Almost every element of Concord was heavily criticized.
Apparently already being done.
Scavenger T.O.M utilizes AI-based environment generation technology to explore and reconstruct the radiation-ravaged surface in real-time. In hazardous areas where players cannot venture directly, AI acts as T.O.M’s eyes and analytical core, dynamically processing the environment based on movement paths.
The Diffusion model-based AI system goes beyond simple image generation, creating explorable spaces at the object level and seamlessly integrating them into the game environment to enhance interactivity.
Through this technology, Scavenger T.O.M provides a unique survival environment that evolves with the player’s exploration. Step into the AI-generated wasteland and craft your own survival strategy.
Planned Release Date: To be announced.
Not really; if you go to the summary on the wiki most of the reviews said the gameplay was fine to good and just criticized the character design.
PC Gamer went all out on criticism, but that wasn’t the consensus.
Imagine an open-world city game like GTA where every house and apartment can be entered, and if it hasn’t been designed in advance for story purposes, the AI will design the interior of every single one the moment you enter it. Think of the weird and wonderful stuff you’ll find.
Those always seem so boring though, just basic rooms with randomly generated filler material and loot. Fortnite is full of them.
I much prefer handcrafted set pieces, personally…
When people talked about how revolutionary Cyberpunk was supposed to be, and then it came out, I wasn’t really paying a ton of attention. So when there was a lot of disappointment in the game on release, I looked into the hype to see what was missing; and the game that people thought they would be getting seemed impossible to create to me.
And then generative AI started getting big, and I figured: the game people wanted could be made in a few years, maybe. Cyberpunk 2077 came too early to be that game. But it’s coming, and it will be truly revolutionary.
Because right now they’re built by simple random generation tables. AI could do much, much more.
We’ll see…
I mean, AI can already write entire novels today. Doesn’t mean people want to read them.
I’m sure the quantity of procedurally generated content will increase, but not sure about quality.
Games will still have plots, key characters, and key areas written and designed by human beings. But even the best games also have a lot of filler and repetition in them; filler and repetition are where AI shines.
Think of it this way: walk around a videogame, and you’ll see the same tables and the same chairs over and over again, in different buildings and even in different countries. That’s just the way videogames are. Now imagine a game in which every table and every chair was different. Think how much that would add to the game’s verisimilitude.
Do they…?
I dunno, even in the procedurally generated maps of Diablo / Path of Exile / etc., the procedural content isn’t really the focus of the game. They could just be colored circles flying at you and the game wouldn’t really fundamentally change. It’s the gameplay balance (of skills, equipment, classes, enemies, etc.) that’s hard to get right, and that takes a lot of permutation and testing. The filler content is really just in the background, not a core part of the gameplay loop.
OK… but what would that actually add to the game? It’s like walking into a chair factory where every chair was handcrafted… but so what?
Content for the sake of content is just filler; only when they meaningfully interact with the challenge and reward systems are they worthwhile. We don’t need more fetch quests, random items with +/- 0.5% different stats, etc. Even generated NPC characters like Nvidia’s seem… bland? To me that just screams “how do I skip the dialogue and get on with the game… and oh god, please not another randomly generated quest”, not “oh wow, this is going to add so much depth and longevity to the game!”
I don’t think this is an inherent limitation of the technology, mind you, just its current progress. Maybe someday (soon?) it will rival the best game designers in being able to create interesting challenges of varying difficulty and skill, but I don’t think it’s there quite yet.
As for graphics/models/storylines/etc, there is already so much crap out there (human or procedural) that it’s not really a problem of not having enough content or not having enough realism, but of having a terrible signal to noise ratio. I’d much rather see a few more indie, low-res, gameplay-heavy games like Balatro than the next Ubisoft cookie-cutter game with holographic VR and hyper-real smells and whatever…
Immersion.
Why do we even need fancy graphics at all? What’s wrong with bit sprites and imagination?
Verisimilitude is overrated. I want to play a game. Not walk around a static environment inspired by Architectural Digest.
I am currently replaying Dragon Age: Origins. It’s just as amazing and satisfying as it’s always been, despite the extremely dated graphics, because it’s a great game.
We don’t…? They’re fine today. They were fine 10 or 20 years ago too. It gets incrementally better every year, but to the detriment of gaming. Expensive games, massive layoffs, poor gameplay, super high risk, $2k graphics cards, shitty story… I don’t think it’s a net positive. All the actual innovation happens on the indie front with games using 2000s and 2010s graphics, not the latest AAA filler material.
They don’t sell as well, so multinational gaming conglomerates who care more about profit over fun don’t tend to make them?
Otherwise… they’re fine?
Yeah, especially compared to Veilguard, an entirely forgettable high-production-value formulaic copycat like all the others that so plague the AAA industry today.
Special effects, in both movies and games, have dramatically increased in realism in recent years, but they’ve taken so much money and resources out of development that the rest of the title tends to suffer because of it. We could do with a little less of that, honestly…
So do I - but there’s no reason why one should come at the expense of the over. You can design excellent gameplay, write great plots and characters; invent original game worlds, and enhance immersion by using AI. It’s just another tool in the toolbox.
90% of everything is crap; fortunately, the remaining 10% is more than enough for me.
Well, there is one big one: Publishers and investors. AI is just another cost-savings measure to them, not a creativity-enhancing, gameplay-improving one.
Easier immersion doesn’t make for better games, but it certainly makes for better sales pitches. Have we learned nothing from the whole VR hype cycle, that several times over ended with more immersive but usually boring games?
Publishers and investors have always been the enemy of quality and always will be. AI won’t change that one way or the other.
VR failed, again, because those damn helmets are uncomfortable and because the controllers are useless, not because of immersion. If I could wear VR rig as easily as I wore pair of glasses, and if it let me use a mouse and keyboard, I’d be playing games on it day and night.
Well, real rooms in real living spaces are usually boring as well.
The Division had a large number of apartments you could enter. Most used the same assets and you’d eventually see the same wardrobes, wall art, unmade bed, etc. But it didn’t really take away from them and they still looked like lived spaces while you were primarily there to scan for loot chests or story echoes. The effect of having a lot of apartments to investigate was still a positive one and it still made for a richer environment. If the apartments had been created by AI rather than a dev saying “Let’s put Cupboard03 and DirtyTable17 in this one”, I don’t think it would have been noticed.