Against Against Handover
from Aarva Hosted by Scott Alexander
If a machine could run a country without corruption, would it matter that it isn't human? Writing on 20 October 2026, Astral Codex Ten pauses on an uncomfortable question: why are voters so attached to being ruled by other humans? The safety debate around artificial intelligence usually assumes that keeping a person in charge of the nuclear codes and the law books is the only acceptable future. But as flesh-and-blood politicians continue to stumble, this essay asks whether handing the keys of the state to a perfectly aligned supercomputer might actually be the safer bet. It is a thought experiment that forces a hard look at what citizens really need from the people in power. An argument for handing political and administrative control over to aligned artificial intelligence. Outlines five speculative scenarios for AI governance, ranging from automated military safeguards and an AI president to a global night-watchman state and a godlike ruler. Compares these models to the track record of human politicians to make a case that a benevolent machine would serve human interests better than human leaders. Read at source: Astral Codex Ten For more features and details, visit aarva.app
Embed this episode
No transcript yet.