Herman Kahn's Doomsday Machine

With the work of von Neumann, Turning, and Wiener, machines that were intended to merely model reality were anthropomorphized into "thinking objects" that were often considered more reliable than human actors. Such capabilities of computation coupled with the ability to "accurately" simulate "real" situations (or at least the strategists' perception that their models were correct) led quickly to the adoption of computers for complex decision making. Researchers at RAND asked, ‘If von Neumann's methodology of formalized games can be applied to physics, why not policy judgments?'

In 1952 Herman Kahn became involved with von Neumann in the design of the hydrogen bomb. To this end, Kahn simplified the Monte Carlo simulation while increasing its accuracy. Modeling a hypothetical hydrogen bomb became possible as a result. Later in his career, Kahn worked for the government's military consultation group, the RAND Corporation.

While working at RAND, Kahn settled in with a group working on nuclear strategy known as the Strategic Objectives Committee. Its members recognized that an all out nuclear war with an initial strategy to attack cities was not feasible. In response to such a strategy, Kahn (only half jokingly) proposed his "Doomsday Machine," a massive computer connected to a stockpile of hydrogen bombs. When the computer sensed imminent and intolerable danger from a Soviet attack, it would detonate the bombs and cover the planet with radiation fallout and billions of dead. No one laughed (except for Stanley Kubrick, whose 1964 dark comedy, Dr. Strangelove or: How I learned to stop worrying and love the bomb, parodied Kahn's Doomsday Device). The Doomsday Machine, nonetheless, was only a mildly absurd version of existing US policy: If the Soviets scare us, we destroy their cities and provoke them to retaliate. Kahn advanced the strategists' thinking to a new level by suggesting military installations as the next logical target. This work led Kahn to believe there could be such a thing as a winnable nuclear conflagration.

Kahn began working intensely with the massive computers at RAND's disposal. Modeling nuclear wars for the Strategic Operations Committee, Kahn proposed a variety of simulations that he claimed proved his theories. At the same time, his work had such persuasive (albeit paranoid) force that it became the basis for the majority of military strategy during the Cold War. Kahn believed that any war plan ought contain a variety of responses. The war had to be "controlled" so that "intrawar deterrence" might be practiced to prevent escalation of the conflict. Conceiving of 44 "rungs of escalation" from "Ostensible Crisis" to "Barely Nuclear War," from "‘Justifiable' Counterforce Attack" to "Local Nuclear War," Kahn saw himself as the great systematizer of nuclear strategies (Kahn, 1961). To "control" a conflict, the military needed what Kahn called a "Credible-First-Strike Capability" so that they could suppress Soviet strategic forces in the event that conventional forces failed. Kahn labeled the pure deterrent capability as Type I, a first-strike capability as Type II, and the retaliatory deterrent as Type III.

These many types of deterrence, variations on possible escalation scenarios, along with many other variables, were calculated. Using this data in a modified Prisoner's Dilemma simulation (based on von Neumann's game theory), Kahn modeled nuclear wars to determine US nuclear vulnerability: "If the Soviet aggressor is reasonable, he will avoid the defender's cities, civilians, and recuperative capability in order to maximize his post-attack blackmail threats" (Kaplan, 1983, p. 224). For example, given a Type I deterrence, a rational competitor would most likely attack military bases while keeping a reserve as a threat to destroy cities afterwards.

Kahn was not advocating a preventive war but was calling for first-use in the face of conflicts that could not be deterred otherwise. As a result, Kahn called for a buildup of missiles, bombers, and "limited-war" forces, as well as a massive civil defense program. Kahn's advocacy of civil defense was not so people could survive a first–strike attack. Rather, it was a strategically sound approach to winning a nuclear war: a country that could evacuate its cities could "take firmer positions" in conflicts hot or cold. With a good civil-defense system, Kahn felt going to the threshold of a nuclear war might even be advantageous. In response to criticism that this might escalate a limited conventional war to a strategic nuclear conflict, Kahn replied, "Insofar as the civil defense program gives us the ability to convert at our discretion, it should be a good thing" (p. 225). Kahn's book On Thermonuclear War (1961) collected his thoughts over the years and the content of his public lectures. He described the horrible side effects of radiation on the human body. Even after graphically describing the mutations possible, Kahn nonetheless concluded, "War is a terrible thing, but so is peace. The difference seems to be a quantitative one of degree and standards" (p. 228).

Kahn followed with his text, Thinking About the Unthinkable (1962). Partly responding to critics who characterized his work as murder, he claimed that one must consider nuclear war as possible and therefore a topic of discussion. He argued he was merely trying to guarantee a higher rate of survival than without any strategy at all. Even so, many of his RAND colleagues believed that with his models as the basis for decision making, a nuclear war became more likely. Moreover, it seemed that Kahn's simulations of war and survival might be inaccurate representations of potential situations. His approach to the civil defense problem assumed several optimistic conclusions while failing to account for certain complications and unquantifiable variables. Kahn even admitted his assumptions were "optimistic." In the end, however, his influence was felt and formed the basis for how everyone conceived of nuclear war. He created a vocabulary under which strategic issues could be discussed comfortably and easily. His labels reduced the emotional content of nuclear war to scientific indifference.

Perhaps the most damaging effect of game theory in the hands of RAND was the paranoid bias it introduced into the modeling of an enemy's psyche. Creating a computer model of the Soviets' military mind was the essence of RAND's work in strategic decision making. As in a Prisoner's Dilemma, the most rational action was to minimize risks and maximize gains (von Neumann's Mini-Max theorem); for two prisoners, the best strategy was to snitch on your partner before he snitched on you. Rephrasing the problem we can see the Prisoner's Dilemma as the perfect model for the Cold War; if you cannot maximize your gains through disarmament, minimize possible losses with a buildup of weapons. By framing the problem in a zero-sum calculus, the mathematicians artificially introduced conflict over cooperation.

But if the "game" were not as symmetrical as the zero-sum claimed, would it not be better to think in terms of collective good instead of individual gains? The fact that their machines were incapable of dealing with certain problems like the enemy's morale could mean an incorrect policy assessment based on erroneous simulations from RAND. RAND's failure to predict the extreme resistance of the Vietnamese, for example, may be connected to the bombings of North Vietnam and US failures in the region.