Zuckerberg's AI manifesto pitches 'personal superintelligence'

Mark Zuckerberg published a 6,500-word essay titled 'The Future Is for Everyone,' pitching Meta's answer to OpenAI and Anthropic: a world of 'personal superintelligence' in which every user controls their own AI agent aligned to their goals. The Verge read the manifesto as a dim vision, and it is not the only outlet doing so. Two experts in our Who's Who directory have shared the piece as well.

The core argument is that concentrating advanced AI inside a handful of companies is itself the danger. 'The notion that AI is so dangerous that the only safe path is an extreme concentration of power seems inherently problematic,' Zuckerberg writes. He frames the philosophy as 'individual empowerment as the source of prosperity, invention as the primary purpose of superintelligence, and balance of power as the foundation of safety.' Superintelligence, in his telling, is an 'invention tool' rather than a labor replacement.

The specific promises land in a lofty register. 'Everyone will have a personalized tutor and coach with a PhD in every subject and unlimited patience,' the essay says. Zuckerberg goes further and asks readers to 'imagine everyone has a superintelligent lawyer,' arguing that in such a world 'justice would be carried out more fairly.' Meta says it will offer basic versions of its products free, with advanced capabilities priced via what the essay describes as a 'dynamic auction mechanism.'

That is where the reception gets rough. In a TechCrunch piece, Russell Brandom argues the manifesto trades in 'hazy generalities' while glossing past the documented ways Meta's products have shaped daily life, and against a backdrop in which existing AI tools in schools mostly enable cheating rather than tutoring. Brandom contrasts Zuckerberg's tone with Sam Altman at OpenAI and Dario Amodei at Anthropic, who at least name the dangers alongside the pitch.

Several concrete questions the pitch invites are left dangling. The essay does not spell out what 'free or affordable' actually costs, how the auction mechanism prices frontier capability, or which future models Meta will release under open weights versus keeping closed. Underneath all of it sits the trust question: the same company promising a personal AI aligned to a user's goals is the company whose consumer products already draw sustained criticism for their real-world effects.

If the plan lands, it reshapes the access story for the AI era, since billions of Meta app users would touch a capable model for free. If it stalls, it will be because the audience is unwilling to lend Zuckerberg the optimism the pitch requires.