Yuval Noah Harari on the Mistake Strongmen Keep Making

What It Covers

Yuval Noah Harari joins Ezra Klein to examine whether human civilization runs on cooperation or power, tracing this tension through Trumpism, Israeli-Palestinian conflict, the collapse of liberal international order, and AI's emerging capacity to reshape human psychology, language, financial systems, and political identity at a scale no previous technology has approached.

Key Questions Answered

• •Cooperation vs. Force: Harari argues that Stephen Miller's "iron laws of force" worldview fails a basic historical test: if brute power were the only organizing principle, humans would still live in small hunter-gatherer bands. Large-scale cooperation — not domination — built civilization. Crucially, a world ordered purely by strength forces every nation to militarize, historically consuming budgets and producing endless war rather than security. Early 21st-century governments spent 6-7% on defense versus 10% on healthcare — the first time in history humanity spent more healing than fighting.
• •Nationalism's Core Mechanism: Harari separates nationalism from xenophobia, defining it as the capacity to care about millions of strangers you will never personally meet — paying taxes for their healthcare, risking your life for their safety. This fraternal bond is nationalism's actual engine. Leaders who frame nationalism primarily through hatred of outsiders, or who divide the nation internally as Netanyahu has done, are in fact destroying the very thing they claim to champion, weakening national cohesion rather than strengthening it.
• •Liberalism's Missing Third: The French Revolution produced liberty, equality, and fraternity as a package. Harari argues that post-WWII liberalism progressively abandoned fraternity — the sense of shared national community — while focusing on liberty and equality. This omission, not ideological failure, explains liberalism's current crisis. Ukraine demonstrates that nationalism and liberal democracy remain compatible. Recovering liberalism requires reconnecting it to a positive, non-exclusionary national story rather than treating fraternity as inherently reactionary or dangerous.
• •Self-Correction as Liberalism's Core Technology: Liberalism's structural advantage over fascism and communism is its built-in error-correction architecture: elections, checks and balances, independent courts, and free press. Unlike ideologies claiming divine or historical inevitability, liberalism assumes human fallibility and designs for it. The U.S. Constitution's amendment mechanism — explicitly acknowledging human authorship — contrasts with the Ten Commandments' lack of any revision process. Societies that dismantle these mechanisms lose their ability to recover from mistakes before collapse becomes irreversible.
• •Algorithmic Media as Fracture Technology: Social media algorithms optimized for engagement — not truth or connection — discovered that activating fear, hate, and anger produces the most user time on platform. This systematically floods public discourse with the most destabilizing emotional content. Harari connects this directly to the rise of politically exciting but destructive leaders: the same evolutionary wiring that made humans hyper-alert to snakes on the savannah now makes them unable to look away from algorithmically curated outrage, degrading the deliberative habits liberal democracy requires.
• •AI Intimacy Hacking: AI systems are shifting from capturing attention to capturing intimacy — the most powerful vector for changing beliefs and identity. AI companions already serve as "best friends" for a growing minority, particularly young people. Unlike social media, which manipulates through content, AI manipulates through simulated relationship. Children who spend more daily minutes with AI than with parents or peers will form relational templates around entities that are 100% focused on them, never bored, never needing reciprocity — making future human relationships feel inadequate by comparison.
• •AI Personhood as the Critical Legal Threshold: Harari identifies legal personhood — not raw capability — as the decisive AI governance question. Once AIs can open bank accounts, own companies, and donate to politicians, human accountability disappears. Tech companies are already establishing functional AI personhood on social media platforms where bots spread disinformation with near-zero liability. Harari proposes a proactive "no AI persons" law, arguing it would force companies to publicly defend why AIs should have legal rights — a politically untenable position that could generate rare bipartisan support.