While Dario Amodei (Anthropic-Claude), Sam Altman (openAi-ChatGpt) and Elon Musk (xAi-Grok) ask to slow down, Mark Zuckerberg (Meta) and Jensen Huang (Nvidia) say they are already responsible enough and indeed invite us to run. The debate on the dangers of artificial intelligence has now involved the main global big tech companies. Corporate leaders were divided as US President Donald Trump stressed the importance of maintaining technological dominance over China.
Zuckerberg and Huang’s theses
Mark Zuckerberg couldn’t help but intervene on the topic. Meta’s CEO summed up his thoughts in a post on “There is a lot of debate about slowing progress on capabilities until alignment reaches the same level,” Zuckerberg summarized and then expressed his opinion, “trust and alignment are quickly becoming the most important capabilities that will differentiate agents and models.”
According to Meta’s CEO, companies “face significant legal liability if their models cause harm, so they have a strong incentive to prevent this.” The risk of economic loss would act as a deterrent to any releases of unsafe models. “Meta delayed the launch of Muse for several months to focus on safety and security. We didn’t ask everyone else to do it before us. We just did it as part of our daily work,” he recalled.
On the same line is Huang, head of Nvidia. From Salesforce’s annual conference, the CEO had his say on the issue. Confronting Amodei on stage, Huang acknowledged that safety “is the number one priority.” But for Nvidia’s CEO it is “an engineering problem”. If a company is not sure about a product – as Zuckerberg stated – it must not put it on the market. This is why he said that “we don’t need new laws”. According to Huang, pitting speed against security is “a false choice. We can absolutely have both.”