Nvidia CEO tries to explain why DLSS 5 isn’t just “AI slop”
Summary
NVIDIA's Jensen Huang argues on the Lex Fridman Podcast that DLSS 5's 'generative AI' glow-ups are not just post-processing and are 3D conditioned and artist-led rather than simple post-processing. The piece discusses concerns about homogenization of visuals, the openness to training models, and the ongoing push to adopt DLSS 5 through partnerships with major publishers, while noting users can disable enhancements. Adoption hinges on developers and artists embracing the tool.