Thought Leadership
Nvidia CEO Jensen Huang says AI safety is a technical problem and does not need new laws.
Nvidia founder and CEO Jensen Huang clearly spoke out against new laws or regulations for artificial intelligence at the Salesforce conference Dreamforce on Tuesday. Contrary to how an OpenAI safety researcher, for instance, had phrased it, AI is not a novel “alien mind” from his perspective, but simply human-built hardware and software – and thus controllable by humans as well as existing laws.
Market forces instead of new laws
“Safety is a technical problem, not a legal one. After all, we build software. After all, we build computer systems. It’s a complicated computer system, but ultimately it’s a computer system.”
Jensen Huang, Nvidia founder and CEO
Huang argued that the free market already exerts sufficient pressure on companies not to bring unsafe products to market. He explained: “If you are not convinced of the safety of a product, then you don’t release it. That is a very obvious way to act.” Companies should take as much time as needed until they are confident of their own safety; the pace of innovation and safe products are not a true tradeoff against each other, but can be achieved simultaneously. New laws or regulations are not necessary for this, according to Huang; it simply requires companies that decide for themselves when to pause if a product threatens to spin out of control.
Debate over self-regulation in the industry
According to reports, Huang did not directly address the question of possible voluntary self-regulation in the industry; instead, he relies more on open AI models with freely accessible weights as a counterweight to closed, proprietary providers. Microsoft CEO Satya Nadella had stated on Monday at the All-In Summit: “China should worry just as much about the same safety issues if the US is worrying about them. Why should it be different for China? It’s not like they don’t have the same hacking problem.” According to TechCrunch, Huang himself had also recently sought direct contact with US President Donald Trump to advocate for a high pace of AI development.
(Editorial Team)
