1 hour ago
39
Mark Zuckerberg wants to give every person on Earth their own superintelligent AI. Not sell it. Give it away, or close to it.
In a roughly 6,500-word essay published on August 10, titled “The Future Is for Everyone: The Path to a Positive AI Future,” the Meta Platforms CEO laid out a vision where billions of people and small businesses gain access to personal superintelligence, powered by open-source models, running across devices from phones to smart glasses.
The $145 billion bet
Meta’s 2026 capital spending budget stands at $145 billion, with the bulk flowing into data centers and AI infrastructure.
A significant chunk of that strategic commitment materialized through Meta’s $14.3 billion acquisition of a 49% stake in Scale AI, the data-labeling and AI infrastructure company. That deal also birthed Meta Superintelligence Labs, a new division led by Scale AI CEO Alexandr Wang.
The groundwork was visible as far back as July 30, 2025, when Zuckerberg published a letter declaring that developing superintelligence was “now in sight.” The August essay expanded that claim into a full thesis: personal superintelligence isn’t just possible, it’s the morally correct path forward.
What “personal superintelligence” actually means
Zuckerberg’s version of superintelligence isn’t the sci-fi scenario where one omniscient AI sits in a government bunker making decisions for humanity. His pitch is closer to a hyper-capable personal assistant that deeply understands your goals, preferences, and context, then acts on them across every device you own. The AI would live on your Meta smart glasses, your phone, your laptop, and would be, in Zuckerberg’s framing, working for you rather than for a centralized institution.
The essay draws a sharp line between Meta’s approach and what Zuckerberg characterizes as the competition’s model: tightly controlled AI systems developed behind closed doors by a small number of powerful organizations. His argument is that concentrating superintelligent capabilities in the hands of a few entities creates risks that outweigh the dangers of broad distribution.
Open-source is the vehicle for this philosophy. By releasing models that anyone can inspect, modify, and deploy, Meta positions itself as the populist alternative in an industry increasingly defined by access gates and API pricing tiers.
The distributed power argument
Zuckerberg’s essay repeatedly returns to a core thesis: distributed power is safer than centralized control. His logic: when power is distributed, no single actor can cause catastrophic damage, and the collective wisdom of billions of users creates a natural check on misuse.
What this means for the competitive landscape
Meta’s strategy creates an interesting dynamic in the AI race. OpenAI, Google DeepMind, and Anthropic have all pursued models where access is mediated through APIs, subscriptions, and enterprise contracts. Meta is essentially saying: we’ll build something just as powerful and hand it out.
The $14.3 billion Scale AI investment signals that Meta isn’t just talking. Alexandr Wang, who built Scale AI into one of the most important data infrastructure companies in the AI ecosystem, now runs a lab with a singular mandate: build superintelligence and distribute it.
Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy.
English (US) ·