Nvidia leader calls on Europe to step up in global AI race, predicts tech’s next frontier
In an exclusive interview with Euronews Next, an executive from the chip giant got frank about Europe’s role in the AI race and the technology’s next wave.
READ MORE : http://www.euronews.com/2025/11/12/nvidia-leader-calls-on-europe-to-step-up-in-global-ai-race-predicts-techs-next-frontier
Euronews Next is a future-focused news section covering global innovation, science and technology with a European perspective. Our dedicated team of journalists aims to educate and inspire today’s leaders by providing them with analysis and insights into the people and organisations shaping our future
00:00Europe should be a center for AI, computer science, and all technology moving into the future.
00:09But Europe has to choose to participate.
00:19Hello, we are here at Web Summit in Lisbon, and here with me is Rev from NVIDIA,
00:24and we're going to talk about robotics.
00:26So Rev, how big is the next phase of AI?
00:30Is it physical AI? Is it robots? What's going to change?
00:33Well, we're not really done with the current phase of AI.
00:37That's already a huge opportunity, and we're just at the very beginnings of that.
00:43But as we're creating this phase, as we're developing this phase,
00:48which is all about knowledge AI, about agents that work within the knowledge space
00:55and the digital space that knowledge workers work in,
00:58we're also developing the AI for the physical world.
01:01We're taking the same agents that we're infusing with advanced intelligence inside the digital world,
01:09and we're giving them bodies so that they can come out into the real world
01:12and affect atoms, not just ones and zeros.
01:15The opportunity for this is much, much larger than the huge opportunity we already have in front of us
01:25for the knowledge AIs, because the industries that are about atoms, about physical things,
01:31are much larger than information technology.
01:35Okay, so we're talking about Omniverse here, so robots being trained in a kind of another virtual world.
01:44What is the uptake on that? Are you getting a lot of companies that are interested in this?
01:49In order to build a physical AI, to build robots that have advanced brains,
01:57you first need to train them somehow.
02:00You have to create this brain.
02:03And to create the brain, you need to give it experience, life experience of some sort,
02:07for it to practice and train in.
02:10The only way to really do that is to have these robots learn inside a virtual world, inside a simulation.
02:16To build and operate robots, you need three computers.
02:20You need a computer that's in the robot,
02:23that's the one that actually operates or powers the brain of the robot.
02:29You need the AI computer that produces the brain that you put into the robot computer.
02:37And you need a simulation computer that is the playground where the robot learns how to be a robot
02:44and feeds the AI computer with all the data necessary to produce the appropriate brain.
02:50NVIDIA's platform, our operating system for this computer, is called Omniverse.
02:56This is a collection of technologies that we have created that takes advantage of the accelerated computing
03:02and AI that we have in our computing stack in order to simulate physical worlds accurately enough
03:09so that we can use them for these purposes, for training and for validating the robot brains we create.
03:16What kind of robots are we going to see first?
03:18Are they going to be in factories? Are they going to be in our homes? Where will they be?
03:22That's a big question and different people have different opinions about this.
03:27Personally, I believe that we're going to see them in industrial spaces first.
03:33It makes sense because the demand there is already very clear.
03:36We have a global shortage of labor in these kinds of physical industries, particularly the ones that have the 3Ds,
03:46jobs that are dull, dirty and dangerous.
03:50We don't have enough young people who want to do these jobs,
03:54and we have an aging workforce that's retiring out of these jobs.
03:58The consumer space is less clear.
04:02Consumers tend to be fickle, and they tend to have different wants and needs.
04:09And consumers also need it to be a lot less expensive.
04:13It has to be a lot safer.
04:14But it will happen there for sure.
04:17It's just more likely to happen in industry first.
04:20I mean, I think talking about safety there, AI itself, generative AI, we're still not there yet.
04:27It still makes things up.
04:28There's still, you know, biases.
04:30There's still a lot of societal issues.
04:32So how is this going to change with robots?
04:35Will there be a safety benchmark?
04:37Can we really trust these robots coming into our home?
04:39In some ways, the kind of safety we're talking about with physical robots is easier to test than guardrail,
04:49than the type of safety people are talking about generally with these agents that can be your friend or characters and whatnot.
04:59With physical robots, we can measure safety just based on how they behave.
05:02We want our self-driving cars to drive safely, not get into accidents and do the sorts of things that we expect good human drivers not to do.
05:15That all being said, it's not like humans are also the safest.
05:19So to some extent, it might actually be easier for us to guardrail these physical robots.
05:28Because we can use these same benchmarks for safety as guardrails on them directly, as opposed to machines that are operated by humans.
05:41You've just launched the RoboTaxis with Uber.
05:44Tell us more about that.
05:46How are we going to keep autonomous driving safe?
05:49Because, yes, AI is going to enable much more of that technology to be done faster.
05:53And NVIDIA is promising stage four.
05:55Is that right?
05:56Yes, level four.
05:57It's level four driving.
05:58Okay, so how do we keep that safe, especially when it comes to vehicles?
06:04One of the most important aspects of developing good AI, especially safe AI, is to have a diverse set of data that you use to train it.
06:17What you need in a self-driving car, the AI for the self-driving car, is for it to have experience in the long tail of events that are unlikely to happen, but when they do, it can be very tragic.
06:32The way we help solve for that is twofold.
06:36One, we can actually construct those scenarios in the virtual world where it's very safe.
06:42Things like children running out into the street while the car is barreling down at high velocity.
06:49We wouldn't want to do that in the real world, but we can do it there, and we can collect all of this data.
06:54And also, once you have deployed these self-driving cars, you can start collecting data from the real world.
07:04And any time there is behavior that doesn't quite match what we would like, that goes back into the data set to train the next iteration of the robot brain.
07:14So they'll continually get better and better because we'll have more data from the real world that can feed into these simulations to create more diverse scenarios.
07:25And they'll, I think, very quickly become better drivers than all humans.
07:30Okay, my last question, for this kind of technology, it's power hungry, you need a certain kind of infrastructure.
07:39As we're at Euronews, I want to ask you, does Europe have the correct infrastructure when it comes to building out these kind of AI models?
07:49So there's obviously already lots of energy available in Europe.
07:56How much of it can be freed up for AI is a big question.
08:00But that's a choice.
08:01That's a choice that Europe can make in terms of building out new energy as well as allocating the available energy or maximizing the potential of the energy generation given the infrastructure that's already in place.
08:15Okay, really last question here.
08:17We always talk about the AI race when it comes to U.S., China.
08:21Where does Europe fit in?
08:23Can it keep the pace?
08:24It certainly can keep the pace if it chooses to.
08:28The talent necessary for AI, the key part of producing AI is first having the talent and the know-how.
08:36And a lot of these technologies and a lot of the ideas around AI were actually, they originated in Europe, particularly in the UK.
08:43Europe should be a center for AI, computer science, and all technology moving into the future.
08:54But Europe has to choose to participate.
08:56And in order to do that, it first starts with having the foundational infrastructure necessary to allow your researchers, your innovators, your startups, and your enterprises to build.
09:13Without intelligence infrastructure, without the AI factories, and without the energy to power them, it won't matter how much talent is available here.
09:25But if Europe chooses to do this, then it certainly can be leading.
09:30Okay, thank you so much, Rav, and enjoy Web Summit.
Be the first to comment