Zuckerberg Wants Superintelligence for Everyone — But Only Big Tech Can Afford to Build It

Mark Zuckerberg is warning that humanity could face a dangerous future if artificial intelligence becomes concentrated in the hands of a few powerful institutions. His proposed solution is to put superintelligence directly into the hands of billions of individuals, giving people unprecedented capabilities to work, create, learn and compete. But behind that promise lies a striking contradiction.

Zuckerberg is making the argument as the billionaire CEO of Meta, one of the handful of technology giants with the money, infrastructure and global reach required to build AI at this scale. His vision promises to decentralize technological power while relying on Meta to help make that decentralization possible, raising questions about who would ultimately control the technology.

In his manifesto, Zuckerberg argues that advanced AI should primarily strengthen individuals rather than corporations, governments or a technological elite. «The defining questions of our age are who will have access to superintelligence and what will we direct it towards. Will it be centralized and restricted to a few institutions, or will it be a tool that empowers everyone?» he writes.

Zuckerberg proposes a balance of power in which widely distributed AI prevents individual institutions from becoming dominant. «The key to a positive future for everyone is achieving a balance of power that favors individuals. The solution is to ensure that superintelligence is broadly distributed to empower people,» he writes. But his next step places his own corporation directly inside that solution.

«Meta is the company primarily focused on building personal superintelligence for everyone,» Zuckerberg declares. His philosophy therefore depends on an unusual proposition: one of the world’s largest technology corporations could become a central force in preventing AI from becoming excessively centralized, even as Meta competes to develop and distribute some of the technology involved.

Zuckerberg imagines personal AI becoming deeply integrated into everyday life. «Everyone will have an exceptionally capable personal agent that understands you, your goals, and everything you care about,» he predicts. The agent could assist with relationships, health, careers, finances, household management and hobbies, giving its provider a potentially important role in the infrastructure surrounding everyday decisions.

Meta plans to offer free versions of its technology to billions of people, while Zuckerberg also promises strong privacy protections. Such access could dramatically broaden the availability of sophisticated AI. But billions of people using superintelligence would not necessarily mean billions controlling it. The models, infrastructure and enormous computing resources behind those systems could remain concentrated among relatively few companies.

That distinction becomes particularly important because frontier AI requires extraordinary amounts of capital, specialized chips, electricity, data centers and technical expertise. Meta expects capital expenditures of $115 billion to $135 billion in 2026, partly reflecting investments supporting its superintelligence ambitions. Such costs leave only a limited number of corporations capable of competing at the highest level of AI development.

Zuckerberg predicts that personal superintelligence could help individuals create businesses and develop new skills, potentially producing more employment rather than less. He also anticipates smaller companies capable of operating at significant scale with relatively few people. Whether enough new businesses and occupations emerge to counter displacement from increasingly capable automation, however, remains uncertain.

The concentration extends beyond money. Major technology companies already control important consumer platforms, cloud infrastructure and other resources necessary to develop and distribute AI. If billions eventually rely on a small number of assistants to obtain information, learn, work and make decisions, the companies developing those systems could retain considerable influence over their design, safety boundaries and access, even while empowering individual users.

That leaves Zuckerberg’s manifesto facing its central contradiction. He wants superintelligence to prevent governments, corporations and AI systems from accumulating overwhelming power, yet delivering that future requires infrastructure affordable primarily to some of the world’s wealthiest corporations. Giving billions access to AI could democratize what individuals can accomplish, but it does not necessarily democratize ownership. The companies building that intelligence could remain among its most powerful gatekeepers.