Would like to point out again, that it doesn't take a data-center's worth of compute to train sub-billion parameter models, which DLSS models are because they have to run in real-time. 

AlexNet was trained on a gaming PC with two GTX 580s in 2010 for about a week, and that is probably on par with Nvidia's CNN DLSS in terms of parameter count (60M.) The DLSS transformer model is about twice that size. Nvidia did use a super-computer for DLSS 1 but that was mostly to render the synthetic data, which they probably do need to extend with newer iterations, but can still reuse old data. 

Also if Deep Learning died tomorrow, there are a dozens of other use-cases for these data centers. Nvidia GPUs aren't FPGAs, they are fully-programmable and can run many different compute workloads. You could see them being used for CGI, scientific computing, big data processing, etc. Nvidia is probably the most hedged of the AI companies. 

Oh shit, you nailed it - that's exactly it, DLSS5 version of her looks exactly like that generic AI porn slop. Now I can't unsee it...

Wasn't DLSS supposed to give us a boost in performance, keeping the visuals as closer as possible to the original while sparing us the need for stronger hardware?

It doesn't looks like the intention for DLSS 5...

Yep, typical corporate hubris; apparently you're wrong if you don't want their slop.

"It's not dogshit, it's hot stinky dog shit."

I don't think we should get too caught on the semantics. We can look at it more holistically at what is happening, and what people are finding issue with.

A game designed with a specific character & look in mind is changed drastically, not just in quality but in tone and personality. Capcom has super talented teams, if they wanted Grace to look like the DLSS 5 example, that would be way more evident especially in the cutscenes and pre-rendered art and even in-game with intentional shading/texturing etc. That is simply not the case.  

The characteristics brought out are specific to this  DLSS 5 experiment and it is removed from what was up until that point established by the art team and accepted by the audience.

It doesn't actually matter who is at the helm of the change, the fact still remains true. This isn't unique to DLSS 5 either, some recent examples include Windwaker HD where the advanced lighting and bloom completely alters the toon/cell shading effect. It is simply against the outcome of the original art direction regardless of whether Nintendo gave it the green light or not. I actually like both (WWHD & OG) but its absolutely fair criticism. And this happens a lot, remasters often mess up art direction.

This is made especially worse in this case because its not a remaster with all the internality that goes into that, but instead a broadbrush AI tool trained on who knows what... It gives people the "the ick". Developer input doesn't negate this and so far they all have a fairly similar "AI" look. The Grace example is the focus here but it just reflects a wider concern with the technology.

Fundementally this particular argument shouldn't be confusing to anyone. It should be very clear to anyone who is looking at that image or who has played that game and built a relationship with that character. This doesn't dictate whether we have to like it or not, but simply this is what people are reacting to and more the spirit of  the "artist intention" argument.

I must not be as picky as others. WWHD, demon remake, colossus remake, etc. Loved them all. Never thought the new art changed the original intent enough to be upset.

Given how great DLSS has been, I will give the newest version the benefit of the doubt.

Actually, SotC remake (and DS to some extent) fundamentally change the original artistic vision. Not in a "it's a matter of preference" kind of way like in WWHD (which just changed the lightning and improves on some textures), but in a "now this desolated and grim land is filled with vegetation, light and beautiful landscapes when it shouldn't be the case and the protagonist's facial expressions frequently contradicts the mood of the scenes".

Which funnily enough, it's exactly what we are seeing in some of the examples of DLSS5.