Kingdom Come: Deliverance Director Says NVIDIA DLSS 5 Will Not Be Stopped by “Haters”

The debate around NVIDIA DLSS 5 continues to divide players, modders, and developers, and now one of the most recognizable voices in European RPG development has stepped directly into that conversation. Daniel Vávra, founder and creative director behind Kingdom Come: Deliverance and Kingdom Come: Deliverance II, shared his view on X and argued that while the earliest examples of DLSS 5 may look uncanny today, the technology’s long term trajectory is far more important than the first wave of backlash. In his view, this is not a gimmick that will disappear under criticism, but a tool that will continue evolving until it becomes too useful for the industry to ignore.

In the post, Vávra said he can imagine a future where developers train the technology around a specific art style or even around particular faces, and suggested that it may eventually replace expensive rendering techniques such as ray tracing in some use cases. He described the current state as “just a little uncanny beginning” and concluded with the blunt statement that there is “no way haters will stop this.” The wording is classic Vávra, direct and provocative, but it also reflects a broader industry reality. Even when new rendering or AI assisted tools arrive with rough first impressions, publishers and hardware vendors rarely abandon them if they believe the long term efficiency and visual upside is large enough.

I can imagine in the future devs will be able to train this tech for particular art style or specific people faces and it might replace expensive raytracing etc. This is just a little uncanny beginning. No way haters will stop this. Its way more than a soap opera effect every tv… https://t.co/SUdxhy6Arj

— Daniel Vávra ⚔ (@DanielVavra) March 23, 2026

What makes his comment more relevant than a typical social media hot take is that NVIDIA has already described DLSS 5 in a way that overlaps with exactly this future facing argument. In a recent Lex Fridman interview, NVIDIA CEO Jensen Huang said the system is open enough that developers could eventually train their own models, prompt the look they want, and even ask for something closer to a toon shader or another stylized visual output rather than strict photorealism. In other words, NVIDIA is not presenting DLSS 5 as a tool limited to making games look more realistic. The company is trying to position it as an adaptable artist facing system that can work across multiple visual styles. That is a critical distinction because one of the biggest early criticisms of DLSS 5 was that its first public examples looked too synthetic and too disconnected from original art direction.

That does not mean the criticism was baseless. If anything, NVIDIA made the launch conversation harder for itself by showing material that immediately triggered concerns about over processing, facial distortion, and what many players described as AI slop aesthetics. Huang has since tried to soften the messaging, stressing that DLSS 5 is meant to be optional and artist controlled rather than a replacement for human visual direction. Still, the reveal clearly damaged first impressions, and that is why the conversation remains so heated. Early demonstrations matter, especially when a technology touches the final image players see on screen.

Vávra’s stance is interesting because it does not really defend DLSS 5 on the basis that it already looks perfect. Instead, it defends the direction of the technology. That is an important difference. He appears to be arguing that the industry should judge the trajectory rather than only the current edge cases. From a development perspective, that is not an unreasonable position. A lot of major rendering advancements looked awkward, overused, or poorly implemented in their early phases before developers learned how to integrate them properly. Motion blur, temporal reconstruction, ray tracing, and frame generation have all gone through that cycle in one form or another.

At the same time, the resistance from parts of the developer and modding community is also understandable. DLSS 5 touches a sensitive issue in game production, which is control over a game’s visual identity. If a system begins generating or heavily altering final frame output in ways that artists did not fully intend, then the concern is not just technical quality. It becomes an authorship problem. That is why this conversation is bigger than performance optimization. It sits right at the intersection of art direction, rendering cost, and AI assisted image generation.

For now, Vávra’s comments add more fuel to a discussion that is not slowing down any time soon. DLSS 5 has clearly crossed the line from being a GPU feature into becoming one of the more contested visual technologies in modern game development. Whether players embrace it, reject it, or eventually stop noticing it altogether will depend less on social media arguments and more on whether NVIDIA and game developers can prove that the system respects artistic intent while delivering meaningful visual and technical gains.

As for Vávra himself, his role around the Kingdom Come franchise is also evolving. Recent reports indicate he is shifting focus toward expanding the IP into live action film or television, a move that has been confirmed through recent coverage around Warhorse’s transmedia ambitions for the series.

Do you think DLSS 5 will eventually become a powerful creative tool for developers, or has NVIDIA already damaged trust too much with the way it revealed the technology?