Nvidia’s DLSS 5 is like motion smoothing for video games, but worse

Nvidia has unveiled DLSS 5, its latest AI-powered graphics upscaling technology, which the company touts as its most significant breakthrough since real-time ray tracing. The tech aims to 'infuse pixels with photoreal lighting and materials,' but the initial reveal backfired. Early demonstrations, shown on games like *Resident Evil Requiem* and *Hogwarts Legacy*, applied a homogenizing filter that transformed distinct character faces into what critics are calling 'AI slop'—a generic aesthetic marked by unnaturally smooth skin, uniform features, and synthetic-looking hair.

Major publishers Bethesda and Capcom have pledged support, with plans to integrate DLSS 5 into titles like *Starfield*. In statements, executives like Bethesda's Todd Howard praised the tech for letting 'artistic style and detail shine through.' However, the immediate public reaction was strongly negative, with players noting that recognizable faces, including that of real footballer Virgil van Dijk, were warped into an anonymous, AI-generated look. Bethesda has since clarified these were 'very early' visuals and that art teams will adjust the final effect.

The controversy highlights a growing tension between AI-assisted graphics and artistic integrity. The 'AI face' aesthetic—prevalent in social media and AI-generated films—is characterized by a homogenized ideal that often falls into the uncanny valley. Nvidia's push, backed by industry giants, represents this aesthetic's most visible infiltration into AAA gaming. While DLSS 5 remains an optional feature requiring powerful hardware, its debut signals a potential future where AI processing could standardize and flatten the unique visual identities carefully crafted by game artists.