Dario Amodei
CEO & Co-founder
Anthropic
Former VP of Research at OpenAI who co-founded Anthropic to pursue safer AI development. Author of "Machines of Loving Grace", a landmark essay arguing that transformative AI could compress decades of scientific progress into years.
Current focus
Sep 16, 2026Dario Amodei, along with other AI leaders, is advocating for slower, more cautious AI development and emphasizing safety, while debating the role of government regulation in the field.
Do you agree with this position?
AI-distilled summary of recent news coverage — see sources above.
In the News
There are few players, but potentially plenty of losers, in so-called AI race
San Antonio Express-News · September 18, 2026
Big AI is begging to be regulated. That should worry you. | Opinion
USA Today · September 18, 2026
Opinion: Don’t believe the calls for slower AI development will work. The juggernaut will keep rolling
The Globe and Mail · September 18, 2026
The Awesome & Alarming A.I. Visions of Dario Amodei
RealClearMarkets · September 18, 2026
Alarmist Intelligence: Dario Amodei, Anthropic and the Existential Debate Over AI Doom, Power and Control
Open Magazine · September 18, 2026
Core Positions & Ideas
Constitutional AI: Align Models to Principles, Not Just Human Feedback
2022Anthropic's Constitutional AI (CAI) approach trains models to critique and revise their own outputs according to a written 'constitution' of principles. The key innovation: you can reduce harmful outputs without needing human labelers to review every example. More scalable and more transparent than pure RLHF.
Your take on this position:
Safety and Capability Are Complementary, Not in Tension
2023Amodei's central thesis: the same techniques that make AI systems more capable (interpretability, reliable reasoning, honest communication) also make them safer. Unlike some researchers who treat capability and safety as a tradeoff, Anthropic bets on 'responsible scaling' — moving fast but pausing if safety evals show danger.
Your take on this position:
Machines of Loving Grace — AI Could Compress 50–100 Years of Progress Into 5–10
2024In his landmark essay (2024), Amodei argued that transformative AI could eliminate most infectious diseases, solve mental health crises, and dramatically extend healthy human lifespans — all within this decade. His vision is the most optimistic from a safety-focused AI leader, striking for its specificity about what AI could actually deliver.
Your take on this position:
Responsible Scaling Policy — Pause If Evals Show Danger
2024Anthropic's Responsible Scaling Policy (RSP) commits the company to halting development if safety evaluations show certain risk thresholds are crossed. First major AI lab to publish such a commitment. Critics say it's self-regulated; supporters call it a meaningful step toward accountable development.
Your take on this position:
Essential Reading & Watching
Constitutional AI: Harmlessness from AI Feedback
The paper introducing Constitutional AI. Proposes training AI systems using AI-generated critiques guided by a written constitution of principles, making alignment more scalable and auditable.
Read / Watch →
Machines of Loving Grace
Amodei's most personal and ambitious piece of writing. Describes a world where AI helps defeat cancer, end mental illness, and dramatically compress the pace of scientific progress. Widely read across Silicon Valley and Washington.
Read / Watch →
Anthropic's Responsible Scaling Policy
Anthropic's public commitment to pause development if safety thresholds are crossed. The first such policy from a major AI lab.
Read / Watch →
