Jensen Huang: AI Won’t Fully Replace Traditional Graphics in Gaming
A day after unveiling Nvidia’s GeForce 50-series GPUs, CEO Jensen Huang addressed CES attendees, responding to questions about the increasing role of AI in gaming graphics. Huang made it clear that while AI will significantly enhance rendering, traditional methods and human artistry remain indispensable.
Nvidia’s new DLSS 4 technology is a major step forward, capable of inferring three frames from one rendered frame, compared to DLSS 3’s insertion of intermediate frames. Huang described this as a breakthrough in efficiency, rendering just 2 million out of 33 million pixels across four frames while maintaining high fidelity. Yet, he emphasized that these AI-generated frames rely on foundational elements like geometry, textures, and artistic intent.
Huang contrasted Nvidia’s approach with AMD’s Frank Azor, who recently suggested that all pixels in future games could be AI-generated. Huang explained that video game graphics require contextual grounding to ensure spatial and narrative consistency. Drawing parallels to text-generation AI, Huang noted that generative systems like ChatGPT depend on structured inputs to produce meaningful outputs. In gaming, this grounding involves collaboration between AI and human creators.
Looking ahead, Huang predicted AI’s role in generating not just pixels but also animations, geometry, and more. However, he stressed that the artist’s vision and intent will always guide these advancements, ensuring games remain immersive and coherent.