Now with its own dedicated hardware: the SPU (Slop Processing Unit)
Apparently you need a second RTX 5090 dedicated just to slopification duty
Guess they didn’t realize games have an art style.
Tampering with that by making all games look the same is destructive to the graphic art style.
Booo It’s bad Boooooo
For realistic maybe but every style would need its own models and parameters to work well, seems like one thing more that will only fit for one flagship game and then realistically not used ever again
It’s like putting ketchup on a hamburger! If the chef wanted you to have ketchup, he would have put it on there himself. You’re spitting in his face by making changes to accommodate your personal preferences. One size fits all!
I think it’s a bit different.
It’s more like changing a human made painting with AI to make the person look like modern IG “models”.
I think it’s fine to add condiments to food. That’s not really the same as painting over real human art , with AI slop.
Is DLSS 5 a hard requirement to play the game? If not the condiment comparison is apt.
Ah, using modern Photoshop to look like fake people, tale as old as time. It is kinda funny that in showing that ai is crap, you showed that humans are crap too.
That’s like saying “you shouldn’t mod games, 'cause that’s not what the dev intended”. Idgaf what the dev/chef intended, if I like mods in my burger, I’m gonna season it to my liking.
You can dislike what this dlss5 is doing all you want to, but in the end, no one is forcing you to actually use it. Sure, some devs will start to rely on it, but those same devs never gave half a shit about what their burger tasted like in the first place. They’re the equivolent of a McDonalds burger. The devs who won’t rely on this, and who have a vision for their artstyle won’t be affected from this, in the same way that indie devs aren’t affected by the newest call of duty.
The devs who give a shit aren’t gonna be forced to even make this available in their games. DLSS has always been a choice, if the chef doesn’t want me to put ketchup in their game, they shouldn’t offer it by putting the ketchup on the table in the first place.
It reminds me more of that old meme of “if Mona Lisa was painted today”.
It is too… clean? The lighting is too “perfect”. Each object is an idea rather than an image to be interpreted. It conforms to what instagram considers to be realistic.
This is a mirror of human delusion averaged and contorted to gaze into itself, and as such I abhor it.
This video has a neat showcase of what’s going on here. So-called “AI” adds detail where there should be none, which makes the faces look like worn-out, but polished leather.
No way. Nvidia is currently the only gpu that do utilize 1440p 500hz oled monitors. Dlss4. 5 is a have charger for high refresh gaming.
Where can I post this ? Seriously I would love nvidia to see this.
And I am fascinated by the number of people who, after all those years, still have no idea how DLSS and neural networks work.
I knew that nvidia will at some point make graphics fully ai generated, skipping the rendering.
This is not the case… yet.
Some companies already tried this multiple times in the last two years, the future of gaming is looking bright
Digital Foundry are gushing over how amazing this is. https://www.digitalfoundry.net/features/nvidias-new-dlss-5-brings-photo-realistic-lighting-to-rtx-50-series
Yeah sure it looks more “real” but you are losing so much expression and you are making the image look like it was badly photoshopped.
I think half of their comparison screenshots look flatout worse with DLSS 5
How is the removal of shadows considered to be better lighting in any way?
Yeah i was thinking that, using the comparison tool i was able to see it better and it completely ruined the atmosphere in the first image. The image without DLSS5 looks darker and personally i think it looks better like that, DLSS5 then just completely removes that darkness and makes it look awful.
Also, as someone who is making a game in godot, lighting is EVERYTHING, i am spending hours tweaking my lights so they look good.
In one scene from Starfield DLSS just removes the shadow from the baseball cap bill and replaced it with a soft shadow. That’s not “realistic lighting”, that’s a smear filter with extra steps.
It looks impressive and bad at the same time to me
Impressively bad.
Im a console player so the background in the Skyrim clip looked impressive to me, but everything else looked really weird and boring even though it ‘technically’ looked better to me
Ehhhh. I saw the text (“The Real Deal” poster on the window in one comparison). There’s still some artifacting, which to me is a deal breaker, especially if I have to DLSS everything instead of selected items.
But I’m also not a game developer, so I don’t know if the modeling differences are worth the tradeoff for something that most people won’t see in motion.
As an embedded systems engineer, I find VLSI shaders to be a fantastic technology.
Simultaneously, AI code overall is absolute fucking trash.
I hate how bad NVIDIA has to be in general, because DLSS is legitimately amazing technology. Even the frame gen in DLSS 4 that was pretty rough at first is shockingly good in Borderlands 4 for example. Too bad they’re the reason no one can buy RAM
No thx. I’m not too big on the last few generations of gaming anyway but if they’re really going to push their slop like this I’m out fully.
Was watching the Digital Foundry review of this tech. They seemed impressed, the people in the comments were scathing
Everyday I think things can’t get worse. That maybe this will be the breaking point. But no.
The main Resident Evil example is just awful. It’s like an Instagram filter.
It’s worse than that guy who has been going around photoshopping “Ugly” female game characters to make them look hotter.
I hate this reality.
Don’t forget that when Sony went back and did the HD remaster, they actually remodeled her that way.
All of the worst people are winning.
Jesus Christ
Well at least the photoshopping is done by a human.
It looks like dog shit lmao
Eh, the examples shown imo were 50/50 improvements vs downgrades. First few were good, couple looked really bad, then a few good ones again. I suppose it’s subjective. As long as it’s just an option that can be turned off I don’t think it’s that bad.
I personally think these looked really good IDK why it’s getting so much hate other than “ai bad” complaints
Bruh she don’t even look like the same person. Also, i’m deeply concerned that if this garbage tech takes off that there will be little to no incentive to make games look as good as they can because gamers are just going to use the AI slop filter to make it “look better” anyway. Also, due to the fact that this is generative AI, It is impossible to guarantee that everyone sees the same “enhanced image”. You might see a completely different face than your friend playing the same game. Everything about this tech sucks. I’m playing Requiem on a base PS5 and I think it looks amazing. I think it’s so disrespectful to the artists who worked hard to make this game look good to just paint over their work with an AI slop filter.
You can just…turn it off? Devs can just… Not implement it.
Not really that big of a deal imo.
impossible to guarantee that everyone sees the same “enhanced image”.
Let’s be honest this thing is not generating totally new faces per session the constraints on this tech are not so wide that we’re going to think it’s different people, you can alter the randomness of the models and it should be tuned accordingly.
How many R’s in room since you’ve got a fancy nvidia card?
Also, twenty feathers and twenty pounds of bricks both weigh twenty pounds.
I would try to interpolate your probably real question here, too. And I’m no AI. Unless I missed your point
Someone posted yesterday with a question asked to AI.
What weighs more, 20 pounds of bricks or 20 feathers?
The useless chat bot will always answer with “they both weigh 20 pounds” because that’s what the training data always says when asked about bricks and feathers.
That’s what I said. I expect it to interpolate silly questions to a sensible meaning. Of course it would sound like you wanna compare a comparable set of data. I would, as said, also assume you meant 20 pounds of both and just forgot to say that. Maybe adding “take this question literally” helps?
No fan of AI here, but hate where hate is due 😁
They did ask it to consider and repeat the exact wording of the question, yes.
Hm, i just had to try myself with chatgpt, and:
so, as suspected, it just tries to ignore one’s nonsensical questions unless asked to take it verbatim. This is actually not even dumb. If it had said “your question makes no sense”, we would call it stupid to not see the obvious error in the question - as it’s a run-of-the-mill-“riddle”.
I’ll have to find the post, but you did it in two steps, changed the units of mass, and object.
The post, which is extremely hard to find with the latest slop release from Nvidia, asked the chatbot to consider the exact wording, without babying it into the correct answer. All because the close variations of the phase “X pounds of bricks and X pounds of feathers weigh the exact same” have been used in various textbooks and such for at least the last hundred years or so.
That means that the chatbot has seen that exact combo of words, in roughly that order, quite a bit more than your use of “100 kilograms of rice”. At least in English.
You can baby it through when the training data is sparse, but not when there are hundreds of uses of the same phrase over and over again in the training.
