Gaming

Nvidia CEO defends DLSS 5 and addresses 'AI slop' claims

Related Video: Nvidia Signals Final Major Bet on OpenAI Before IPO

Cheddar / VideoElephant

Nvidia CEO Jensen Huang has further addressed the widespread backlash to DLSS 5 and defended the new tech.

There have been widespread claims that DLSS 5 makes games designed by human developers look like they were created using generative AI, with some describing it as an "AI slop filter".

Huang previously told gamers they are "completely wrong" if they think developers do not have control over this.

And in a new interview on the Lex Fridman Podcast, while he defended the new tech, he softened his stance - to a point.

Huang said: "I think their perspective makes sense and I could see where they're coming from because I don't love AI slop myself. All of the AI generated content increasingly looks similar and they're all beautiful and I'm empathetic towards what they're thinking.

"That's just not what DLSS 5 is trying to do. I showed several examples of it.

"DLSS 5 is 3D conditioned, 3D guided. The artist determined the geometry, we are completely truthful to the geometry that maintains in every single frame. It enhances but it doesn't change anything."

- YouTube youtu.be

After Nvidia revealed DLSS 5 and side-by-side comparison screenshots and videos showed the same content with and without it, it was absolutely panned online with widespread claims DLSS 5 makes games look like they have been created using generative AI.

Although background details look much sharper, which has been noted by some, character faces do look as though they have been created using generative AI and lighting does not looks as realistic because it does not simulate the behaviour of light properly, resulting in the complete removal of shadows in other examples.

It was widely mocked across the internet and the memes quickly flowed across social media.

When claims about developers losing artistic control of their work because of DLSS 5 were previously put to Huang, in a press Q&A with Tom's Hardware, he responded: "First of all, they're completely wrong.

"The reason for that is because, as I have explained very carefully, DLSS 5 fuses controllability of the of geometry and textures and everything about the game with generative AI."

He added developers can "fine-tune the generative AI" to match their style, saying it "doesn't change the artistic control" despite the technology adding generative capability to the existing geometry of the game.

"This is very different than generative AI; it's content-control generative AI," Huang said. "That's why we call it neural rendering."

DLSS stands for Deep Learning Super Sampling and is Nvidia's AI-powered image enhancement technology that renders games at lower resolutions and upscales them in real-time to help improve performance.

DLSS 5 is described as Nvidia's "most significant breakthrough in computer graphics since the debut of real-time ray tracing in 2018" as it introduces a real-time neural rendering model that infuses pixels with photoreal lighting and materials. It will be available in Autumn.

Elsewhere from indy100:

How to join the indy100's free WhatsApp channel

Sign up to our free indy100 weekly newsletter

Have your say in our news democracy. Click the upvote icon at the top of the page to help raise this article through the indy100 rankings.

The Conversation (0)