Nvidia just released a video aiming to prove the power of generative AI in video games and it’s certainly interesting… but likely not in the way it was intended.
The video showcases Nvidia ACE, a technology that aims to simulate “intelligent” characters via generative AI. Created through a partnership with Convai, it shows how the player can speak to an NPC named Jin including speech recognition, speech simulation, audio-to-face, and so forth, in order to have an unscripted conversation.
The ramen shop scene is displayed using RTX tech and DLSS 3.0 for additional Nvidia brownie points.
The video claims that what we see is “dynamic and immersive” as the technology “sparks life” into virtual characters, but I’d say that what we see is the opposite of immersive, and I definitely don’t see any “life” sparked into Jin. He’s as robotic as he could be and the interaction really isn’t that impressive. As a matter of fact, this is a video I’d expect to see from someone trying very hard to prove that generative AI does a very poor job of creating immersive and lifelike characters.
At least the shop itself looks nice, but I haven’t ever seen a ramen shop in which an entering customer doesn’t even get a simple greeting. Luckily, real ramen shops aren’t manned by AI bots, at least for now.
This is not to say that generative AI doesn’t have the potential for certain limited applications, but this video hilariously fails at proving it. You can watch it below and see for yourself.
This is just the latest of many rather humorous outcomes of the application of generative AI. You can check out what happened when we asked ChatGPT to review a popular visual novel.
Recently, Take-Two CEO Strauss Zelnick expressed excitement about the potential of generative AI, but also cautioned against overestimating said potential, mentioning that it’ll never replace human genius and “there is no such thing as artificial intelligence.”
