Sixty times. That is how often the word "superintelligence" appears in the 6,000-word essay that Mark Zuckerberg published on Monday — a figure that says as much about Silicon Valley's current mood as it does about any particular technology. Titled "The Future is for Everyone," the post doubles as a manifesto, a product announcement and a direct challenge to the two companies that currently dominate the AI industry: OpenAI and Anthropic. On the same day the essay went live, Meta released Muse Glimmer, a new open-source AI model designed to compete directly with the proprietary systems those rivals have built.

The timing was deliberate. Where OpenAI and Anthropic have argued that cutting-edge models are too dangerous to be released freely, Zuckerberg made precisely the opposite case: that locking powerful AI behind corporate walls creates a far greater risk than sharing it openly. His essay ranged across topics as varied as datacenter infrastructure, bioweapon risks, cybersecurity, surveillance and the disruption of labour markets — a scope that reflected how seriously Meta now treats AI as an existential competitive arena. The core of his argument was about power.

"If the power of superintelligence is held by a small number of individuals, businesses, governments, or AI itself, then that will naturally lead to outcomes that are less favorable for everyone else," Zuckerberg wrote. He proposed letting individual users define the values their AI models hold, rather than having companies impose fixed ethical guardrails — a position that will strike critics in Brussels and Beijing alike as either admirably libertarian or dangerously naive.