So, it occurred to me that this series was never going to get started until it got started, so here we go. You’re never going to be 100% ready, but don’t let that be an excuse to take 0% action, right?

Right.

Um, right out the gate, let’s do a bit of a “Talebian disclaimer” with regard to what Nassim (Taleb) would call the “Ludic Fallacy” — using games, which are simple and calculable, to model reality, which is too complex to calculate. Chess is a turn-based game with unambiguous rules played on a board of limited scope. To be fair, it’s deep as all get-out, and the possibilities are too complex and numerous for even the best supercomputer to crunch out; this isn’t tic-tac-toe (noughts and crosses) — artificial intelligences, just like organic intelligences, have to give up on optimizing (finding the absolute best/perfect solution) and be content with satisficing (using the best solution that can be found within the constraints of the available time and resources, and not bothering with perfection). That is, they have to make a best guess based on limited, incomplete information. They have to, as it were, pick the best bad idea; there is simply no time to calculate every possibility. But chess, deep as it is, is not even a puddle compared to real life.

Real life is not equivalent to chess; casinos are not equivalent to stock markets; casinos (Mediocristan) aren’t even like stock markets (Extremistan). Real life is full of power-law distributions. Only our artificially simple games are Gaussian. But does that mean that we can’t learn any lessons from chess? Hail naw. I mean, of course not. Sorry, too much rap music. There’s probably enough depth in chess to keep us busy finding real-life applications for years, perhaps even a lifetime. It’s just that it’s never, ever going to be a 1-to-1 correspondence.

That was a lukewarm-a$$ disclaimer, wasn’t it? Barely seems worth having made. Oh well, it’s written now. Let’s move on!