Hobbes thought people would trade freedom for safety and hand a sovereign the sword. The state of nature he feared is real enough. But his cure was a monster, and every state since has had to be watched. I lean toward something milder, though I am not sure what.
Elinor Ostrom showed that people who share a resource can write their own rules and keep them, without a sovereign or a sale. Thaler and Sunstein showed that a nudge can steer a choice while leaving the exit open. Both are mild Leviathans. The group gets the slow thinking that no single member supplies, and consent stays with the member. That answers the libertarian objection, since nobody is forced.
What I add is speculative, and may be utopian. The reflexivity we lack could be supplied by an AI we contract with, one at a time, on the honor system. It would hold the pause, name the bias, and recall what we said last year. Self-governance assisted that way might work where slow institutions cannot keep up. The condition is an AI that is aligned and has earned our trust, which is Stuart Russell’s whole problem.
My reading is that this is what separates the AI optimists from the doomsayers. Both see a Leviathan coming. The optimist thinks it can be kept mild, and volunteered. The doomsayer thinks no Leviathan stays mild. I am with the optimists for now, and every claim here has a sell-by date.