NVIDIA's DLSS 5: Graphics Breakthrough or AI Filter?

NVIDIA is positioning the upcoming release of DLSS 5 as the most significant graphics advancement since the 2018 launch of real-time ray tracing, as Engadget first reported. During a March GTC conference, CEO Jensen Huang described the technology as the "GPT moment for graphics," claiming it uses a real-time neural rendering model to infuse pixels with photoreal materials and lighting. Unlike previous iterations—such as DLSS 4.5, which NVIDIA says draws 23 out of 24 pixels through a combined pipeline to increase speed—DLSS 5 actively redraws elements like hair, skin, and fabric.

However, reporting from Engadget reveals a significant backlash from gamers and developers who characterize the technology as an "AI slop filter." Critics argue that the software overwrites a game's original art direction. This sentiment was amplified after GTC demonstrations of *Starfield* and *Resident Evil Requiem* showed characters with smoothed, glossy faces, leading internet users to describe the effect as "yassified."

Professional pushback includes artist Karla Ortiz, who has worked with Blizzard and Ubisoft, stating the technology is "disrespectful" to intentional art direction. Other concerns involve convergence, where games may lose their distinct visual identities if they all utilize the same realism model. While NVIDIA representatives claim developers maintain full artistic control, Engadget reports that Capcom was allegedly unaware its title would be featured in the initial demo. Other early partners include Tencent, NetEase, Warner Bros. Games, Ubisoft, Capcom, and Bethesda. Following a late August leak via *NBA 2K27*, modders have begun testing the software in titles like *Monster Hunter Rise* and *Control* ahead of the official fall release.