“Friction is the difference between war on paper and war as it actually is.”
Carl von Clausewitz, On War
From the start of his war against Iran, US President Donald J. Trump has misunderstood nuclear war and nuclear deterrence. Conspicuously, he has assumed, without tangible evidence or sound logical argument, a direct correlation between the magnitude of his nuclear threats and the likelihood of successful dissuasion. Among other things, this assumption has ignored the potentially plausible counter-argument that threat credibility could sometimes vary inversely with threat destructiveness.
There is more. It follows, inter alia, that Trump’s explicit threats at the UN General Assembly (September 2026) to “annihilate” Iran were at a minimum injurious to American national security. Beyond any reasonable doubt, the American president’s lawless declaration, flush with criminal intent (“mens rea”), will generate, accelerate and/or enlarge military operations against the United States.
There are multiple particulars. The absence of reason-based US military policies signifies Trump’s visceral embrace of self-delusion and anti-reason. From the clarifying standpoint of strategic thought, such postures reflect Carl von Clausewitz’ concept of “friction.” Recalling On War (1832), they reflect the difference between “war on paper” and “war as it actually is.” Today, at a unique historical moment when proliferating nuclear weapons coincide with autonomous superintelligence and recursive AI self-improvement, this difference could more easily produce irremediable harms. In essence, we are living at a moment when the traditional military idea of “friction” has been “AI-enhanced.”
For Washington it’s not just about Iran. North Korea remains on a determinedly belligerent trajectory with the United States. Though militarily inferior to the US, the “hermit kingdom could still represent an unpredictable and fearsome adversary.
How might a US-North Korea war actually begin? One increasingly compelling scenario would engage active North Korean military support for Iran in a war re-started by new Trump attacks against the Islamic Republic. In a discernibly worst case scenario, North Korea would act as Tehran’s nuclear proxy, a bewildering stance that could produce irreversible American losses during a mutual (US and North Korea) search for “escalation dominance.”
To provide more meaningfully-precise clarifications, history will deserve pride of place. At the Singapore Summit on June 12, 2018, Donald Trump promised Americans that North Korea was no longer a nuclear war threat. This was the case, reassured the president in his first term, because he and Kim Jung Un had “fallen in love.” In other words, Trump reasoned openly, “attitude” is more important than “preparation.”
The still-dissembling Trump presidency reveals rising risks of military confrontation between Washington and Pyongyang and also between Washington and other adversaries. Whether or not such confrontations would result in a nuclear war, however, cannot be calculated scientifically. In logic and mathematics, determinations of probability must always depend on the frequency of pertinent past events. Prima facie, there have been no such events.
That’s a good thing. Nonetheless, the United States still needs to render informed judgments on the plausibility of a nuclear war in this and other potential theatres of military conflict. To a significant but indeterminable extent, these judgments will depend on whether the belligerent risks would arise from deliberate or inadvertent nuclear actions and escalations.
For the American president, though there will be variously consequential nuances, it is unlikely that he (or his designated counselors) would recognize and acknowledge them. To wit, in an implausibly optimistic view, Trump’s advisors would learn to understand that an inadvertent nuclear war with North Korea could be the result of decision-maker miscalculation, enemy hacking, IA-related computer intrusion or mechanical/electrical/computer accident. In the final analysis, while such risks could be traceable expressions of individual leadership personality, they are most apt to stem from generically corrosive competition for “escalation dominance.” In part, at least, these risks would reflect the continuously “Westphalian” structure of world politics, an inherently-unstable 17th century foundation (1) that could at any point descend into an unprecedented chaos; and (2) that could coincide with previously unimaginable “equalizer” capacity. Most worrisome for the United States would be bestowals of this capacity on non-nuclear adversaries such as Iran.
In facing newly emerging strategic threats, it is high time for candor. President Donald J. Trump’s expansive military policy misunderstandings are already setting the stage for future and unique nuclear crises with North Korea and other adversaries. Here, core expressions of adversarial “uniqueness” will lie in the “weaker” state’s ability to offset the “stronger” state’s “normal” nuclear advantage. Such ability, in turn, will stem from a new capacity for AI-enhanced intrusions that could (3) compromise American command and control communications and/or (4) immobilize US financial, medical and information networks.
Invariant Obligations of Correct Reasoning
During any upcoming crisis negotiations with North Korea, Iran, China or other adversary, Trump should take scrupulous care not to exaggerate or overstate America’s military risk-taking calculus. In part, such diplomatic caution would derive from the absence of analytically-comparable nuclear crises. In world politics, as in every other topic of human interaction, probability judgments can never be derived ex nihilo, out of nothing.
This does not mean, however, that Trump’s senior strategists ought to steer clear of dispassionate assessments regarding nuclear costs and risks. It does mean that such assessments ought to be systematically drawn from constantly shifting geopolitical trajectories. Such hard-to-decipher trends should include variously complex considerations of worldwide nuclearization. Though set back tangibly by US and Israeli preemptions, Iran has quietly recommitted to certain modified forms of nuclear weapons development. Such conflict could also become plausible via surrogate nuclear enhancements (e.g., if Tehran decided to call upon Pyongyang as its belligerent nuclear proxy), but Iran could prevail even without any nuclear weapons or doctrine.
All this would owe to the Islamic Republic’s equalizing capacity to inflict AI-enhanced harms on the United States. Significantly, these harms could involve biological warfare attacks on American cities, and would likely include variously resistant forms of plague. In science-based scenarios, the human and material outcomes would be as devastating or more devastating than they were during naturally-occurring medieval outbreaks.
Intersections and Synergies
There is more. In all such complex assessments, the geo-strategic world should be examined as asystem. For Trump policy-planners focused on Iran, North Korea or even China, there will be variously fixed obligations to consider, including intersections between Pyongyang’s nuclear escalation threats and Tehran’s clandestinely-restarted nuclearization. These obligations could shape potentially synergistic intersections. In these force-multiplying intersections, the “whole” of any particular conflict outcome would be greater than the sum of its “parts.”
Such many-sided strategic issues ought never to be approached from the obfuscating perspectives of American domestic politics. In further clarification, some component problems and crises would emerge as more complicated and problematic than others. In deference to essential intellectual background, world security processes should always be approached by Washington as a totality. What is happening in such far-flung places as India-Kashmir, China-Taiwan, Russia-Ukraine, Iraq, Iran, Pakistan, Afghanistan and Yemen could at some point have “spillover effects” in northeast Asia and beyond. These effects could be of highest possible importance.
“My Button is Bigger Than Yours”
In one form or another, military threats from already-nuclear North Korea remain substantive and “robust.” President Donald Trump’s first term metaphors notwithstanding, the fact that his nuclear “button” is “bigger” than Kim’s is anything but meaningful. In strategic deterrence relationships, a condition of relative nuclear weakness by one of the contending adversarial states need not imply any corollary diminution of military power. Even the presumptively weaker party in such asymmetrical dyads could deliver “unacceptable damage” to the stronger.
Though bewildering by definition, complexity will always be crisis-defining. President Trump will need to bear in mind that many of Asia’s continuously transforming military developments could be impacted by radically changing US-NATO relationships. Regarding Russia’s war of aggression against Ukraine, it has long been apparent that Trump is more sympathetic to dictator Vladimir Putin than to victim Volodymyr Zelensky. In this connection, nothing was changed by Trump’s September 24, 2026 agreement to Patriot missile licensing for Ukraine.
There are related issues. Important here will be Trump’s willingness to acknowledge the consequential limits of “expert” military advice. Widely unseen, these limits are not based on any presumed intellectual inadequacies in the United States, but on the irrefutable knowledge that no person has ever fought in a nuclear war. In scientific terms (i.e., from the standpoint of disciplined probability theory), this bit of knowledge ought never to be underestimated or disregarded.
Going forward with time-urgent considerations of Iran, North Korea, China and Pakistan, American strategic calculations will be fraught with uncertainties. Still, it will be necessary that Donald Trump and his designated counselors remain able to offer best available war-related estimations. Among prospectively causal factors – some of them overlapping, interdependent or “synergistic” – the risks of a nuclear war between Washington and certain adversaries will ultimately depend on whether the conflict would be intentional, unintentional or accidental. In principle, this tripartite distinction could prove vitally important to any hoped-for success in US nuclear war prediction and prevention.
There is more. In facing future Iran, North Korean and Chinese negotiations, it will be necessary for competent US policy analysts to capably examine all foreseeable configurations of nuclear war risk. Expressed in the game-theoretic parlance of formal military planning, shifting configurations in the “state of nations” would present themselves singly (one-at-a-time, the expectedly best case for Washington); suddenly (unexpectedly, with apparent “diffusiveness”) or in variously-overlapping “cascades” of strategic complexity.
Quo Vadis?
What is to be done? To properly understand such analytically dense scenarios will require properly-honed, well-developed and formidable intellectual skills. This will not be a task for half-educated politicians or the intellectually faint-hearted. It will require, instead, sharply-refined combinations of historical acquaintance, traditional erudition and a demonstrated capacity for advanced dialectical thinking. Clarifying elucidations of such disciplined thinking should go back to the dialogues of Plato and the ancient but timeless awareness that reliable analysis calls for continuously sequential asking and answering interrelated questions.
This will be a challenging task. Indeed, nothing could be more daunting. It will require American strategic thinkers who are as comfortable with the classical prescriptions of Plato and Descartes as with the more narrowly technical elements of modern military theory. Consequently, it will have to be tackled by the “few,” not the “many.”
There is more. Not all nuclear wars would have the same origin. It is conceivable that Washington, Pyongyang and Beijing are currently paying insufficient attention to residual risks of an unintentional nuclear war. To this point, each pertinent president would seem to have assumed the others’ decisional rationality. After all, if there were no such reciprocal calculation, it would make no ascertainable sense for any side to negotiate further security accommodations with the others.
Viable nuclear deterrence (not “denuclearization”) should be the overriding US strategic goal vis-a-vis North Korea, but this complicated objective is always contingent on basic assumptions concerning enemy rationality. There is also an antecedent question: Are such assumptions realistic in the case of a potential war between two already-nuclear powers?
If at some point President Donald Trump should fear enemy irrationality in Pyongyang, issuing calibrated threats of US nuclear retaliation would no longer make any sense. At that unprecedented and opaque stage, American national security could come to depend on presumptively optimal combinations of missile defense and defensive first strikes. In any event, determining such dense combinations would lack decisional input from quantifiable historical data and from “wise counsel.”
In a still-imaginable worst-case scenario, the offensive military threat could require either a situational or comprehensive preemption – a defensive first strike by the United States – but at that late stage, all previous hopes for bilateral reconciliation would have become moot. There would then exist no “ordinary” circumstances wherein a preemptive strike against a North Korean nuclear adversary could be considered “rational.” What then? Whatever the answer, this question should always be approached as an intellectual problem, not a political one.
None of these referenced strategic decisions should be reached casually. With the expanding development of drones, EMP ordnance and “hypersonic” nuclear weapons, figuring out optimal US military policy combinations from one security crisis to another would quickly become overwhelming. Though counterintuitive amid prominently intersecting complications, the fact that one “player” (the US) was recognizably “more powerful” than the other (e.g., North Korea) could prove effectively irrelevant. In the “real world,” let us be candid, threatening a “bigger button” would represent little more than buffoonery.
Law and Strategy
There is more. In all foreseeable circumstances, there would be overlapping issues of law and strategy. Under international law, which remains an integral part of US law, the option of a selective or comprehensive defensive first-strike might sometime be correctly characterized as “anticipatory self-defense.” But this would be the case only if the American side could argue persuasively that the security “danger posed” by North Korea or other specific adversary was “imminent in point of time.”
Such discernible “imminence” is required by the authoritative standards of international law; that is, by the formal criteria established after an 1837 naval incident famously called “The Caroline.” Presently, offering aptly precise characterizations of “imminence” could prove confusingly abstract and operationally problematic.
For the moment, it seems reasonable that Kim Jong Un would value his own personal life and the survival of his nation above every other imaginable preference or combination of preferences. Despite his occasional bluster, Kim does appear to be visibly and technically rational, and should therefore remain subject to US nuclear deterrence. Still, going forward, it could become important for a negotiating American President to distinguish between authentic instances of enemy irrationality and instances of contrived or pretended irrationality. This vague prospect adds yet another layer of complexity to the subject at hand, one that could at some point include force-multiplying synergies.
Throughout history, wars have often been the result of leadership miscalculation. Though neither side here would likely seek a shooting war, either Kim or Trump could still commit various errors in the course of rendering their strategic calculations. At times, such consequential errors could represent the unintended result of jointly-competitive searches for “escalation dominance.“ These errors are more apt to occur in circumstances where one or both presidents had first decided to reignite threatening and hyperbolic rhetoric.
Even in reassuringly calm periods of polite diplomatic discourse, major miscalculations, accidents or “cyber-confusions” could accumulate. Not only might such ill-fated accumulations be hastened by the effects of widespread disease pandemic, the tangible likelihood of another pandemic (this next time, including plague) could be made greater by super intelligent AI interventions and recursive AI self-improvements.
In several worst case scenarios, negotiations gone wrong could result in a nuclear war. This largely unpredictable prospect ought never to be downplayed or overlooked. In the obvious but necessary words of Swiss playwright Friedrich Durrenmatt, “The worst does sometimes happen.”
Inadvertent Nuclear War with North Korea
An inadvertent nuclear war between Washington and Pyongyang (or a still non-nuclear Iran) could take place not only as the result of misunderstanding or miscalculation between rational national leaders, but as the unintended consequence (singly or synergistically) of mechanical, electrical, computer malfunctions and/or “hacking”-type interventions. These interventions could include clandestine intrusions of “cyber-mercenaries” or autonomous AI “aggressions.”
While an accidental nuclear war would necessarily be inadvertent, certain forms of inadvertent nuclear war would not necessarily be caused by mechanical, electrical or computer accident. These difficult to anticipate but consequential forms of unintentional nuclear conflict would represent the unexpected result of specific misjudgment or miscalculation, whether by singular decisional error of one or both sides to a two-party nuclear crisis escalation or by still-unforeseen synergies arising between variously singular miscalculations.
In an upcoming crisis between Washington and Pyongyang or between the United States and another enemy, each side would strive to maximize two critical goals simultaneously: (1) to dominate the dynamic and largely unpredictable process of nuclear crisis escalation; and (2) to achieve “escalation dominance” without sacrificing overriding national security interests. This second objective would mean preventing one’s own state and society from suffering catastrophic or existential harms.
In such increasingly plausible circumstances, having a “bigger button” (Trump’s earlier metaphor) could quickly become a source of weakness, not of strength. Nuclear “size” would matter, but only in unexpected or counter-intuitive ways. It follows, among other things, that to take comfort from observing that North Korea has been testing mostly shorter-range ballistic missiles would “miss the point.” Already during Trump’s first term, several of North Korea’s nuclear test firings had expressed yields at least 16x larger than the Hiroshima bomb. The 14KT WW II bomb (a “small” nuclear weapon by today’s ordnance standards) produced almost 100,000 immediate fatalities.
Vital understandings about nuclear “button size” would continue as long as Kim Jung Un’s “inferior” nuclear arms seemingly remained invulnerable to American preemptions and capable of penetrating missile defenses in the United States, Japan and South Korea. Because of the extraordinary harms generated by “low-yield” nuclear weapons, even a small percentage or tiny fraction of Kim’s “inferior” nuclear arsenal could produce “unacceptably destructive” outcomes in Washington, Tokyo or Seoul. Also worth noting is that in all such venues of strategic competition, the only reality that would figure in adversarial calculations would be perceived reality.
An Issue of Overwhelming Complexity
The bottom line of informed assessments concerning a US-involved nuclear war is that all underlying issues of contention and calculation would be starkly complicated and potentially indecipherable. Faced with endlessly challenging layers of complexity, both operational and legal, the United States should now proceed warily and in a fashion that is recognizably risk-averse. Though such prudent counsel could at first appear contrary to meeting US obligations of “escalation dominance,” any future Trump-involved nuclear crises would take place in “uncharted waters.”
There is more. Any aggressive over-confidence (or what the ancient Greeks called “hubris” in tragic drama) by President Trump ought strenuously to be avoided. While everything at some upcoming negotiation might initially appear simple and calculable, history should call to mind Prussian strategist Carl von Clausewitz’s classical observation about “friction.” Succinctly, this ubiquitous fly-in-the-ointment references the differences between “war on paper” and “war as it actually is.”
To avoid intolerable crisis outcomes between the United States and North Korea or other adversary, a prudent, science-based and informed nuclear posture should be fashioned, not with silly and self-destructive clichés but with refined intellect and cultivated erudition. The ancient Greeks and Macedonians had already understood that war planning should be treated as a disciplined matter of “mind over mind,” not one of “mind over matter.” Going forward, any purposeful assessment of America’s nuclear posture should include the impact of more-or-less integrated AI decision-making on “friction.”
Averting a Final “Nightmare”
It would be best for the United States to plan carefully for all strategic eventualities and not to stumble into a nuclear war – whether deliberate, unintentional or accidental. The fact that any such “stumble” could even take place without any adversarial ill will or base motive should provide little sensible consolation. In essence, an ounce of diplomatic prevention will be well worth avoiding a strategic nightmare.
Nightmare. According to the etymologists, the root is niht mare, or niht maere, the demon of the night. Dr. Johnson’s famous Dictionary claims this corresponds to Nordic mythology, which identifies all nightmare as some unholy product of demons. This would make it a play on the Greek ephialtes or the Latin incubus. In all of these fearful interpretations of nightmare, the idea of demonic origin is integral and indispensable.
But America’s current security worries are of a different and more secular sort. Now there are inherent complexities in problem solving that must be accepted, understood and suitably overcome. At a time when our planet is imperiled by variously intersecting threats of nuclear war, there is no reasonable alternative to proper analytic emphases.
The risks of a nuclear war between North Korea and the United States should always be assessed on solid intellectual foundations. By correctly acknowledging that North Korean denuclearization is a futile expectation and diplomatic non-starter – Kim Jong Un would never accept such a condition of codified inferiority – an American president should focus more explicitly on creating a viable system of mutual deterrence. Though such an “egalitarian” focus might first appear demeaning for a Trump-era “Great Power” like the United States, national security policy should always be founded on rigorously-accurate theoretical assumptions and a correlatively apparent national “will.”
Summing up, an “inferior” nuclear power such as North Korea could create unacceptable security harms for the United States. In part, at least, these harms might be the result of Clausewitz’s difference between “war on paper” and “war as it actually is.” In any final reckoning, the security harms of a nuclear war with North Korea could be quasi-existential for the United States and be impacted by coinciding nuclear crises involving Russia, China, India, Pakistan or Iran.
In worst case scenarios, synergistic interactions between seemingly-separate crises could generate harms that are sui generis and no longer subject to human oversight. In handling each threats, it would be essential that President Trump first understand the AI-enhanced importance of “friction” and accomplish this critical understanding in the perplexing context of unmanageable machine power. While the problem of “friction” has been around for thousands of years, it is exponentially more serious today, as superintelligence becomes increasingly autonomous and the outcomes of recursive self-improvement become unknowable.
All things considered, underlying questions of knowledge and power remain essentially unchanged from earlier centuries. In The Republic (4th century BCE), Plato asserts famously: “Until philosophers can rule as kings or those who are called kings can become philosophers, that is, until political power and philosophy coincide, there will be no rest from evils.” Accordingly, whatever tangible outcomes are being spawned by artificial intelligence, whether directly or indirectly, human survival will still require wise and virtuous human governance.
To meet this immutable requirement, Americans and others should learn to identify gainful leadership with a dignified philosophical spirit. Any state and society that ritualistically abjures such spirit must condemn itself to sudden or incremental disappearance. In this regard, warned 20th century philosopher Karl Jaspers: “The enemy is the unphilosophical spirit which knows nothing and wants to know nothing of truth.” This is the same enemy celebrated by Donald Trump’s brazenly incoherent declaration of affection on February 23, 2016: “I love the poorly educated.”
