The Physicist Who Kept Talking

Einstein published this collection a year before he died, and the organizing principle is less "ideas and opinions" than it is one long argument that the smartest person in any room still has to beg the room to behave. The book sprawls — relativity sits alongside pleas for disarmament, reflections on Jewish identity neighbor tributes to Marie Curie, the economic crisis of the 1930s gets diagnosed with the same cool precision as the equivalence of mass and energy. What holds it together is a voice that refuses to separate knowing from caring. Einstein's central conviction, repeated in different registers across decades, is that science tells us what is and religion (or ethics, or conscience — he uses the terms loosely) tells us what should be, and that neither domain can cannibalize the other without catastrophe. In 1954, this was a distinguished elder's plea. In 2026, after years of AI alignment debates that rehearse almost exactly this partition — capability versus values, optimization versus ought — it reads less like philosophy and more like an engineering specification we still haven't implemented.

His prescience is sharpest where it's least scientific. The essays on nationalism, militarism, and the failure of international institutions could have been written about the UN Security Council's paralysis over Ukraine, or the collapse of arms control frameworks that began unraveling well before 2022. He saw that the League of Nations was a gentleman's agreement with no teeth and said so plainly; he would recognize its successor's identical condition. His analysis of the 1930s economic crisis — that technological productivity gains, unevenly distributed, create structural unemployment and political instability — maps with uncomfortable accuracy onto the automation anxieties and AI displacement debates of the 2020s. He didn't foresee the specific technologies, but he identified the mechanism: when production requires fewer hands, the system must redistribute or fracture. What he got wrong, or couldn't imagine, was the degree to which information itself would become the contested resource. His model of intellectual cooperation assumes that scientists and thinkers, given channels of communication, will naturally form a cosmopolitan counterweight to nationalism. He never anticipated that the channels themselves — social media, algorithmic feeds, state-controlled platforms — would become instruments of the nationalism he feared.

The blind spots are era-specific but not trivial. His Zionism is idealistic and communitarian, advocating Jewish-Arab cooperation in Palestine with a sincerity that now reads as almost heartbreaking given eight decades of subsequent history. He speaks of a cultural and spiritual homeland, not a militarized state, and his explicit warnings against chauvinism within the Zionist movement were ignored by the very project he supported. His socialism is similarly gestural — he wants collective ownership and rational planning but shows little interest in the institutional mechanics, the incentive problems, the information failures that made twentieth-century socialist states what they became. And his "science and religion" framework, while elegant, assumes a version of religion that is essentially ethical humanism wearing traditional clothing. He has almost nothing to say to fundamentalism, to religion as identity politics, to the weaponization of faith that defines so much of twenty-first-century conflict. The God he defends is Spinoza's God, and Spinoza's God has a very small constituency.

What hits differently now is the cumulative weight of a genius pleading for restraint. When Einstein wrote his letters about disarmament in the 1930s and 1940s, the bomb was new and the horror was fresh. Reading them in 2026, after the resumption of nuclear saber-rattling by multiple states, after the development of autonomous weapons systems he never dreamed of, after the quiet normalization of existential risk as a policy variable, the letters feel less like historical documents and more like evidence in a case we are still losing. His insistence that individual conscience must override state authority — that "the minority which consists of men of high moral and intellectual caliber" bears a special responsibility — sounds both necessary and quaint in an era when expertise itself is a target. The book sits in the corpus as a hinge: it inherits the Enlightenment confidence of Kant and Spinoza, the ethical urgency of the early pacifist movements, and it bequeaths a template — the scientist as public intellectual, the physicist as moral authority — that Oppenheimer, Sagan, and every signatory of every open letter about AI risk has since followed, with diminishing returns.

One question the book raises now that it could not have raised then: if Einstein was right that science cannot generate values, only serve them, what happens when the science builds systems capable of generating their own?