Nvidia CEO tries to explain why DLSS 5 isn’t just “AI slop”

In a recent podcast, Nvidia's CEO Jensen Huang addressed the backlash surrounding the unveiling of DLSS 5, a technology that uses generative AI to enhance gaming visuals. The announcement last week sparked significant concern among gamers, who feared that the new features might lead to a homogenization of game graphics, resembling what Huang referred to as "AI slop." During his nearly two-hour discussion on the Lex Fridman Podcast, Huang acknowledged the skepticism from the gaming community. He expressed understanding for their concerns, admitting that he shares a dislike for generic AI-generated content. "All of the AI-generated content increasingly looks similar and they’re all beautiful, so… I’m empathetic towards what they’re thinking," Huang stated. However, he emphasized that DLSS 5 is fundamentally different from the simplistic output associated with AI slop. According to Huang, the technology operates on a foundation that is artist-driven, where game creators define the structural geometry and textures that DLSS 5 enhances without altering the original design. "Every single frame, it enhances but it doesn’t change anything," he clarified. The primary worry among gamers is not about DLSS 5 generating entirely new content but rather the potential for it to standardize visual elements across various games, leading to a less diverse aesthetic. Huang countered this notion by explaining that DLSS 5 is not merely a post-processing tool but is deeply integrated into the artistic workflow, providing game developers with the power of generative AI to enhance their creative vision.