This month in Humans Overreacting to AI Breakthroughs, the contingent of folks who obsess over Facebook privacy yet still manage to get their email hacked twice a year are unsettled by a program that has just defeated the most brilliant human mind in the strategy game of Go!, a significantly greater challenge than chess in terms of possible moves and outcomes to predict. As usual, this category of human fears the implications of such a complex learning system, following the idea of “if it can out-think, or beat a human at a game, it can beat a human in other, more dangerous ways.”
It’s a fair leap in logic, though not particularly helpful. The world is made of numbers, and playing the odds has been a method of survival since the dawn of time. To succeed at game theory is a powerful skill. AI programs have a natural inclination towards numbers and predictive technology, since numbers are governed by rules, which can be taught. But that’s exactly where AI and Man can succeed together: imagine any military scenario where a great number of lives are at stake. Hell, imagine any important human job involving a very large specific, rule-bound but constantly changing set of data. You don’t necessarily need to prove that a robot can beat a human at this sort of strategy game because we’re not looking to replace the human element (before you say no robot should have any say over human lives, MOM.) Rather, a robot that is capable of winning is capable of keeping up, of supporting a human when the human needs to be more accurate than a human can typically be.
You’ve seen this sequence in a hundred movies: the student wants to be the best at his craft but lacks artistry. They challenge the master. The master only accepts them as a partner or a student once they have surpassed, at least once even by luck, the skill of the teacher. It’s a classic Batman and Robin scenario (and, of course, I refer to the deliciously terrible Joel Schumacher Batman & Robin, and invite you to fight me at your leisure, nerds.) Batman doesn’t feel threatened by Robin because a) he needs support and b) Robin doesn’t understand real-life consequences because he’s a teenager with a deathwish. Should you feel threatened by a robot support system? No. Robin (robots) will support Batman (humanity) in ways that Batman lacks, but has no real desire to actually usurp Batman (yet,) because Batman is furnishing him with all his cool toys and a home and family and love. Robin can’t succeed without Batman, and Batman needs a strong support system because he’s a psychological trainwreck prone to emotional mistakes. I’m not saying an ambitious robot couldn’t revolt, but as yet they don’t have ambitions at all unless we decide they do. We’re looking to build a team relationship.
To get back to the larger point, all this development tells us, the Dumb Public, is that intricate thought processes are moving forward in a spectacular way, and rather than be afraid of a robot surpassing a human’s strategic capacity, we should be overjoyed that we’re much closer to a processor that can keep up with the most brilliant and intricate human reasoning and strategy so as to help, challenge, and push humans to greater achievement. You have no reason to be a sore loser; you don’t even know how to play Go!.