DLSS 5 criticism has become the latest flashpoint in gaming’s uneasy relationship with artificial intelligence. Nvidia CEO Jensen Huang has waded into the firestorm, dismissing critics as fundamentally mistaken about what the technology does. His blunt rebuttal—”they’re completely wrong”—marks a rare moment of a tech CEO directly confronting online backlash, and it exposes a widening gap between Nvidia’s vision for photorealistic gaming and what players actually want to see on screen.
Key Takeaways
- Jensen Huang claims DLSS 5 critics misunderstand the technology’s purpose and capabilities.
- DLSS 5 uses AI to add photorealistic lighting, materials, and details described as “bridging rendering and reality”.
- Social media backlash calls DLSS 5 a “garbage AI filter” that alters character appearances and removes shadow detail.
- DLSS 5 requires high-end RTX GPUs; older cards experience artifacts and performance issues.
- The tech echoes past divisive Nvidia innovations like frame generation, sparking community debate over authenticity in gaming graphics.
What DLSS 5 Actually Does (According to Nvidia)
Nvidia unveiled DLSS 5 as a breakthrough in gaming graphics, powered by 24/7 AI supercomputers that learn from games and tweak visual output per driver update. The company describes it as “bridging the divide between rendering and reality, empowering game developers to deliver a new level of photoreal computer graphics previously only achieved in Hollywood visual effects”. This is not simply upscaling—it is generative AI that synthesizes lighting, materials, and fine details that the game engine never rendered in the first place. DLSS 5 represents the logical evolution from DLSS 3.5, which used five times more training data than DLSS 3 and delivered better-than-native 4K quality in some games. The technology is not yet widely deployed; full utilization is expected with future RTX 60-series cards, with games featuring DLSS 5 expected to arrive in fall 2024.
The core pitch is seductive: gamers get photorealistic visuals without the computational cost of rendering every detail natively. Nvidia’s cloud AI does the heavy lifting, learning game-specific quirks and artifacts to improve quality over time. This approach differs fundamentally from AMD’s FSR, which relies on bilinear upscaling plus machine learning—a less sophisticated pipeline that Nvidia argues cannot match its cloud-trained results. For players with RTX 4090 cards running 4K at 120Hz with ray tracing and HDR, the promise of Hollywood-grade graphics without the frame rate hit is genuinely compelling.
The “Garbage AI Filter” Backlash
What Nvidia sees as a revolution, social media sees as an AI abomination. Gamers have lambasted DLSS 5 for adding “overexposed” lighting with no structural shadows, excessively sharpening character details until they look plastic, and creating an artificial “filter” overlay that fundamentally alters how games look. The criticism goes beyond aesthetics—it touches on authenticity. One common complaint reframes the technology as “slop tracing,” inverting the frame generation debate where Nvidia previously faced accusations of generating “fake frames”. Comparisons to AI misuse in other contexts, such as generating explicit content, have divided the community despite Nvidia’s offer of toggle options to disable the feature. A YouTuber defending the tech noted that “anything you add AI to, you’ll get hatred for it,” a sentiment that captures the broader skepticism toward algorithmic image synthesis in gaming.
The backlash reveals a deeper tension: players trust hand-crafted graphics more than they trust AI, even when the AI produces technically superior results. This distrust is not irrational. DLSS 5 operates as a black box—gamers cannot easily see what the AI is changing or why. Previous DLSS versions have been praised in specific titles like Microsoft Flight Simulator 2024 and Cyberpunk, where the technology ignores screen regions intelligently and improves glass cockpit rendering on newer cards. Yet the same players recommend older alternatives like TLAA for cards like the RTX 3060 Ti to avoid blur artifacts. Trust is fragile and game-dependent, making a one-size-fits-all AI solution inherently controversial.
Jensen Huang’s Defense and the Credibility Gap
Huang’s dismissal of critics as “completely wrong” is a bold gambit that sidesteps the actual complaint. Critics are not claiming DLSS 5 does not work—they are claiming it does not look good, or that it alters games in ways developers did not intend. These are subjective judgments, not technical errors. By framing disagreement as ignorance, Huang risks widening the gap between Nvidia’s marketing narrative and player perception. The company has a history of this: frame generation faced similar skepticism before becoming normalized, and ray tracing was once dismissed as a gimmick. But DLSS 5 is different because it is generative—it is not enhancing what the developer created, it is synthesizing new pixels based on probabilistic learning. That distinction matters to players who care about artistic intent.
Nvidia’s confidence rests on the assumption that photorealism equals quality. The company has invested heavily in cloud-based AI training, driver updates that continuously improve DLSS per game, and partnerships with game studios. This infrastructure is genuinely impressive and technically ahead of AMD’s approach. However, confidence in technology does not translate to confidence in the company’s judgment about what gamers want. Huang’s blunt rebuttal suggests Nvidia believes it knows better than the community what gaming should look like—a posture that has backfired for tech companies before.
Who Can Actually Use DLSS 5?
Hardware requirements present a practical barrier to adoption. DLSS 5 ideally requires an RTX 4090 for 4K 120Hz HDR with ray tracing; older RTX cards have partial support but experience artifacts. This means the technology is accessible only to players with flagship hardware, limiting its real-world impact and amplifying the perception that it is a feature for enthusiasts rather than a meaningful upgrade for the broader player base. Older GTX cards offer only partial support, making DLSS 5 a next-generation story rather than an immediate shift in how games look today. The fall 2024 rollout timeline means games with full DLSS 5 support are still months away, giving Nvidia time to refine the technology and perhaps address community concerns before launch.
Is DLSS 5 actually better than native graphics?
DLSS 5’s quality depends entirely on the game and how well Nvidia’s AI has learned its specific rendering patterns. In some cases, DLSS 3.5 delivered better-than-native 4K quality, but this is game-dependent and comes with acknowledged artifacts. DLSS 5 will likely follow the same pattern: excellent in some titles, questionable in others. Without independent benchmarks from neutral reviewers, claims of superiority remain Nvidia marketing rather than verified fact.
Why do gamers hate DLSS 5 if it works?
The backlash is not about functionality—it is about trust and artistic control. Gamers perceive DLSS 5 as an AI filter that alters the developer’s vision, adding synthetic lighting and detail that was never intended. This echoes broader skepticism toward AI-generated content. Even with toggle options, the default behavior feels invasive to players who value authenticity over photorealism.
Jensen Huang’s dismissal of DLSS 5 criticism as simple ignorance misses the point. The community is not wrong about what DLSS 5 does—it is expressing a legitimate concern about whether algorithmic synthesis belongs in gaming at all. Nvidia has the technical advantage, but it is losing the cultural argument. Until the company acknowledges that photorealism and player preference are not synonymous, the backlash will only intensify as DLSS 5 rolls out to more games.
Edited by the All Things Geek team.
Source: Tom's Guide
