Cyberpunk Video Games • Hacker News Board

Re: Cyberpunk Video Games

#71

Post by freehorse » Wed, Aug 05, 2026, 9:42 AM UTC

ghosty141 wrote: post #67

freehorse wrote: post #53

Earlier quoted context omitted.

> they did increase or tweak crowd density in an update I think I assume something like that to be the case, as I played the game last year and "empty streets like covid" definitely don't describe my experience. But apart from stuff like the NCPD-marked encounters spread around the city, you probably do not want to look too closely at the random npcs, as it does get repetitive fast. But they do give an image of a liv…

> and "empty streets like covid" definitely don't describe my experience I think I gotta explain myself a little better here. If you play the story it doesn't feel like that too much, but if you just walk through the city you often encounter quite empty roads and walkways. See here for example: https://youtu.be/0J55aRPrgOM?t=1285 Like mentioned above, you sadly can't properly fix it by just upping the population dens…

Well, that's in the US (sort of), most people use the cars /s

No, you are right. I see what you mean now, it does not have the population density that corresponds to a superpopulated megacity.

There is a mod that makes it much higher (apparently activating a setting that gdpr had introduced but hidden)

https://www.youtube.com/watch?v=wgsRpxZU-0M

But even then, apart from the absolute amount of people, there is imo an issue with how people are distributed as they increase. People irl are not distributed uniformly and walk alone, they cluster (walk together, hang out in groups etc). Even with increased density with a mod, model repetitions aside, it puts them in a way that feels imo fake/weird. And in a game you have the time to notice these details.

Re: Cyberpunk Video Games

#72

Post by freehorse » Wed, Aug 05, 2026, 9:17 PM UTC

reactordev wrote: post #62

freehorse wrote: post #48

Earlier quoted context omitted.

Which games currently use NN's for managing decisions-consequences in the story?

coming... What you have today is the result of 5 years ago of decisions. Give it time to catch up. There's a couple on the horizon that are going to blow you away. However, there are a few indie titles that have already incorporated some of this in smaller, unique ways. OMEA for example, or Wanderfolk.

From what I see these games simply use LLM apis, not some kind of sophisticated NN mechanics handling the decision trees.

They are not new in that, I have seen a couple of games that were basically slop (and terrible games, judging them as games) wrapping chatgpt.

Wanderfolk seems does not seem particularly inspiring, even the trailer has terrible audio mixing (there are times you cannot hear the narrator because of the game sounds, it is either fully made by an llm agent or by humans that do not seem capable of making a game if they cannot even make a proper short video) which does not make it feel very promising. Everything seems ai generated in general, and more like slop.

OMEA seems a bit more decent, and I would be curious to try it, but I do not like its visual art very much (it is ai generated "and curated by an artist" and it actually shows), and in particular the ai generated images of humans. Irrespective of what I think about ai image generation in art in principle, I think there is a deeper problem here as a game in (broadly) the visual novel genre must imo have a distinctive visual style, and their images just reads as ai generated (before I even read whether they are ai generated, it is that obvious). In general, I find this style of ai generated images of humans very offputting/uncanny (other assets, backgrounds etc I dont get bothered aesthetically as much) and I think I will find it hard stay engaged for long with such a game. Regardless, it seems a more decent attempt than most others I have seen. They do employ writers etc at least to create narratives, so it is probably not a total slop (otherwise, I can just roleplay with chatgpt directly or whatever).

In any case, these games are closer to being "llm wrappers with a certain harness" than having some actually new and sophisticated way to handle decision trees. Another issue with these games is that they do not even use some finetuned local llm, but use apis to providers, with which I am not that comfortable either.

Re: Cyberpunk Video Games

#73

Post by izacus » Thu, Aug 06, 2026, 1:42 AM UTC

A_D_E_P_T wrote: post #44

izacus wrote: post #15

Earlier quoted context omitted.

Calling art design uninspired and plot that brings "nothing new" has to be the dumbest take about art I've seen in a while. I wonder if they actually played the game to write such nonsense.

Dude they literally just aped the styles that were popular on tumblr in the decade before the game was released. Here you go: > https://tacticalneuralimplant.tumblr.com/archive/2013/4 > https://cyberpunktech.tumblr.com/archive/2015/8 Aesthetically, what the game did was literally the very definition of "nothing new" and "uninspired." It wasn't bad, but they definitely weren't pushing any new angles or visions.

This is such a bizarre take - a piece of media set in "New York" isn't "uninspired" because it uses NY as a backdrop. It's what it does with it.

Same as a historical/fantasy game isn't "uninspired" and "nothing new" if it grounds itself in medieval style.

Same here - CP2077 is an established gameworld and putting the game in it is what the title promises. It's so strange to see complaints that a Cyberpunk franchise game looks like Cyberpunk IP. Especially when you see just how beautifully realized it is in IF YOU ACTUALLY PLAY IT.

Re: Cyberpunk Video Games

#74

Post by reactordev » Thu, Aug 06, 2026, 12:17 PM UTC

freehorse wrote: post #72

reactordev wrote: post #62

Earlier quoted context omitted.

coming... What you have today is the result of 5 years ago of decisions. Give it time to catch up. There's a couple on the horizon that are going to blow you away. However, there are a few indie titles that have already incorporated some of this in smaller, unique ways. OMEA for example, or Wanderfolk.

From what I see these games simply use LLM apis, not some kind of sophisticated NN mechanics handling the decision trees. They are not new in that, I have seen a couple of games that were basically slop (and terrible games, judging them as games) wrapping chatgpt. Wanderfolk seems does not seem particularly inspiring, even the trailer has terrible audio mixing (there are times you cannot hear the narrator because of…

I think you are taking things way to literally. Walking around in GTA and in RDR2 was a NN trained in bipedal locomotion (Dynamic Motion Synthesis - NaturalMotion Euphoria). LLM api's are how many are going to use the GPT's without wasting 6GB of VRAM on your GPU for a language model. No, there won't be some novel local neural net powering your next video game but there will be a neural net trained on the games narrative to provide you a more robust experience.

https://www.youtube.com/watch?v=eyP5Pl7pzG0

People are finding ways to include it in existing games but its going to take some time for the game devs themselves to release the games where narrative and dialog are more free-flowing.

Outside of that, there's also a huge huge opportunity in the TRELLIS.2 / Pixal3D space for quickly generating props and 3D assets.

I don't disagree with the points you made, I just never made it clear that upcoming games that utilize AI like this are not exactly how you would picture it. Still, expect some deeper integrations in the years to come.