What Nvidia presented as a major step forward in gaming graphics has quickly sparked backlash online, with critics saying that the new feature is an AI beautification filter that limits artistic direction in video games.
The chip and AI giant unveiled on Monday Deep Learning Super Sampling (DLSS 5), an image-enhancing technology, which it describes as a breakthrough that uses generative AI to deliver near-photoreal visuals.
DLSS is Nvidia’s AI-driven graphics technology that began as a way to upscale resolution and now generates entirely new frames, with support in more than 750 games.
DLSS 5 will arrive this fall, but Nvidia presented a sample of what games will look like with the new technology earlier this week.
But across YouTube, Reddit and X, the conversation has focused less on realism and more on how different the images look, especially when it comes to faces.
Clips comparing “DLSS 5 off vs. on” began circulating within hours of the announcement.
While backgrounds and textures appear sharper, many users pointed out that characters themselves look altered and AI-generated.
In a “before and after” post on X by Nvidia, facial features of a popular game character, Grace Ashcroft from Resident Evil Requiem, appeared smoother, with brighter skin, and details like eye bags or imperfections were heavily reduced.
From realism to ‘yassification’
A large part of the criticism centres on how characters are presented.
In several demo clips, characters appear to be noticeably changed with near-perfect lighting and skin.
Some users said the results resemble beauty filters rather than graphical improvements.
On YouTube, one commenter wrote: “Why does every character look like they walked out of a beauty filter app?”
The term “yassified” quickly became a common way to describe the effect. Online, it usually refers to images that have been edited to look more polished or conventionally attractive, often at the cost of realism.
Memes take over
Memes soon became the main form of criticism.
Users began posting side-by-side images labelled “DLSS 5 off vs. on,” often exaggerating the effect for comedic impact.
One widely shared example used Kratos, a character from God of War, showing a side-by-side where the original model is placed next to a version with exaggerated makeup and softened features, labelled “DLSS 5 on”.
Other posts used actors, game characters, or even animated figures, placing stylised originals next to hyper-realistic or overly polished versions.
One Reddit user captioned a comparison: “DLSS 5 when it forgets the game has an art style”.
Artistic Control and the Uncanny Affect
Behind the humour, many players are raising concerns about how far the technology can go.
DLSS originally focused on improving performance and upscaling resolution but transformed into generating frames. Some users argue that this shifts the role of the technology from enhancement to modification.
“The obsession with fidelity over art direction is reaching terminal levels,” one commenter wrote on YouTube.
Others on Reddit pointed out that the feel of hyper realistic graphics is not right for video games, “At some point it doesn’t feel like a “game” anymore. It just looks like an IRL video or movie with characters you can control… It feels off somehow.”
Another user also said on Reddit that the new look felt like fan art: more “real” on paper, but wrong in spirit.
This feeling is what social media refers to as the uncanny effect the filter creates.
One YouTube comment described it as: “Everything looks high quality, but nothing feels right.”
Tech Concerns
Some Reddit users mocked the fact that Digital Foundry’s hands-on demo used two RTX 5090s, one to run the game and one to apply the effect.
On Nvidia’s own subreddit, posters questioned how far the tech can really spread if most GPUs cannot run it well, and whether studios will want to spend extra time implementing a feature for a relatively small number of players.
That has fed a wider suspicion that DLSS 5 is being marketed as a shiny future while ordinary PC gamers are still dealing with expensive hardware in the present.
Nvidia and Companies respond
Nvidia has pushed back against the criticism.
CEO Jensen Huang said critics are “completely wrong,” arguing that developers still have full control over how DLSS 5 is used. According to the company, studios can fine-tune the AI to match their intended visual style.
“The reason for that is because, as I have explained very carefully, DLSS 5 fuses controllability of the geometry and textures and everything about the game with generative AI,” Huang told media platform Tom’s Hardware.
In a pinned comment on its YouTube demo, Nvidia also said developers retain “full, detailed artistic control” over the final look.
Bethesda was one of the first game studios to address the reaction.
The studio said the footage shown was only a “very early look” and stressed that any final implementation would remain “under our artists’ control” and be “totally optional for players”.
At the same time, Bethesda director Todd Howard took a more positive view. Speaking during Nvidia’s showcase, he said the company had already tested the technology in Starfield and that it was “amazing how it brought it to life”.
A divided reaction
For Nvidia, it represents a shift towards combining traditional rendering with generative AI.
But for many players, it raises questions about how much control should remain in the hands of developers and creates concerns about the increasing demand to use more expensive gaming tech to run such features.