The presentation of NVIDIA's DLSS 5 has generated considerable buzz in the gaming and developer community. What was supposed to be a major leap forward in PC graphics has instead been met with a wave of criticism, doubts, and unkind comparisons to the so-called "garbage AI" that many users are beginning to detest.
Far from avoiding the topic, the NVIDIA CEO, Jensen Huang, has had to come out and clarify his own words Regarding the detractors of DLSS 5 and how the company intends to use artificial intelligence in video games, in a lengthy conversation, the executive acknowledged that he understands the public's mistrust and dropped a phrase that has spread like wildfire: he He also dislikes "AI-generated garbage".
From “they are completely wrong” to understanding the controversy
Following the initial announcement of DLSS 5, The first official demonstrations did not convince everyonePart of the community interpreted the presented images as a kind of generative filter that "beautifies" everything in an unnatural way, leading many games towards a photorealism that is flat, uniform, and very similar to each otherThis fear that AI will homogenize the appearance of titles sparked controversy.
Tensions increased when, in response to these criticisms, Jensen Huang responded with the phrase "they are completely wrong"The comment, which was widely circulated in the media and on social networks, sounded dismissive of the players' concerns and was interpreted as an unwavering defense of the technology, lacking empathy for the weariness many feel with the omnipresence of AI.
In his participation in the Lex Fridman's podcastThe tone was different. When the presenter directly brought up the controversy surrounding DLSS 5, Huang admitted that The critics' perspective "makes sense" And he understands where the rejection comes from. There he made it clear that he himself cannot stand what he calls "AI slop," carelessly generated content that ends up all looking the same and that, even if visually appealing, is generic and repetitive.
This change in discourse shows an attempt to to get closer to the perception of the average consumer, who is increasingly tired of AI features touted as “revolutionary” that later feel intrusive, unhelpful, or simply unnecessary. In a context where many users associate these solutions with bloatware, Huang attempts to distance DLSS 5 from that “AI junk” image.
What DLSS 5 aims to do, and what this technology is not.
Beyond the media noise, NVIDIA insists that DLSS 5 is not a shoddy filter that is applied on top of a finished game.According to Huang, the new version of the technology is "3D-driven" and "guided by real structural data," which in practice means that it operates respecting the geometry and original composition of each scene.
The CEO himself points out that DLSS 5 It draws on information from meshes, textures, and artwork created by the development team.The model works frame by frame to refine the visual result, but without altering the underlying artistic intention. Or, in the words of the executive, it "improves, but doesn't change" the scene: it raises the perceived quality while maintaining the essence of the artist's work.
One of the points Huang emphasizes most is that DLSS 5 does not function as a generic post-processing tool that is applied automatically over any game. The technology is integrated into the development itself and is conceived as one more tool for creatorswho can decide how and to what extent to use it, or even not implement it if it does not fit their project.
In fact, NVIDIA envisions a future where the system is open enough to allow training of custom modelsThis would open the door to specific style adjustments: from requesting cartoon-like shading to adapting the aesthetic to a particular artistic universe, always maintaining consistency with the original design. AI would thus become a creative assistant, not a substitute for human art.
In this approach, DLSS 5 distances itself from the idea of "garbage AI," where a generic algorithm generates brilliant but impersonal results. Here, the promise is that the tool act as a multiplier of the artist's worknot as a standard filter that would erase nuances and visual personality.
Neural rendering, photorealism, and fear of homogenization
DLSS 5 relies on Neural rendering to improve the graphic quality of all game elementsIt's not limited to faces, although NVIDIA's initial demonstration focused on faces because that's precisely where the visual leap is most evident and easiest to show in comparisons.
Tests shared by the company and by specialized media show that the technology It affects characters, clothing, hands, small objects, and entire settings.For example, there is a more realistic interaction of light with surfaces such as fruit, fabrics or book pages, as well as superior definition in fine details that were previously blurry or unclear.
In larger scenes, such as castles or complex environments, DLSS 5 brings More natural lighting and shading, greater sense of depth and a richer perception of geometric volume. Without this processing, many scenes can appear flatter and more generic, with less visual impact, something that is especially noticeable at high resolutions where the flaws become apparent.
According to those who have analyzed these demonstrations, the difference is large enough to to give the impression of one or more generational leaps in graphic qualityThe goal is to approach photorealism or, at least, a level of detail that until recently would have required exorbitant resources in terms of polygons and geometric load.
At the same time, NVIDIA emphasizes that all of this is achieved without increasing the complexity of the meshes nor the polygon count to unmanageable levels. The idea behind neural rendering is precisely to rely on AI to recreate materials, lighting, and micro-details that previously had to be explicitly modeled, with the consequent impact on performance and hardware requirements.
A European community tired of “AI everywhere”
The context in which DLSS 5 arrives also does not help to soften the reception. Many consumers in Spain and the rest of Europe They show clear signs of fatigue given the proliferation of AI functions appearing in operating systems, office applications, social networks and, of course, video games.
On PCs especially, the average user has been seeing how they are integrated for months. assistants, automated functions, and AI modules which often consume resources, reduce performance, or complicate the experience, without a clear benefit. Amid concerns about component costs and with a graphics card market still marked by years of tight pricing, patience is not exactly endless.
This climate explains why a part of the community views any new AI announcement in gaming with suspicion. The line between innovation and passing fad It's a delicate matter, and many players fear that resources will be allocated to features that are flashy for marketing purposes but of little use in day-to-day operations. Hence, they describe some offerings as superfluous software or simply "garbage."
At this point, it becomes clearer why Huang's statement acknowledging that He doesn't like AI-generated garbage. It has had such a significant impact. It is interpreted not only as a change in tone, but also as a message directed at those who doubt that these technologies will be used responsibly and judiciously.
For NVIDIA, the key will be in Demonstrate with facts that DLSS 5 provides real valueIt improves the gaming experience and respects the creative vision of European and Spanish studios that choose to integrate it. If the final result is perceived as a useful tool for creating better games, the "garbage AI" label could become just an anecdote from the initial debates.
Timeline, expectations, and room for improvement until launch
Another factor that reduces tension is that DLSS 5 is not yet available to the publicAccording to NVIDIA itself, the technology will not reach gamers until at least the fall of 2026, giving the company ample time to refine both the model and how to integrate it into graphics engines.
This period allows that PC game developers experiment with the toolAdjust parameters, test different styles, and define how you want to use neural network generation in your projects. For European studios, this could be an opportunity to work with NVIDIA to adapt DLSS 5 to the specific aesthetics of their franchises and creative projects.
One of the big questions remains the impact on hardware requirements and compatibility with affordable graphics cardsMany gamers in Spain and other EU countries use mid-range or even lower-end systems, so the success of the technology will largely depend on it working reasonably well on budget-friendly systems. Hardware requirements and optimization will be key.
NVIDIA is aware of this concern and has hinted that the goal is for the final product to be less extreme in resource consumption that the first internal tests, and that they take into account the H20O production and its distribution.
Meanwhile, the debate surrounding DLSS 5 also serves to to open a broader conversation about the role of AI in the future of video gamesThe question is not only whether more spectacular graphics can be achieved, but how it is done, with what degree of control for the artists, and what kind of experiences are prioritized for the end player.
Based on everything known so far, the new version of DLSS is shaping up to be a very big step in the application of AI to renderingcapable of enhancing lighting, materials, and detail without resorting to impossible geometries. At the same time, the initial controversy and the CEO's statements highlight that the public is not willing to accept any innovation simply because it carries the "artificial intelligence" label, and will demand that its real contribution to next-generation games be noticeable and significant.
