this post was submitted on 07 Apr 2025
67 points (100.0% liked)

games

20871 readers
191 users here now

Tabletop, DnD, board games, and minecraft. Also Animal Crossing.

Rules

founded 4 years ago
MODERATORS
all 25 comments
sorted by: hot top controversial new old
[–] KobaCumTribute@hexbear.net 33 points 3 weeks ago* (last edited 3 weeks ago) (3 children)

It looks so much like that thing that happens when one gets way too into a game or other piece of media and then one's dreams are just an incoherent loop of something vaguely resembling it where nothing works right or makes any sense. No progression, no mechanics, just an endless replay of nonsense that vaguely resembles what one was doing or thinking about.

Although that seems to serve an important role in parsing and learning from it to do the actual task better, while this is just gibberish presented as a sideshow which goes nowhere and does nothing.

[–] JoeByeThen@hexbear.net 13 points 3 weeks ago

Tetris Effect.

[–] Acute_Engles@hexbear.net 10 points 3 weeks ago (1 children)
[–] umbrella@lemmy.ml 4 points 3 weeks ago

belts.... belts everywhere

[–] ElChapoDeChapo@hexbear.net 9 points 3 weeks ago

I've had those with Smash Bros, Pokemon and Dark Souls

[–] Deadend@hexbear.net 33 points 3 weeks ago

Microsoft is desperately trying to find an AI use case that isn’t cheating on homework or scam related.

[–] MolotovHalfEmpty@hexbear.net 27 points 3 weeks ago* (last edited 3 weeks ago) (1 children)

The amount of 'generation' happening here - if you take that to mean remixing or creating new geometry or features rather than just a lower-res, non-permanent visual copy of the exact thing it was trained on - is next to nothing. The first corridor, the T junction, the outdoor area, the bridge across it, the corridor into the small indoor arena are all exactly the layout of that section of level in Quake 2. The only difference here seems to be that enemies appear and disappear at random (who wouldn't love to play a game where an enemy shoots you in a corridor and when you try to turn around and fight back it's gone forever?) and when the 'player' gets itself stuck in a dark corner it just teleports them somewhere else and starts again. Oh, and the small indoor arena presumably dead-ends because it's meant to have a staircase around a large pillar in it but the AI doesn't seem to be able to replicate it or deal with verticality in general.

Even scraping the absolute bottom of the barrel of any possible, moderated, benefit-of-the-doubt, ongoing research expectations this is such dogshit to the point of being more of a demonstration of why this tech absolutely will not ever be a viable product. Doubly so when you account for the amount of resources, cost, time, and even people involved in this process compared to what a dozen or so people made 30 years ago and one person could knock up in little to no time comparatively now.

[–] KobaCumTribute@hexbear.net 5 points 3 weeks ago

Even scraping the absolute bottom of the barrel of any possible, moderated, benefit-of-the-doubt, ongoing research expectations this is such dogshit to the point of being more of a demonstration of why this tech absolutely will not ever be a viable product. Doubly so when you account for the amount of resources, cost, time, and even people involved in this process compared to what a dozen or so people made 30 years ago and one person could knock up in little to no time comparatively now.

It really is weird that that they can't even think through what it needs to do what they're trying to do. Like they're just grabbing a video generator algorithm and training it to predict what controls will make the video look like and throwing more training and processing power at it until it makes a shitty video for them, instead of approaching it from a gamedev perspective of what is the actual structure it needs to be building off to do this right. It's like how all the chatbots are just text-predictors with no internal modeling or long-term data storage just getting more and more processing power thrown at a dead end, just this surface level aesthetic shit that while it works better than one would expect still fundamentally does not work correctly or consistently or do anything useful.

[–] BelieveRevolt@hexbear.net 26 points 3 weeks ago

There are actually people in the comments defending this and one of them used the word luddite, which is ironic, because the Luddites were right and did nothing wrong.

[–] git@hexbear.net 26 points 3 weeks ago

So glad we burned through 9000 solar panels worth of energy to produce a facsimile of something that a 486 can handle on 10 watts of power.

[–] Krem@hexbear.net 20 points 3 weeks ago (1 children)

i love that looking up or down will teleport you away. the world has no permanence. enemies turn to semi-transparent ghosts if you back away and either come back to life or disappear completely if you come close again.

at my most cynical i guess new games will be developed this way to save money of put "ai" in the name or something, and people won't notice or care about the obvious flaws same way people don't notice how machine-generated images look like garbage and can't tell a machine-generated photo from an actual photo. i personally like when care and thought are put into things like small details, when things have some kind of vision or idea behind them, but i'm in the minority i'm sure

[–] BelieveRevolt@hexbear.net 15 points 3 weeks ago

If you thought there were lots of shovelware porn games on Steam before…

[–] neo@hexbear.net 20 points 3 weeks ago

I played it the other day for a minute and I thought it would be a little interesting in a horror context. You look down then look up and suddenly you're not even in the same room. Yhe enemies all looked like shapes that resembled human figures but were distinctly... wrong.

I think those ideas could be transported into a horror game (made without AI because the microsoft thing was genuinely unplayable and a whole host of other reasons) because it was a bit unsettling, to be honest.

[–] ElChapoDeChapo@hexbear.net 15 points 3 weeks ago

This is so bad it doesn't even qualify as slop

[–] FortifiedAttack@hexbear.net 15 points 3 weeks ago (1 children)

It's a cool proof of concept and that's where it should stay.

[–] Des@hexbear.net 15 points 3 weeks ago

you know what i never asked for. i never asked for billions of dollars to make a game exactly like when i "play" a 1st person game in my dreams

[–] sweatersocialist@hexbear.net 13 points 3 weeks ago

great use of finite resources

[–] NuraShiny@hexbear.net 12 points 3 weeks ago

This is still the best they got. Pathetic.

[–] WeedReference420@hexbear.net 11 points 3 weeks ago

Of course all the comments love it inexplicably

[–] 7bicycles@hexbear.net 7 points 3 weeks ago (1 children)

what do my brainworms need to whisper to me for this to make sense. How do I get from "quake was a very influential game in the 90s" to "we should do one that looks and runs like ass, with AI"?

[–] peppersky@hexbear.net 5 points 3 weeks ago (1 children)

"quake is a videogame we own, that has a very robust demo playback system and is easily moddable, which means we can make a robot play hundreds of thousands of hours of it, record videos of it and then use thousands of gpus to recreate those videos, also as long as we pretend ai is the next big thing, stock prices go up and we get even more stinking rich, also will you hand me that cocaine?"

[–] 7bicycles@hexbear.net 4 points 3 weeks ago

I follow for most of this but even if you were sniffing conga lines of the nose coffee how do you look at this and think "yes, this is what quake is about. 5fps"

Interesting project. Cool idea but not at all worth it.