| sc94597 said: If the measure of "AI filter" is that it works on 2D image data + buffer data, then all DLSS with the exception of Ray Reconstruction are "AI filters." Personally I think having access to buffer data is a very important difference between this and the "AI filters." I also think that the final release version probably will end up having G-buffer data in its training set and inference to solve some of the criticisms we've been seeing, if it doesn't already (that is still ambiguous.) I think Nvidia's idea was that neural shaders would be responsible for pre-processing of materials and lighting, then DLSS5 would be a final touch up, but they really should just merge the technologies. If DLSS RR can be a DLSS (brand-wise) despite not being purely post-processed, then so can neural shaders. Or maybe it is time for DLSS to be abandoned as a brand and just go back to describing their super-sampler? |
The significant difference being that the final image is really no longer something an artist who worked on that game can say is their art work really.
It's a different image, now Nvidia is being clever and knowing that this will be controversial so they're keeping the image manipulation to look fairly close to the original image for now and trying BS like saying "it's just a lighting filter" (when its not), but likely there's no real limit to how far this could be pushed, it just comes down to which image data set you feed the AI algorithm. If you wanted for example to make the main character look just like a photo real Megan Fox and gave the AI algorithm enough image data, that likely is possible for example. The generative AI doesn't care, it just has its 2D input (image) and then will create from its dataset something it deems "looks better".
At some point then why even hire a full art staff. What art staff's will become is basically a small handful of people that are there probably just to create reference images for the generative AI to understand roughly what the art style/look of the game should be and then it takes over. And even that at some point probably you're not even going to hire reference artists as generative AI will understand basically every kind of art style there is.
