To no one's surprise, NVIDIA CEO Jensen Huang isn't too concerned about a future full of robots and superintelligent artificial intelligence. Moreover, he welcomes it. During a Q&A session at CES with media and analysts, Huang was asked whether he thinks intelligent robots will eventually side with or against humans. “With people, because that’s how we’re going to build them,” he answered confidently.
“The idea of superintelligence is not unusual,” Huang continued. “I have a company that employs a lot of people who are super intelligent in their field of work. I am surrounded by super intelligence. And I'd rather be surrounded by superintelligence than the alternative.”
Given that the hype around generative AI has had a huge impact on NVIDIA's business, it is currently competing with Apple and Microsoft for the world's largest valuation—it makes sense that Huang would be all for a future in which we rely more on smarter AI. While he fails to herald the emergence of god-like artificial general intelligence (AGI) like OpenAI CEO Sam Altman and Tesla's Elon Musk, Huang's vision instead sounds more task-focused.
“This is the future: You will have superintelligent AI that will allow you to write, analyze problems, do supply chain planning, write software, design chips,” he said. “This technology, of course, can be used in different ways, but it is people who cause harm. I think cars are cars.”
During the morning question and answer session that took place after Huang's long talk at CES in front of a largely unenthusiastic audience, he admitted that he had done a poor job of communicating his vision for AI in the real world. Huang believes that the combination of NVIDIA Omniverse technology for rendering 3D programs, as well as NVIDIA Cosmos for creating photorealistic environments for AI training, will make it easier to train robots for real work. Ultimately, this could be one small step toward superintelligent AI—or at least slightly more capable robots.