Dario Amodei
LII-D "設計者" AI Researcher · Italian-American · 21st c.Italian-American AI researcher and entrepreneur (1983–). Having worked in Google's Brain team, he founded Anthropic with AI safety as the top priority. Creating Claude, he develops AI based on "Constitutional AI." The most serious voice in the AI safety movement.
主導機能+Ti-p (Organization & Law)
"Constitutional AI" — logically systematising the value system and behavioural rules AI must uphold: the world's first "value design for AI." The "Responsible Scaling Policy" as institutional architecture.
創造機能-Ne-c (Common Sense & Peace)
Choosing staged, cautious implementation over radical AI deployment — institutionalising AI safety as common sense. Prioritising safety over competition as an expression of -Ne-c.
脆弱機能1+Se-p weak (Achievement & Protection)
Weak +Se-p: known as "quiet and introverted" — fundamentally uncomfortable with heroic, front-stage charisma. "The person who quietly designs institutional frameworks."
脆弱機能2-Fi-c weak (Sincerity & Reconciliation)
Weak -Fi-c: almost no interview statements about personal feelings. Speaks about "AI risk" but not personal inner life.
クアドラ・気質・クラブ
クアドラ: Anti-Gamma Quadra (Utopia) — "AI should benefit all of humanity" — service to humanity rather than a specific company as -γ values.
気質: Balanced-Stable temperament: maintaining principles of institutional design without being pulled by competitors' aggressive moves.
クラブ: Researcher Club: Anthropic's internal culture is the Researcher Club itself — papers, discussions, and systematic analysis at its core.
世界観・変化への態度
"The existential risk of AI development without adequate safety exists" — the direct statement of present reality. The "absent reality" (AI already made safe) is not proclaimed.
変化への態度: Executing Anthropic as a realistic plan for safe AI transformation — a practitioner of staged, failure-resistant safety-first development.