Mark Zuckerberg says ‘Meta AI could surpass human intelligence and…’

Meta CEO Mark Zuckerberg has outlined a bold vision for the future of artificial intelligence, revealing that the company’s latest AI systems have begun showing signs of self-improvement without human input. In a policy paper published on July 30, 2025, Zuckerberg described this development as the first step toward artificial superintelligence (ASI), a form of AI capable of surpassing human cognitive abilities. He believes this technology could fundamentally transform humanity, driving scientific breakthroughs and empowering individuals in unprecedented ways, while also urging caution in how these powerful systems are released.

Mark Zuckerberg and Meta AI’s journey toward self-improvement