The Most Powerful Man in AI Isn’t Worried at All

In recent weeks, the CEO of Nvidia, Jensen Huang, has been dismissive of the existential concerns surrounding artificial intelligence (AI). At a Goldman Sachs conference on September 10, Huang accused industry executives of fabricating cybersecurity fears to boost their own products' market demand. He stated that creating a problem is a more effective way to generate demand than genuinely addressing a problem.

This cavalier attitude extends to his interactions with other key players in the AI industry, such as AI developers Dario Amodei and OpenAI's CEO Sam Altman. Neither Amodei nor Altman can afford to alienate Huang, as access to Nvidia's high-end semiconductors is crucial for training their advanced models. Huang's disinterest in AI risks is evident in his past statements.

In 2023, more than 350 industry leaders, including Amodei and Altman, signed a statement acknowledging AI's risk of causing extinction or societal-scale risks, but Huang did not sign. Later that year, Huang told The New Yorker that he had never worried about AI risks, emphasizing that AI is merely processing data. Huang's skepticism about AI risks has grown as concerns about AI's potential impact have spread beyond the industry.

Huang has dismissed calls for regulation, arguing that safety is an engineering problem, not a legal one. He has also lobbied President Trump to lift restrictions on chip sales to China, potentially enabling AI research and associated risks to spread globally. Huang's viewpoint contrasts with that of Amodei, who regularly publishes essays on AI risks.

Huang has never explicitly defended his minority view, but he seems confident in his stance, suggesting that AI will self-regulate and that falling costs will facilitate automated supervision of powerful models.