Predictions Old And New

     I think I’ve mentioned here, or perhaps at Eternity Road of loving memory, that I was once deeply immersed in international strategic analysis. That immersion arose from my studies in physics. Once I became aware of the immense power available from nuclear technology, and the ways in which it could be put to destructive purposes, I had to get a sense for the possibilities.

     That began, roughly, in 1966, when I happened upon a peculiar book in my high school’s library. It wasn’t the sort of book one would expect to see in a high school library, but there it was. Having found it, I simply had to read it.

     The book was Herman Kahn’s magnum opus On Thermonuclear War.

     Pretty strange fodder for a fourteen-year-old kid, eh? Even my father thought I’d spend my time better by improving my screwball and chasing girls. But I felt a need to understand Kahn’s analysis of that most horrifying of subjects. After all, we were doing regular “duck and cover” drills back then. People still built backyard bomb shelters. “Civil defense” was a lively subject even at town hall. So I read Kahn’s book carefully and thought hard about its implications.

     I was aware that many things people argue about have no relation to any real danger or opportunity. But I knew about the tensions in Europe, the NATO / Warsaw Pact face-off, the Cuban Missile Crisis, the Nuclear Test Ban treaty, and (of course) the “space race.” I knew about America’s steadily increasing expenditures on strategic weaponry. I kept abreast of the angry rhetoric that emanated regularly from the USSR. The matter was serious enough to deserve attention and study.

     Strange things were being said in classrooms, too. The mid-Sixties were the leading edge of antinuclear activism. Any number of people argued that the mere existence of nuclear weapons endangered the very existence of Mankind. Some of those people were teachers determined to impress their views on us young folks.

     It’s been six decades since then. A great deal has changed, but not this: nations that fancy themselves as “great powers,” actual or potential, still strive to acquire weapons of mass destruction (WMD). Whether nuclear, biological, chemical, or whatever else there may be, those who seek power over others want all of them.

     It’s unclear just how widely nuclear weaponry has spread. We know that nine nations definitely have them:

     But other nations may have acquired them as well. As for the other varieties of WMD, the picture is considerably cloudier. The tableau makes the need to understand international relations, competitions, and tensions graver than ever.

* * *

     Haretina Kukuri may be the most valuable thinker alive today. She’s produced a stunning compilation of analyses and inferences about international stability. If you have any interest at all in international relations and how the various rivalries affect world events, please read her piece in its entirety.

     From its beginning:

     Nobody built 2030 on purpose. I need you to sit with that sentence before you read another word of this piece, because everything that follows is simply proof of it, repeated across five domains until the pattern becomes impossible to unsee. Every actor examined here did the single most defensible thing available to it, at every decision point, across four straight years, and the sum of all that defensible behaviour produced a condition none of them would have chosen and none of them can now exit alone.

     From its middle:

     Twentieth-century stability rested on forces you could observe, count, trust to change slowly, tank divisions, missile silos, the sort of thing a satellite could photograph and a treaty could cap. The decisive capabilities of this decade meet none of those conditions.

     And from near its end:

     Haretina is an honest prognosticator. Immediately before that chart of probabilities, she writes:

     Every generation's predictions, [Freedman’s] own research shows, were shaped by the anxieties of their own present as thoroughly as by any genuine foresight…. The weights below are my own judgement.

     Once you’ve acquainted yourself with the breadth of her studies and ponderings, you’ll be inclined to trust that judgment.

* * *

     This is one of those encounters that can alter a thinker’s entire mental model of geostrategic causality. While there are themes that reach backward to fears we’ve faced in other guises – consider the movie WarGames for an example – there are new factors that transform those themes in critical ways. Probably the most important of those factors is the emergence of what we’ve been calling artificial intelligence (AI).

     The greatest of all threats to world peace is the powerful decision maker who sees his nation – and by implication, himself – brought to the edge of the abyss. In times past, his judgment, even if bolstered by the counsel of sober, intelligent others, was a human thing. It functioned at human speeds from a catalogue of a human’s priorities. It may include economic calculations, but only as supports to decisions with broader consequences.

     Digital AI, founded on calculation as is all digital technology, will “see things differently.” And it will see them at a speed far above the human. When that powerful decision maker’s analyses of his nation’s situation derives partly from AI, the factor of greatest import will be the degree of trusts he – and whoever might be in a position to constrain or overrule him – invests in that AI.

     We once thought the arms-control agreements of the Eighties and later decades would make us safer. But those agreements only limited certain kinds of weapons. The destructive capabilities of the world’s States have grown in other directions. AI hasn’t yet been adequately assessed in that light.

     When Haretina wrote:

     A single leader's miscalculation, a single accident nobody modelled, could make every careful trend in this piece irrelevant in an afternoon, and no survey, no book, no probability table protects against that possibility, because that possibility is precisely what none of them can see coming. I have built this piece on the better-supported case. I have not proven it is the only one.

     … I’d bet she was thinking about AI.