Nvidia CEO Jensen Huang addressed gamers' concerns over DLSS 5's generative AI features in a Lex Fridman podcast interview, admitting he dislikes 'AI slop' while emphasizing the technology's artist-guided nature that enhances frames without changing core game structures. Partnerships with major studios signal broad upcoming adoption.
Nvidia announced DLSS 5 last week before March 23, 2026, introducing generative AI enhancements for gaming that sparked backlash from the community, who compared it to an 'AI slop filter' homogenizing visuals. Initially, CEO Jensen Huang dismissed critics as 'completely wrong.' However, in a nearly two-hour Lex Fridman Podcast episode published on March 23, he adopted a conciliatory tone, stating, 'I don’t love AI slop myself… all of the AI-generated content increasingly looks similar and they’re all beautiful,' and expressing empathy for gamers' perspectives.
Huang defended DLSS 5 as distinct from generic AI content, describing it as '3D conditioned, 3D guided.' Artists provide the 'ground truth' geometry and textures, and DLSS 5 'enhances but doesn’t change anything' in frames. Unlike simple post-processing, it's integrated with artists, offering generative AI tools to train models for specific styles—like 'a toon shader' or non-photorealistic looks. This builds on prior DLSS versions for resolution/frame rates and aligns with Huang's earlier remarks to Tom’s Hardware on 'generative control at the geometry level.' Artists can opt out, and gamers can disable features.
Nvidia, now a $4 trillion AI powerhouse, has partnerships with Bethesda, Capcom, NetEase, NCSoft, Tencent, Ubisoft, and Warner Bros. Games for DLSS 5 integration, with debuts months away.
