Mark Zuckerberg, CEO of Meta, stated that the responsibility for technology safety should lie with each AI company and distanced himself from his competitors' call for a coordinated halt in AI development.
Distancing from Competitors' Proposals
Zuckerberg, in statements published on social media, referred to the proposal by Dario Amodei, CEO of Anthropic, which was quickly supported by tech leaders including Sam Altman from OpenAI and Elon Musk from SpaceX. Warnings from researchers about the potential threats of AI technology have sparked a serious discussion among industry leaders about how to control this technology before it gets out of human control.
Read more: Trump Dismisses Concerns About AI Safety
Responsibility of Each Lab for Safety
Zuckerberg said that AI companies naturally have the incentive to develop their models safely and face "significant responsibility" to prevent harm. He wrote: "Every lab has the responsibility and incentive to move at the pace necessary to train its models safely, and it has the ability to take actions to ensure this."
Zuckerberg made these remarks a few days after Amodei outlined a three-step plan for AI safety. This plan included a proposal for companies and countries to coordinate in slowing down AI development.
Zuckerberg also noted that Meta delayed the release of its AI agent named Muse for several months to ensure its safety and security, stating, "We didn't ask others to do this before us."
Zuckerberg emphasized that companies should prioritize building safe AI models over systems that can self-improve automatically, known as self-improving systems. He wrote: "Allocating a significant majority of processing to serve people instead of competing for self-improvement is one of the best ways to ensure the safe development of this technology."
Read more: China Responds to Anthropic CEO's Request for AI Development Restrictions · NVIDIA CEO Attends State Dinner with Trump and Xi Jinping
Source: abcnews.com
