Future Definitions Of Human
This is a transcript of Eliezer's talk on a panel discussion at World Futures.
Humans are a very tribal species - we divide the world into "us" and "them" in steadily widening circles. Myself. My family. My tribe. Those other tribes we sometimes trade with, but not "our tribe". Strangers. And the enemy. There was a time when being sufficiently "un-American" was a crime; you'd get hauled up and interrogated by the House Committee on Un-American Activities. Which is odd, because if you look at it logically, "un-American" doesn't directly say that you've done something awful. It says you've done something which is not characteristic of one particular country's current culture. "Un-American" doesn't say that you've done better or worse than the American average. It just says you've done something different.
Let's talk about the word "human". We look, for example, at Adolf Hitler, and we say he's "inhuman". We want to pretend that Hitler was an alien pod person sent in to sabotage our species from the outside. We don't like Hitler so we want to make him into a nonperson, where nonperson is defined as anyone from outside the tribe. We call Hitler "in-human", not-of-our-tribe, in the same way that Hitler might have called me "non-Aryan". Contrary to Hitler's ideology and our own tribal instincts, humans really are alike under the skin; we have ways of thinking that are humanly universal, and one of those human universals is tribalism. Ironic, isn't it.
I think that in the future, "human" will probably be used to refer to the kind of people we are right now. In that sense I doubt the future definition of human will be very much different from the definition we use today. A "human" is a person having our unique cognitive and emotional architecture, including elements like tribalism, which are not spectacularly well-designed or even very nice.
Now, there are nice things to say about humans. And perhaps it will be possible to say these same nice things about a great many people of the future, humans or not. It is possible that we would look at these nonhuman people and see something in them that causes us to recognize them as... fellows. Neighbors. Even... people of a larger tribe, to which we also belong. Because they have something in common with us, something we care about. So how to describe that? One strategy is to distinguish between "human" and "humane" and say that many of these future people will not be human, in the sense of being members of the particular species Homo sapiens, but they will be humane - they will be, broadly speaking, nice people. You might say that "human" is what we are, and, being human, "humane" is what we aspire to be.
So... let's say that there's a word, "humane", that applies beyond the humans of this present day. What does the word "humane" mean? Is it the same as the word "person"? Is it identical with the phrase, "legally and morally deserving of citizenship rights"? How about the word "intelligence"? These are also all important generalizations of the word "human". But "humane" seems to be something over and above those other qualities. I would say, for example, that if you are a person you therefore have citizenship rights, and that to be a person you must be intelligent, but I can imagine an intelligent citizen person that is not humane. And by that I don't mean Hitler. Hitler had an actively human kind of evil that overwhelmed those humanly good qualities that he must, in fact, have had, because they too are human universals. If you took Hitler, and subtracted all the evil qualities, you would not be left with a Blank Slate, but with humane qualities that previously did not express themselves. Hitler minus evil equals Gandhi. Conversely, Gandhi minus good equals Hitler. That's part of being human. Everyone gets the entire human set of emotions, even if they aren't all visible at once.
So does being "humane" require the entire human set of emotions? I don't think so. The human emotions as they stand lack an important kind of self-consistency. The human emotional architecture has a fundamental imbalance in it. We are out of kilter. Usually this particular kind of inconsistency would correct itself very rapidly, but in humans it can't, because unlike an Artificial Intelligence, we don't have access to our own source code.
The inconsistency I'm talking about is that we don't want to want what we want. We like eating chocolate, and we don't like eating lettuce. But we wish that we enjoyed lettuce as much as we enjoyed chocolate, because then we'd be thinner. We stay on the treadmill, and we hate that, but we wish we didn't hate it, because then we wouldn't have to expend mental energy. If we had access to our own source code, we could fix the problem. But unfortunately humans were not designed by humans. Humans were designed by evolution. So we eat the cheeseburger, and give up on the exercise plan, and don't read all the books we want to read, and we're never quite as nice as we think we ought to be. It's an unstable condition that can only persist as long as we don't have a choice in the matter. We are not who we would choose to be, if we were free.
If I were to define "humane", it would be as a specific kind of continuity with humanity. Not all the human qualities, but the renormalized human qualities - the ones we'd choose to keep if we had the choice. If "human" is what we are now, then "humane" is what we would be, if we were free. I think that in many ways the human condition is tragic, but it is not beyond redemption. If it were beyond redemption we would be incapable of judging it as tragic. All we lack is the freedom. If we can obtain the ability to choose who we are, then our wish to be better people will, in the end, finally win.
