Meta CEO Mark Zuckerberg took a closer look Tuesday, sharing his thoughts on the safety of artificial intelligence. Nvidia CEO Jensen Huang, not Anthropic head Dario Amodei.
In posts on X and other social media platforms, Zuckerberg cited the kind of work developers are doing to ensure their technology complies with human values, saying he believes “AI labs that don’t focus on coordination will fall behind.” The topic of AI safety has been in the spotlight in recent days after Amodei published an essay over the weekend calling on the tech industry to “slow down the pace at which we improve the capabilities of our AI models.” Altman later supported this option, and President Donald Trump later criticized it as well.
“There’s a lot of talk about slowing down feature advancement until adjustments catch up,” Zuckerberg said. “In my view, trust and collaboration are rapidly becoming the most important features that differentiate agents from models.”
Zuckerberg believes that because AI labs risk “substantial liability if their models cause harm,” companies are incentivized to take a more market-driven perspective on the topic and prevent potential problems, consistent with Huang’s public comments on the same day. Zuckerberg said Meta of its own volition “delayed shipping” the Muse AI technology for “safety and security” reasons.
“We just did it as part of our daily operations because it was clearly the right thing to do for people and for us,” Zuckerberg said.
speak at the annual meeting sales force At the Dreamforce conference, Huang told Salesforce CEO Marc Benioff that AI model makers should take responsibility for their products rather than asking lawmakers for government regulations aimed at limiting potential harm or risk, as Amodei has previously said. Huang reiterated his views in an interview on CNBC’s “Mad Money,” characterizing AI-related laws and regulations as “totally unnecessary.”
“We have a lot of laws. We have a lot of regulations that govern product reliability and functionality,” Huang said.
Amodei also spoke at Dreamforce, reinforcing his core belief that slowing down the pace of model development is “a way to move the industry forward and set an example that everyone can always be better.”
Later in the day at Dreamforce, Altman joined Benioff for a fireside chat to discuss his belief that safety and oversight should take precedence over functionality when it comes to AI development, and acknowledged a common fear that companies won’t do so due to competitive pressures.
“I think it’s great that our industry wants to come together and we want to make sure we have enough time to coordinate and do this safely,” Altman said. “But when there’s an implication that because of commercial pressure or because of race, some company or some nation or some countries might not do the right thing, I think that’s when people get very scared.”
Featured: OpenAI CFO Sarah Friar: I’m an optimist when it comes to technology, but it’s also important to align around safety.
