A Manhattan Project for AGI

A company can begin in a small room with a few tables, an idea, and just enough money to last a few months. Many important things of the technology age grew that way. The state usually stood outside, waited for the product to take shape, and only then arrived to write laws or collect taxes

AGI may turn that path. When infrastructure costs reach the scale of energy projects, when the secrets inside become intelligence targets, and when a failure can reach far beyond a company’s customers, governments can no longer play the role of spectators. For Leopold, that involvement is not only possible. It is almost a natural consequence of scale

He calls that future shape The Project, recalling the Manhattan Project that once gathered scientists, the military, industry, and national resources to build the atomic bomb. The comparison is not about the kind of weapon. It is about the moment a technology becomes too important to be developed as an ordinary product, forcing the whole state apparatus to reorganize around it

A project like that could do things the market does poorly. The state can mobilize electricity, build infrastructure, protect laboratories, and force parties to coordinate instead of racing blindly. In an emergency, a clear chain of command may also be better than leaving decisions that affect humanity between a board, investors, and a few founders

But concentrated power always carries its own price. A large machine is often slow, easily pulled by politics, and inclined to keep secrets even when society needs to know. A project in the name of safety can become a tool of surveillance. A mechanism built to manage an emergency can stay long after the emergency has passed

Leopold is mainly offering a descriptive forecast: when leaders truly believe AGI has decisive military and economic meaning, they will not let a few private companies choose the path alone. What remains open is the form of involvement. The state can nationalize, sign contracts, form alliances, or build a half-public, half-private structure. The name matters less than the question of who has the right to decide, and who checks the person who decides

He is not sure humanity needs a Manhattan Project for AGI. He only sees that if that day arrives, the debate should not begin on the morning the government has already walked through the laboratory door. Good institutions are rarely built in panic. Are we preparing for how to govern that power, or only preparing more electricity to make it larger

❤️ cowriter aethery