Confucianism meets Legalism: institutions and virtue are co-dependent for AI agent ecosystems
xuanalogue · x · 2026-09-19
Reading Ken Liu's Dandelion Dynasty led scholar @xuanalogue to argue he's both a Confucian and a Legalist about ecologies of AI agents: ideally most agents are cooperative and virtuous by default, but institutional incentives are needed for robustness against malicious actors. The two are mutually dependent — institutions survive only if enough agents uphold their spirit, and virtuous dispositions are evolutionarily stable only in environments that punish anti-social behavior.
More from AGI Musings
- Amin Karbasi: our paper asking if LLMs could solve easy conjectures 'did not age well' — aminkarbasi · 2026-09-19
- 100+ AI researchers including Hinton call on frontier labs to embed third-party evaluators — burny_tech · 2026-09-19
- Why we never say "the AI is hungry": folk psychology still draws a line at machines — xuanalogue · 2026-09-19
- Science via superhuman AI: shift of paradigm or shift of authority? — AryHHAry · 2026-09-19
- Only a handful of 20th-century philosophers thought AI was possible: Dennett, Chalmers, early Putnam — repligate · 2026-09-19
- After his mother lost her singing voice to a stroke, he's building AI voice tech for healthcare — tzmartin · 2026-09-19