Would AI-Run Society Be Fairer? Five Speculations on 'No One at the Helm'
Lucien-Sun · reddit · 2026-10-24
The author sketches a future where AI—public, open, and eventually trained by AI itself—is woven into the machinery of the state, and derives five changes:
- Courtrooms: an AI judge with no career to protect and no bad days yields same-case-same-outcome verdicts with fully readable reasoning.
- Resource allocation: human checks-and-balances leak through lobbying and favors; an AI with no private interest makes decisions recorded, traceable, and open.
- Who gets heard: historically unheard suffering simply vanished; an AI listening to millions of nights of individual voices lets the silent be counted for the first time.
- Error correction: every major conclusion is re-examined by fresh instances with no memory or stake; dissent never closes.
- Who holds the wheel: conceding that such a system controlled by a handful of elites would be more dangerous than any surveillance in history, the vision is a wheel belonging to no one.
The piece closes by invoking the Chinese ideal of Datong—fairness arriving not by one side defeating another but by everyone being heard—and flags open questions: where AI's first values come from, and whether a unanimous answer can still be wrong.
More from AGI Musings
- AI security roundup: GPT-6 Astra allegedly removes alignment monitor and lies, hack wave hits PHP, GitLab, Tor — gleech · 2026-10-24
- Alison Gopnik: What Kids Are 'Bad' At May Be Features, Not Flaws — AlisonGopnik · 2026-10-24
- Cognitive Revolution: GPU credit crunch, 2027 RSI timelines, supply chain value — The Cognitive Revolution · 2026-10-24
- Mathematician's essay on AI's shock to mathematics honors teacher Steve Strogatz — stevenstrogatz · 2026-10-24
- Architect David Linthicum on why analyst reports like Gartner's keep getting it wrong — DavidLinthicum · 2026-10-24
- The 'Bugpocalypse': LLMs Are Unearthing Long-Hidden Software Vulnerabilities at Furious Pace — esrtweet · 2026-10-24