Einstein on politics and physics

Wednesday 12 August 2026

Albert Einstein is often credited with the observation that “Politics is more difficult than physics.” Whether or not the phrase was uttered in precisely those words, the sentiment has endured because it captures an enduring truth. Physics concerns itself with discovering immutable laws governing the universe. Politics concerns itself with persuading imperfect human beings to cooperate, compete or compromise under circumstances in which there are no immutable laws at all. The contemporary world, in which revolutionary advances in military technology coexist with increasingly fragmented geopolitical order, demonstrates the wisdom of that observation more vividly than perhaps at any time since Einstein himself lived.

The twentieth century was an era in which scientific achievement repeatedly transformed the character of war. Einstein’s own work on relativity indirectly contributed to the intellectual foundations that eventually culminated in nuclear weapons, although he himself became profoundly uneasy about the consequences of those discoveries. The scientists who unlocked the atom solved an immensely difficult technical problem. Yet once the bomb existed, the scientific challenge was complete. The political challenge had only begun.

Nuclear deterrence became perhaps the greatest political experiment in human history. It required adversaries to trust that each possessed sufficient rationality not to destroy civilisation. Maintaining that delicate equilibrium demanded diplomacy, communication, restraint and mutual understanding far beyond anything required to build the weapons themselves. The equations governing nuclear fission remained constant. Human judgment did not.

Today’s technological revolution presents an analogous dilemma on a broader scale. Artificial intelligence, autonomous drones, quantum sensing, satellite constellations, cyber warfare and biotechnology are advancing simultaneously. Engineers continue to solve extraordinary technical problems at remarkable speed. Yet political institutions struggle to adapt even to technologies that became commonplace years ago.

Ukraine offers perhaps the clearest illustration of this divergence between technological innovation and political complexity. The battlefield has become an enormous laboratory for autonomous systems. Small, inexpensive drones have altered tactical doctrine more profoundly than many billion-dollar weapons programmes. Artificial intelligence increasingly assists in target recognition, logistics, battlefield mapping and electronic warfare. Commercial technologies originally designed for civilian purposes have become essential military assets.

Yet none of these technological advances has simplified political decision-making. On the contrary, each innovation has multiplied the number of difficult choices confronting governments.

When should autonomous systems be permitted to identify targets without human intervention? How should responsibility be allocated when artificial intelligence makes an operational error? Which dual-use technologies should remain freely traded internationally, and which should become subject to export controls? How should democracies balance military necessity against civil liberties when cyber attacks increasingly blur the distinction between civilian and military infrastructure?

These questions have no mathematical solutions.

They require value judgments, legal interpretation and political legitimacy.

The same pattern is visible far beyond Ukraine. Artificial intelligence promises enormous economic productivity while simultaneously threatening employment, altering social structures and concentrating unprecedented power in the hands of those controlling computational infrastructure. Quantum computing promises revolutionary scientific advances while potentially rendering existing cryptographic systems obsolete. Biotechnology promises extraordinary medical progress while creating new opportunities for biological warfare and genetic manipulation.

The technical trajectory appears relatively predictable. The political consequences do not.

Indeed, technological superiority has often failed to produce political success. Throughout history, militaries possessing overwhelming technical advantages have frequently struggled against politically resilient opponents.

The United States possessed unmatched military technology during the wars in Vietnam, Iraq and Afghanistan. Russia entered Ukraine believing its larger armed forces and sophisticated weapons would guarantee rapid victory. Superior technology proved insufficient because political assumptions underpinning military strategy proved flawed.

Wars are ultimately contests of political will rather than engineering capability.

This distinction becomes even sharper as technological barriers to military power decline. A sophisticated drone costing only a few hundred dollars can destroy equipment worth millions. Commercial satellite imagery is accessible worldwide. Open-source artificial intelligence models increasingly democratise capabilities previously restricted to great powers.

Power therefore becomes less dependent upon exclusive access to technology and more dependent upon political institutions capable of integrating technology effectively.

This may explain why democratic societies often display unexpected resilience during periods of technological disruption. Democracies frequently appear inefficient, argumentative and slow. Yet those very characteristics permit continuous adaptation. Public criticism exposes failures earlier. Independent institutions distribute decision-making. Private innovation flourishes without requiring central planning.

Authoritarian systems may mobilise resources more rapidly, but they frequently suppress precisely the information necessary for effective political adaptation. Leaders surrounded by officials reluctant to contradict them may receive technically accurate reports but politically distorted assessments.

Technology cannot compensate for institutional weakness.

Artificial intelligence itself illustrates Einstein’s insight. Enormous resources are devoted to improving machine reasoning, pattern recognition and autonomous decision-making. These achievements are genuinely remarkable. Yet artificial intelligence cannot resolve conflicts arising from incompatible political objectives.

No algorithm can determine the appropriate balance between liberty and security.

No neural network can establish the legitimate boundaries of national sovereignty.

No optimisation function can reconcile fundamentally opposed conceptions of justice.

Artificial intelligence may inform political decision-making. It cannot replace politics.

Indeed, AI may increase political complexity precisely because it accelerates the speed at which decisions must be made. Autonomous military systems compress decision cycles from hours to minutes or even seconds. Political leaders may find themselves required to authorise responses before fully understanding unfolding events.

Technology reduces reaction times while leaving human deliberation unchanged.

This widening gap between technical capability and political capacity may become one of the defining strategic challenges of the twenty-first century.

Meanwhile, international institutions struggle to keep pace. Arms control agreements designed for nuclear missiles offer only limited guidance for regulating machine learning models. International humanitarian law must continually interpret principles developed for earlier generations of warfare in light of autonomous systems and cyber operations. Diplomatic negotiations move at the pace of governments; software evolves at the pace of software developers.

Scientific progress compounds itself exponentially. Political consensus usually develops incrementally.

Einstein himself understood both worlds unusually well. His scientific genius enabled him to describe the structure of space and time with extraordinary precision. Yet his later years increasingly focused upon political questions: nuclear disarmament, international governance, peace and human rights. He appears to have recognised that understanding nature, however difficult, was ultimately easier than persuading human beings to govern themselves wisely.

That insight remains profoundly relevant.

The greatest dangers confronting humanity today do not arise because scientific progress has advanced too far. They arise because political wisdom has not advanced at the same rate. Humanity has become extraordinarily capable of transforming the physical world while remaining only modestly better at managing disagreement, mistrust and competing interests.

The challenge facing contemporary statesmen is therefore not merely to encourage innovation but to build institutions capable of governing innovation responsibly. Military planners must understand algorithms, but they must understand human psychology even more deeply. Engineers may invent the weapons of tomorrow, but politicians must decide when, why and whether they should ever be used.

Physics describes what is possible. Politics determines what becomes reality.

Einstein’s observation therefore acquires renewed force in an age of artificial intelligence and autonomous warfare. The scientific challenges remain formidable, but they are increasingly yielding to human ingenuity. The political challenges remain as difficult as ever because they concern not the predictable behaviour of matter, but the unpredictable behaviour of ourselves. It is in that uncertain territory, rather than in the laboratory, that the future of international security will ultimately be decided.