On The Most Human Human
by Thomas Brett
In his book The Most Human Human, an engaging account of competing in the annual Turing test, Brian Christian ranges far and wide through the literature of AI (artificial intelligence), linguistics, computer science, philosophy and even poetry to figure out what exactly makes us distinctly human and distinctly different from machines. The Turing test was conceived by Alan Turing, an English mathematician, in 1950. The test is whether or not a computer can fool a human into thinking that it–the computer–is also human through interrogation only. If a computer can fool us, then it could be said to “think.” Today, the Turing Test pits both computer software programs and human “confederates” against one another, each trying to convince a human judge that they are human and not machine. The catch is that each interaction has a 5 minute time limit. The winner in the human confederate category, of course, is deemed “The Most Human Human.”
One of the more interesting of the book’s digressions is Christians’s discussion of chess playing, specifically the different ways humans and software programs approach this decision-making terrain. Chess is a space for thinking about what makes humans human precisely because the game offers such a vast array of possible moves to get one’s brain around. And the amazing thing–at least from the perspective of a non-chess player such as myself–is that the very best chess players can navigate the terrain of possible moves by intuitive means. What this means is that not only can a Gary Kasparov draw on vast experience but he can also make unusual choices as to how he proceeds. Case in point: making a somewhat random opening move is a great way to stymie a computer software opponent like Deep Blue, who, of course, proceeds through the game only by crunching millions of possible moves per second. The human player has the poetics of randomness and intuition on his side, against which the machine can only number crunch the relative merits of the next move. Here Christian hits on an important point about human creativity: on some level it requires the practitioner/artist to not know exactly what he or she is doing. Or in the words of Donald Barthelme, one of the interviewees in Christian’s book: “Not knowing…is what permits art to be made.” Barthelme is referring to that aspect of the creative process that is inherently random, accessible only by our vaguest of intuitions.
This discussion of randomizing one’s opening moves had me thinking about how I begin working on a new piece of music on the computer. For some time now I have fretted over the sheer number of possibilities open to me as I try to decide how to proceed with a piece. The software programs on my laptop are like a chessboard in that they invite millions of possible creative moves from me. It’s a deeply exciting prospect but also potentially paralyzing.
In the past, I would try to systematically think my way through the “best” option: maybe I’d start by searching for a nice pad sound and then…The problem is that my systematic thinking would always be interrupted by a rogue sound, an unexpected by-way, or an accidental juxtaposition that would instantly charm me, as if asking: “But have you considered this?” Well, no, I hadn’t considered that because I was under the impression that I was doing something else. And then I would fret some more about having let myself be undermined by my own digressions . . .
After reading The Most Human Human, I decided to try applying the idea of randomizing an opening move to writing music. What I did was just jump into making sounds–any sound that seemed interesting–so that I could get the music “game” underway and remove from the equation my anxiety about having too many options. Randomizing my opening move–”Let me just build a little pattern using this drum sound…”–let me get on with the more satisfying business of interacting with and building a new sonic organism that could grow.