Playing Games with our Future and Our Lives

Donald Trump's early Christmas present to the world was a right-up-to-the Twitterbrink 140 character policy tweet: "The United States must greatly strengthen and expand its nuclear capability until such time as the world comes to its senses regarding nukes". In case you didn't get the point, he "clarified" it by saying "Let it be an arms race". In doing so, our incoming Tweeter-in-Chief goes against decades of bipartisan policy of gradual reduction of nuclear stockpiles, thoughtfully enacted to as to simultaneously maintain the safety of not just our own country, but the entire world.

What kinds of thoughts are racing through his head when he decided to drop this tweet-bomb on the world? What is the strategic thinking that guides the idea that an increased nuclear stockpile, already large enough to destroy a good chunk of the world, is going to be the path to safety and security in a world rife with stateless terrorist actors? Who exactly would he target to stop ISIS or Al Qaeda or how would a greater stockpile deter them? How is it that he wants to go head to head with Russia on accumulating nuclear weapons for an imagined world-ending war no one in his or her right mind would want to fight, but dismisses the notion of the existing cyberwar that is actually already being fought here and which we may be losing?

The first arms race was in part an outcome of the applications of the mathematics of game theory to the phenomenon of conflict. Game theory is a formalization of strategic thinking, introduced as a mathematical discipline by mathematicians Oscar Morgenstern and John von Neumann in the context of economics. The most famous example in game theory is the Prisoner's Dilemma in which two prisoners are separated for interrogation and given the choice to confess or stay silent. Simultaneous silence gives them each a lighter sentence than simultaneous confession, but if either stays silent when the other confesses, the confessor gets off lightly and the other goes away for the long time. If they don't trust each other, they will almost surely both confess, not optimal for either individually. Analogized to the nuclear situation, each nation will opt to develop - and then overdevelop - the ability to absolutely be sure that under any circumstance, the one can annihilate the other: "Mutual Assured Destruction" (MAD). This strategic approach to nuclear armament eventually faded away - with the help of thoughtful and patient statesmanship - to a necessarily more nuanced view of arms control, intermingling dynamics of conflict with cooperation, necessary for a world with more than two nuclear states.

But any notion of nuance - much less the history or application of game theory - in the nuclear context seems lost on Donald Trump - and that is terrifying. Trump sees a black and white world (or rather a white world) in which there are winners and losers, more a game like the reality TV Show "The Amazing Race" than a mathematical game. The winner-loser paradigm doesn't generally make for great policymaking. Winning a battle against the unions or small contractors can trigger social unrest. Winning an election at any expense (e.g., inciting and inflaming racial and ethnic tensions, exploding general norms of social decency) can subvert the general tenets of democracy and the democratic process. Winning an arms race can bankrupt a nation and place our people and the world at the precipice of extinction.

In short, win on one dimension, but lose on another. In particular, there are few if any battles that Trump has ever won, that were also wins with respect to a larger moral or ethical value function. His value functions are obvious ones - number of dollars, number of warheads, number of likes and laser-focused on them, they create a blindness - willful or otherwise - to other less obvious, but in the long run more important, value propositions.

The winner-loser paradigm has no time or space for the subtleties and realities of our global and interconnected world. This is well known to economists, ecologists, climate scientists, sociologists, and basically every community that spends all of their professional life thinking about interacting, evolved systems of particles or creatures, and is is in part the result of applying modern game theoretic studies. In the best cases, there can be multiple "winners" - successes for many different players able to coexist in "equilibrium". At the risk of inserting a mathematical summary statement - these are systems operating in many, many dimensions, and any time you try to encode them in a single dimension, you lose a lot - maybe everything.