Ever since Nvidia announced DLSS 5 last week, the reaction has been, well, controversial. And while Nvidia CEO Jensen Huang initially came out and said gamers were wrong for their distaste of the technology, he seems to recognize why people didn't like the trailer now.
Jensen Huang went on the Lex Fridmann podcast to talk about AI technology, and for a moment the topic changed to gaming. When Fridmann asked the Nvidia CEO about the controversy around DLSS 5, he said "I think their perspective makes sense. And I could see where they're coming from, because I don't like AI slop myself." He then clarified, saying "You know, all of the AI-generated content increasingly looks similar and they're all beautiful, so I'm empathetic towards what they're thinking."
While that sounds like Huang was walking back the technology, he insists "that's just not what DLSS 5 is trying to do." Instead, he insists that DLSS 5 is "3D conditioned, 3D guided. It's ground truth structure [is] data guided." And that the model respects the geometry and the "artistry of the artist" and that it "enhances but does not change" the scene.
One of the ways he suggests Artists have control over DLSS 5 is that it's an open model, so game developers can train their own models to control what the output will ultimately look like. Ultimately, Nvidia's intent is to give artists "the tool of generative AI" and that they can "decide not to use it."
Either way, DLSS 5 won't be available to the public until at least Fall 2026, so Nvidia has time to really tune what it wants the technology to be able to do, and how developers and artists will be able to implement it into their PC games. Hopefully the final product is a little less … extreme and will be able to run on graphics cards that people can actually afford.
Jackie Thomas is the Hardware and Buying Guides Editor at IGN and the PC components queen. You can follow her @Jackiecobra
