Why Jensen Huang, CEO of Nvidia, says now is the moment of generative AI

Nvidia reported this week that revenue for the fourth quarter ended Jan. 29 was $6.05 billion, down 21% from a year ago. But Nvidia CEO Jensen Huang said generative AI is expected to create a significant opportunity that will accelerate later this year.

Nvidia’s stock price surged on the occasion, and Huang noted how generative AI will take center stage at its GTC event the week of March 20. panel on the enterprise metaverse during Nvidia’s GTC event on March 22).

Sharon Goldman of VentureBeat wrote a long piece about how Nvidia dominated the AI ​​market. Goldman said that “the AI ​​hype explosion of 2023, when major language models like ChatGPT and DALL-E 2 brought generative AI into the public consciousness in a way that may not have been seen since the inception of the iPhone in 2007 .”

I spoke with Huang this week and he said the combination of user-generated content and generative AI will help create content for the metaverse at an even faster pace than previously anticipated. We talk about that and other gaming trends in our conversation. I also asked him if he has a chance of landing Nvidia chips in a future Nintendo game console.

Here is an edited transcript of our interview.

GamesBeat: Generative AI is having its moment. You’ve been talking about it for a relatively short time. What convinced you that it is more important?

Jensen Huang: Of course, ChatGPT has only been here for a short time, about three months. But as you know, the industry has been playing around with it for a while when it comes to big language models. If you look at some of the breakthroughs over the last year, whether it’s image generation models, which started with our ProGAN work and now our GauGAN work, all the GAN work that we’ve done and the variable automatic encoders we did. A cousin of the variable auto encoder became diffusion models. The stability of it, the scalability, it’s turned out to be incredible. That all added up last year.

Now we have generative models for proteins. We have generative models for chemicals. We have generative models for language, for text. We have generative models for images and video. As you know, we are working on generative models for 3D. You won’t be able to fill the world’s Omniverse, the metaverse, with human-designed content. It must be observed through computer vision, or generated, or a combination of both.

GamesBeat: What I’ve seen over the past month or so are five different startups combining generative AI with user-generated content in games. Even Roblox showed their demo of it last Friday. It seems that users are not very talented professionals, but if you give them generative AI, they can make things that are usable or playable.

Huang: It’s exactly as you say. You can generate the first version and then let me modify it. I don’t think I can make a model from scratch, but I bet I can modify a model you present to me. It’s no different than people using illustrations in Powerpoint slides. They always combine other people’s work and adapt it. It’s much easier to create something around that. I think it will boost content creation.

GamesBeat: It’s more believable now that we have these different things. We have Omniverse. We have generative AI. We have UGC. It all adds to the metaverse.

Huang: That’s right, exactly. The pieces fall together. It’s very exciting. The characters – you can ask them questions, right? These are characters you can actually talk to. They can talk in different languages. They can understand what you mean. Using fetch models, a company building a medieval game or a sci-fi game or a battlefield game, they can use the entire knowledge base of the story, that game’s story and teach an AI just that. The whole world view of the game AI could be completely medieval and narrowly focused on the gameplay. It will be safe. It will be game specific. We now have all the necessary pieces of technology to help people do that.

GamesBeat: Do you think we’re getting close to a gaming metaverse and an enterprise metaverse benefiting each other through something like Omniverse?

Huang: I’m less sure about that. We don’t spend as much time on the consumer side. But on the industrial side, the energy is very high. With the combination of generative models and proprietary models coming together, you can get started very quickly. The industrial metaverse time is upon us.

GamesBeat: For today’s game results, you noticed that gaming is recovering. It is 16 percent higher than in the previous quarter, but still 46 percent lower. What’s the pattern we’ve seen in gaming with respect to the results you’re reporting?

Huang: We subtract the full COVID time. Say, from the end of 2019, 2020, 2021 and 2022. Game resale averaged about $1.5 billion per quarter that went to COVID. We are in a world today where sales are likely to reach $2.5 billion over the next year. A little bit lower in the beginning as we continue to normalize the channel, then probably higher than that in the back half of the year as we have seasonality. The difference between $1.5 billion versus $2.5 billion, that’s basically, if you will, stripping the pandemic off.

If you look at what’s going on at Steam, it’s pretty much like this. The growth of Steam, the number of active players on Steam kind of reflects that. And then, of course, China is going to recover, which we’re pretty excited about. New games are now being approved. China is recovering. We’ll see how it turns out. I think the gaming market has undoubtedly grown over the past three years.

GamesBeat: Do you think there’s a chance of getting into the next Nintendo console?

Huang: We keep our fingers crossed. We are very good at building energy efficient gaming systems, number one. Number two: we’re pretty sure the next generation of video games will be heavily ray-traced, if not fully ray-traced, and based on generative AI. You know we do a lot of work: RTX is really based on two fundamental technologies: ray tracing and AI. I think you’ll find that the next generation of video games will increasingly use these two technologies. We are incredibly good at that. I hope it happens.