But now that Watson conquered the world's best and brightest: What is next? The world itself?

Before competition started, Rutter joked, "When Watson's progeny comes back to kill me from the future, I have my escape route planned just in case." When Jennings gave his final Jeopardy response on the last day of play, in parentheses he wrote "I for one welcome our new computer overlords." (In true trivia nerd fashion, our geniuses were referencing pop culture--the Terminator Sage and The Simpsons, respectively.) Whether it be tongue in cheek or not, this worry seems to be ubiquitous. Everywhere, both in conversation and online, we have heard the same thing in response to Watson's success: "Well, here comes Skynet."

For those who don't know: Skynet is also a reference to the sci-fi Terminator saga. Skynet is a supercomputer that will be, in the future, in control of all U.S. military technology for the purpose of national defense. But when it becomes self-aware, humans try to pull the plug. Skynet decides that this is a threat, that humans are thus its enemies, and turns against them. Most of us die. (For further discussion see Brown and Decker's book Terminator and Philosophy.)

The hope of Watson's creators is that he, or computers like him, will help humanity. A computer, like Watson, that can scan two million pages in three seconds to form Jeopardy answers might, with a bit of tweaking, be able to offer medical diagnoses more accurate than those of television's Dr. House. An updated Watson could analyze scientific data (or even form scientific hypotheses) that could lead to scientific breakthroughs. His progeny might solve economic crises, deal with political problems (without bias or subservience to special interests), and yes even run our entire military with precision and cat-like reflexes.

Is this really something to worry about? Is it likely that such a computer would, say, become self-aware and then turn on us? We would, of course, install safeguards on such computers, but true believers will simply suggest that a super computer could always find a way around any safeguard. That's pretty doubtful, but the real problem with such reasoning is this: the fact that something happened in a movie is not a good reason to think that it will happen in real life. Every technology has its upside and its downside (e.g., nuclear power vs. nuclear weapons) but thinking that a new technology will destroy the entire world, because you saw it in a movie, is silly. We are notoriously bad at predicting the future, especially in movies. (Think about the 2015 future of Back to the Future.) The truth is, we have no idea what dangers and opportunities such technology will bring; did Gottlieb Daimler or Karl Benz, the first inventors of the gasoline engine, envision international plane travel or environmental pollution? This is not to say that we shouldn't be mindful of potential dangers, but we can't let the possibility of something going wrong stand in the way of real progress. Everything has potential dangers. We can (and should) press forward while being careful along the way.

For a testament to our inability to predict the future, just look at what so many have said, in the past, about the future of computing. Speaking a human language, beating a grand champion in chess, and processing and understanding human language, are all things that many had said would be forever outside the reach of computers. Yet, with the development of Watson, his predecessor Deep Blue, and your GPS, all three have now been achieved. (We realize, of course, that there is still some debate concerning whether computers, like Watson, truly understand language). This doesn't mean that computers will definitely one day do all the things the naysayers say they can't. But it does mean that anyone who unequivocally rules anything out is greatly overestimating their ability to know. Humans are just really bad at predicting the future. We don't know what is possible and what is not; we don't know what will happen and what will not.

That said, let us (cautiously) make a little prediction of our own. If the field of artificial intelligence continues to develop as the hopeful predict, Skynet will not be the biggest danger. Computers that understand language will only be the beginning. Computers that have emotions, feel pain, solve problems, and are intuitive will be next. And they won't be housed in giant buildings. We may one day create androids, artificial beings that look and behave like us in every way. (Think of Data from Star Trek: The Next Generation, where he had emotions. Or, think about the Cylons from Battlestar Galactica.)

Androids would not be the danger. Sure, they might turn on us. But more likely, we would find them very useful. They could take out our trash and fix our cars. They could build our skyscrapers and clean our bathrooms. They could fight our wars. They could explore the far reaches of outer space. They could do all the dirty work that we don't want to, or can't, do.

Sounds wonderful, right?

Ah, but here's the rub. We don't know for certain if androids would have minds--if they would truly feel pain, make decisions, and be self-aware. In fact, we can't know for certain. Just like neither of us can crawl inside your head and verify that you actually have experiences like we do (philosophers call this the problem of other minds), none of us could never crawl inside an android's head and verify that it has a mind. We also couldn't verify that it didn't. But since androids would behave just like us, it would sure look like they had minds. They would wince in pain, get upset, express remorse--all the things that you do to make us think that you have a mind. Even if you asked an android if he had a mind, he would say yes. And since we don't know (and may never know) what it is about our brains that produces consciousness (is it biology, or simply the way it functions?) we wouldn't know for sure whether or not an android's brain has what it takes.

So would it be acceptable to make Watson's descendents do all our dirty work? It's impossible to tell for certain. If they are not conscious, sure. But if they are conscious, then they are people. Artificial people, but people all the same. People that feel pain, have wants, and desires, and all the rights and responsibilities that we do. To force them to do all our dirty would be slavery, in the most literal sense. Some people will insist that the androids aren't conscious--that they aren't people, don't have minds, and can't feel any pain that we inflict on them. But Nazi's said the same thing about the Jews. Some plantation owners said the same thing about the Africans. The truth is, since we don't know what is required for consciousness, we would have no good reason for thinking that androids would not be conscious. But since they would act just like we do, we would have a really good reason for thinking they are. If we treat androids with respect, and they are not people, what have we lost? But if we don't, and they are, we will have fallen right back into the worst of our own past follies.

The idea that androids are real-minded people will meet fierce resistance. There will be no good arguments--just like there were no good arguments for the non-mindedness of Jews and Africans. People will point to differences, without any reason for thinking those differences are relevent.They will attempt desperate fallacious justifications  to protect our own biases and self-worth. And people will latch onto such arguments with all the gusto they can muster--even though they fail.

So, it seems, the coming Artificial Intelligence that Watson heralds has more reason to fear us, than we have to fear it. Skynet is not the biggest danger, human ignorance is.

(For more on these issues, we recommend Jason Eberl's Battlestar Galactica and Philosophy.)

Copyright: David Kyle Johnson