Nvidia has taken the wraps off its next leap in graphics tech, and while the promise is sharper, richer game worlds, helping those who don’t have monster gaming rigs to experience the high-end visual fidelity, not everyone is thrilled about what that might mean for how games actually look. Why? Because it uses generative AI to override existing art direction in favour of filtered, commonplace AI art.
The debate erupted after Nvidia revealed DLSS 5, a new version of its Deep Learning Super Sampling technology that takes AI-assisted rendering further. In simple terms, it’s designed to enhance frames in real time using neural networks, adding extra lighting detail, materials, and subtle visual touches on top of what the game engine already produces.
Article continues below 
Since then, the feature set has expanded to include frame generation and smarter reconstruction models. With DLSS 5, the focus shifts again, this time toward AI-assisted scene enhancement, where the system analyses each frame and enriches the lighting and material details before they reach the screen.
Real-time graphics have long pursued the holy grail of film-quality realism, and neural rendering could be a magic bullet that brings it to everyone, including developers, affordably. If it works as advertised, it may enable developers to achieve more complex lighting and surface detail without crushing performance or raising costs. But the reveal has already sparked a wave of scepticism among parts of the gaming community.
Shortly after the announcement, clips circulating on social media drew criticism from players who felt the effect looked less like improved rendering and more like an AI overlay applied to the original game geometry. Some gamers went so far as to call it ‘AI slop’, arguing that it risks smoothing over the deliberate look crafted by game artists.
The Nvidia demo didn’t help, as it showed how DLSS 5 could ‘improve’ games' visuals with a mix of partner developer examples, including Capcom’s Resident Evil Requiem and Bethesda’s Starfield. In RE Requiem, we see improvements to the Grace character, with added ‘realism’, but it’s the kind of uncanny valley AI is renowned for – her hair is softer, lips redder and fuller, skin smoother. Sure, it looks ‘better’, but it hardly represents an FBI character who is fighting for her life against the undead. The grit, hustle, and stress of her character design are removed in favour of generic glamour.
Game artist Karlo Ortiz wrote on X: “Please take it from an artist, all of it being so detailed kills the balance of the image, making cinematic games lit terribly, bringing too much noise that kills focal points of the image, and turns interesting characters into yassified same ol.”
Another artist, Dave Rapoza, chipped in, tongue in cheek, writing: “I don’t think you understand, the public doesn’t want art direction, they want to play “high res photos - the game” starring their AI girlfriends as protagonists - not carefully curated ideas that create a fingerprint for an IP - all IP will look the same, I can’t wait for meta glasses to use this tech so everyone on earth is my AI girlfriend.”
Game visuals are rarely accidental, and colour palettes, lighting styles and surface detail are usually carefully tuned to support a particular mood or art direction, sometimes due to tech constraints, sometimes for artistic meaning – Marathon could load up on detail be chooses a graphic design look, Romeo is a Dead could be photorealistic but opts for a mix of artistic styles to tell its story. If an AI model starts modifying those elements automatically, some players worry the final result could drift away from the developer’s intended look.
However, a key point worth highlighting is Nvidia's GTC announcement, which stated that developers and artists would remain fully in control. Detailed levers enable them to tweak intensity, colour grading, and masking, so enhancements enhance, not override, each game’s unique visual identity. Partner studios at launch include Capcom, Bethesda, and EA – with details on the Nvidia blog showing how each approached using DLSS 5.
In a statement, Jun Takeuchi, executive producer and executive corporate officer at Capcom, said: “At Capcom, we strive to create experiences that feel cinematic, compelling, and deeply believable – where every shadow, texture, and ray of light is crafted with intention to enhance atmosphere and emotional impact. DLSS 5 represents another important step in pushing visual fidelity forward, helping players become even more immersed in the world of Resident Evil.”
Likewise, Charlie Guillemot, co-CEO of Vantage Studios, had this to say: “Immersion is about making the world feel real. DLSS 5 is a real step towards that goal. The way it renders lighting, materials, and characters changes what we can promise to players. On Assassin’s Creed Shadows, it’s letting us build the kind of worlds we’ve always wanted to.”
And indeed, many who have seen how DLSS 5 improves the look of landscape and environment detail in the Assassin’s Creed Shadows demo have been impressed… but character 'upgrades' have been less welcome.
For me, good art direction stands the test of time. Long after technology has moved on, great design and innovative creative choices make some games stand out and last. Think of Persona 5, Nier: Automata, Okami, and even the original PS2-era GTAs with their pop art aesthetic, which is a style Rockstar has evolved, not made photoreal; all the games had a style designed for a reason and with purpose.
That debate taps into a wider anxiety around generative AI creeping into creative pipelines. For critics, DLSS 5 represents another step toward algorithms reinterpreting art originally designed by humans. But for many developers, it's also opening the door to higher levels of creativity, and control is handled properly. As ever, with anything around AI, the debate is going to consume us.
Of course, DLSS itself remains hugely popular. Many PC players rely on it to squeeze higher performance from demanding games, and recent versions have earned praise for how clean their reconstructed images look compared to older upscaling methods. So the reaction to DLSS 5 is likely to remain mixed. Some players will welcome anything that pushes graphics closer to photorealism, while others will watch carefully to see whether the technology enhances a game’s visual style or quietly rewrites it.
What do you think? Is this a good thing for everyone? After all, is it optional to enable DLSS 5 and use the ‘AI filter’, or should developers' and artists’ creative decisions be respected? Take our poll and have your say.
You must confirm your public display name before commenting
Please logout and then login again, you will then be prompted to enter your display name.
