I've said it before, I'll say it again. When it comes to predicting how humans are going to behave, or respond or think about a certain proposition or artifact, I have very little faith in science. Maybe even less faith than I have in sorcery. Tea leaves. Bird entrails. Crystal balls. It seems to be knowing what's going to happen next, or if someone is innocent or guilty, or how they'll behave given a set of circumstances has been a human wish since the beginning of human time. I've heard stories of inveterate gamblers choosing one rain drop over another and saying that one will descend the windowpane first. Or people betting on turtles in county fairs, or trying to figure out where fish will be so we can catch them. We've been trying to predict the unpredictable since the early days of bi-pedalism. In fact, it's been conjectured (speaking of bi-pedalism) that our ancestor stood on two feet so they could get a better view of where food was, or enemies, on the veldt. Standing upright was a new way of seeing things, and edge. People and companies have always looked for that edge. Very few people and/or programs and/or institutions have been right much more often than they've been wrong. Hans Rosling, in one of his great Ted Talks, claims (to make a point) that the professors from the institute that awards the Nobel Prize, can be right less often than a random group of chimps. (Watch the first four minutes of the YouLube link above.) Late last week, I read a book review in The Wall Street Journal of a book called "Dots and Lines: Hidden Networks in Social Media, AI, and Nature" by Anthony Bonato. Smack dab in the middle of the review, I ran across these sentences: It brought to my mind once again my thesis (once I get a thesis stuck in my head, the chances of me straying from it are as low as a cockroach doing the limbo in Death Valley. To say I'm a dog with a bone is unfair to dogs, maws and bones.) When I first came across the "science" of Customer Relationship Marketing, messages (usually in the form of variable field letters) were customized according to six or ten bits of data. The customer--aka victim--might have filled out a survey that said she had children. Every letter to her from the car company that owned that data therefore was about seating capacity and safety. The "data" was a blunt instrument based on very little, but because we trade in certainty, we pretended it would portend great accuracy. We knew what that customer with children wanted--we had data on them, after all. And data is the answer. What we continually fail to see is the very complicatedness of the human brain. We fail to appreciate the near impossibility of understanding how it works and how much more baroque it is than even them most ornate of human-made systems. The most sophisticated AI systems cannot reckon with the idea of 100 billion neurons and 100 trillion synapses, much less 300,000 miles of neural connections. We look to find order or predictability despite these numbers, ignoring the very idea that decision-making, like any causality, can and will mutate based on a wisp of a chimera. A short review from the Economist. I send these things to clients to give them more ways to ignore me. We tried to explain through metaphor how the brain worked. First, when automata were the rage, say in 16th century France, we explained the brain that way. Then as clocks and watches grew in sophistication, they were enlisted to explain the mechanistic calculus of your grey matter. Next, scientists compared the brain to telegraph system, or a computer. Then, they compared it to the world-wide web and giant networks. The right metaphor might be the universe itself. (Walt Whitman knew this: "I contain multitudes.") The same universe that is unfathomably complex and expanding and contracting at the same time. Take one 101 course on astronomy and you realize the vastness of things. I understand the salesmanship behind the perennial promise that "I know what will happen next." It makes a compelling promise. And people buy it. It's like a magical drug. The Greeks had a word they used when humans believed they were more powerful than the gods: Hubris. It's not hard to get from "tech bro" to hubris. Things that start in hubris almost always end in humiliation. It seems though we might be an industry ruled by hubris. Hubris is an all-you-can-eat-buffet. Where you're eating yourself. |