Mark Zuckerberg defends an AI that is widely disseminated rather than concentrated in a few hands, and promises "personal superintelligence" for everyone.
Meta's boss published on August 10 a long text presented as his most accomplished vision of the future of AI, halfway between the corporate blog post and the manifesto. Zuckerberg challenged the idea, suggested by actors such as OpenAI or Anthropic, that an extreme concentration of power would be the only sure path: in his view, a wide access to models would instead diffuse power and stimulate competition.
He developed the concept of "superpersonal intelligence", an AI that would multiply the abilities of each individual rather than replace human work in bloc, and imagined an education where everyone would have a personalized tutor as patient as well as expert. It acknowledges, however, that if automation goes faster than the ability of workers to convert, a "hard period" could follow — a nuance that has fed criticism of a text judged by some too optimistic about the social consequences of AI.
Read the full article on the original source
Go to the source →