to your antics, which then has an effect on how NPCs interact with you, each other, and the game world, as well as how the game world's design changes in response.

Keep committing crimes in a neighborhood? Higher police presence. Houses in the area devalue and can be purchased for cheaper.

Blow up a major bridge? Traffic in alternate routes is now heavier. Businesses you own in the area such as gas stations profit more.

Shut down a major in game company/facility? The game world loses that company, industries shift. Npcs lose jobs and are more prone to hostile behavior, etc. You see where I'm going with this.

Using machine learning in games would be really cool if it wasn't just used to try to smash the player. That's bad game design. The player is playing the game and a good game will play with the player. An ML AI/opponent should maybe work by settings up goals that are unknown to the player and then try to achieve them. In the zombie horde scenario the AI might set up the goal to drive the player up on top of the roof of the shack they're defending, or push them into the nearby caves, or the shoreline etc. Not to defeat the player, just to play with them. It will try to make this happen using the finite units/zombies it has at its disposal.

The player doesn't know what the game is up to, the player might defeat the AI before this happens, but it will appear as if each round/scenario is different.

Exactly, AI/ML isn't just some difficulty tool, it's a way for designers to subtly interact with players. Imagine if the zombie horde can push the player to use weapons that have abundant ammo near them. Or build small scripted events that are created, dropped and activated on the fly. You could even do subtle horror by having zombies talk with the player, and find the right triggers for each player.

And if it's not just zombie hordes, runtime AI voice acting, dialog and interaction creation makes it seem fresh every time.

Unfortunately, all this is going to be siphoned into "cloud gaming" and used to add incentives for live service games. When the internet became popular in the 90's the first big normie deal was AOL, and it was shit. The same thing will happen with AI/ML before it gets better.

 

[Punished Benis]

A primitive form of this has been used in the Armored Core franchise. Besides the trainable unmanned mechs in Master of Arena, which were very primitive and a pain in the ass, Armored Core 3 and several sequels contained enemies that progressively changed up their strategy based on your weapon loadout and playstyle.

 

[The Patrician]

Gran Turismo 7 is trialling something like this, called Sophy. Given how pants on head retarded the general AI is in that game, it can’t help but be an improvement.

 

[Elysian]

This is literally just the Borg in Star Trek Voyager: Elite Force. They adapt to your weaponry so you can only use each weapon in your roster on a drone for a short time before they and every other drone in the level all become completely immune to it, incentivising you to be stealthy and avoid conflict with them (until you unlock the infinity modulator which they can’t adapt to at which point it’s pew pew time).