Also, I am not an expert in the subject of Artificial Intelligence, but I won’t let my ignorance stop me. There is one element of the human intellect that will probably get missed by the people working on AI. This facet of human decision-making is, despite all appearances, vital to the function of our big big brains. I call this element ‘the self-destruct compulsion’. It’s not a catchy name, but I just made it up and I’m tired from the gym, so you do better if you’re so smart.
This is more than taking a risk on a course of action that has a very low chance of success. No, this is the urge that every human being has to occasionally make the worst choice possible on purpose, just to see what happens. To paraphrase Alfred from The Dark Knight, sometimes we just want to watch the world burn.
If you’re a parent, you have seen your child engage this part of their logic. They look you straight in the eyes, with a fixed glare of determination, and they haul off and punch you (or pinch or kick or whatever the most verboten action is at the time). They know it’s a bad choice, they know it’s against the rules, but they do it anyway. It’s an exploration of boundaries, but it’s also a way to test how you recover from a terrible choice.
On a bigger scale, I suspect this urge sometime rises to the surface during election time. the tea party in America is a prime example of people voting for disaster. They want to see how bad it can really get, when you reject fact and logic and just barrel ahead with your hands off of the wheel. I’m worried that the election here is going in the same direction. The guy who is currently winning (according to the polls) is really running on an anti-government platform, because he seems to hate the political process more than anyone. Strangely, he’s capitalizing on the anti-politician sentiment by re-enforcing the stereotype of ‘all politicians are crooked’. It’s a shame, because there are so many good and honest people running for office, and I hate to see them painted with the same hateful brush. (Can a brush be hateful?)
So this compulsion to occasionally be your own worst enemy is a big part of our learning process. Learning to recover from catastrophe is a vital skill for evolutionary survival, so we make sure to keep practicing by making our own messes to clean up. Hopefully we can, as a species, stay one step ahead of our ability to make a mess, so we don’t create a disaster that is unrecoverable. And we won’t see a fully realized artificial intelligence until they add a subroutine that actively works against the good of the whole. I am a nerd.