Ben Goertzel, a leading voice in artificial general intelligence research, is making the case for a decentralized approach to building AGI. He’s pushing back hard on the idea that governments should own or control the technology, arguing that keeping development open and distributed could prevent a handful of entities from dominating one of the most powerful tools ever created.
Why government ownership is a problem
Goertzel sees state-controlled AI as a threat to the very promise of AGI. In his view, centralizing AGI under government authority risks creating a surveillance and control apparatus on a scale no one has seen. He doesn't mince words: handing AGI to governments, he believes, would concentrate decision-making power in ways that are antithetical to the technology's potential for good. The researcher has consistently warned that while governments may claim they need AGI for national security or public welfare, the real danger is that such concentration would lead to an AI elite that could shape society without broad input.
Decentralization as a democratic force
Goertzel argues that a decentralized model could democratize AGI innovation. Instead of a single lab or agency calling the shots, a distributed network of researchers, developers, and even hobbyists could contribute. This isn't just about fairness—it's about speed and diversity of ideas. When more people can tinker with the underlying architecture, the argument goes, breakthroughs happen faster and from more directions. The model echoes open-source software movements, but applied to the most advanced form of artificial intelligence.
Power concentration and its risks
One of Goertzel's central points is that AGI in the hands of a few—whether a government or a large corporation—creates extreme power imbalances. He points out that such concentration could lead to scenarios where the controlling entity uses AGI to entrench its own authority, stifle competition, or manipulate information. A decentralized build, by contrast, spreads that risk. No single node can hijack the whole system. The security argument is straightforward: a decentralized network is harder to corrupt, harder to turn into a weapon of control.
Collaboration over competition
Goertzel envisions a future where AGI development is collaborative by default. Instead of rival labs racing to a proprietary AGI behind closed doors, a decentralized approach encourages sharing of techniques, data, and safety protocols. That cooperation could accelerate progress while building in safeguards from the start. The idea is that collaborative technological progress, guided by many hands, produces more robust and ethical outcomes than a winner-take-all race. Goertzel's vision isn't just technical—it's a statement about how humanity should handle transformative technology.
The debate over who controls AGI is far from settled, but Goertzel's stance offers a clear alternative to the state-centric models gaining traction in some capitals. For now, the question remains whether the research community can actually build a decentralized AGI that's safe, efficient, and truly open. That's the challenge Goertzel and his allies are taking on.
