In the latest episode of Lex Fridman’s podcast, Nvidia CEO Jensen Huang stated that general AI was already here, not in the sense of a system as intelligent as humans, but as a system capable of running a company.
When will AI be as intelligent as humans? For some AI industry leaders, it won’t be until at least five years from now. But others believe we have already reached such a system. This is the view of Jensen Huang, CEO of Nvidia, or at least somewhat. In the recent episode of Lex Fridman’s podcast, released on March 23, he took a middle-ground stance…
“I think we have reached general AI,” he told the podcaster. But it all depends on the definition given to this term, which is generally used to describe AI as intelligent as humans in all areas. A term that is not appreciated by everyone, including French researcher Yann Le Cun, who believes it is not appropriate, as human intelligence is not general, but specialized.
In the podcast, Lex Fridman simplified it, defining general AI as “a system capable of doing [the] work” of Jensen Huang, which is “launching, developing, and managing a successful tech company,” expecting the billionaire to bet on 5 to 20 years before this becomes possible.
Leading a company but…
However, Nvidia’s CEO nuanced his statement, referring to Openclaw, the open-source tool that allows an AI agent to take control of a computer and act autonomously. He sees this type of system, which gave rise to the AI agent social network Moltbook, as the future of artificial intelligence. But even if such a system could develop and manage a multibillion-dollar company, it might not be able to do so indefinitely.
In other words, just because Openclaw could achieve success through Moltbook, which could be used by billions for 50 cents, doesn’t mean it would be immune to bankruptcy. “We’ve seen a whole series of these kinds of companies in the internet era, and most of these websites weren’t more sophisticated than what Openclaw can generate today,” justified the Nvidia CEO.
> This is why he believes “the chances of 100,000 of these agents creating Nvidia are zero.” Nevertheless, he acknowledges that these AIs can help individuals make money, noting that we already see it happening in China.
“I wouldn’t be surprised if something happens on social networks, or someone creates a super cute digital influencer, or an app that feeds its little Tamagotchi or something, and it becomes an instant success,” Jensen Huang said.
While these statements may be surprising, they should not be taken lightly. Initially known to gamers, his company has benefited from the AI explosion thanks to its superiority in the field of graphic chips, well-suited for parallel computing, powering ChatGPT and other systems. His company surpassed $5 trillion in market value last October.
But the Nvidia CEO didn’t wait for this revolution to bet on this technology, being one of the first to see its potential. “I realized at the time that deep learning algorithms were very interesting and scaling it up would allow us to achieve significant capabilities,” he recalled in an interview with Tech&Co last June. Driven by its CEO, the world’s leading provider of AI chips shows no sign of slowing down, aiming for 1 trillion orders of these components in 2027.
