ADVERTISEMENT
  About the SA Blog Network













Assignment: Impossible

Assignment: Impossible


Exploring the area between the unknown and the impossible.
Assignment: Impossible Home

Visions: A Different Point of View

The views expressed are those of the author and are not necessarily those of Scientific American.


Email   PrintPrint



In the series “Visions,” science fiction about the very latest research will be paired with analysis looking into the facts behind the fiction. The goal is to marry ripped-from-the-headlines science fiction with analysis into the possibilities hinted at by new discoveries.

Everyone hated the Turing Tests in school, but everyone begrudged them. They could really make or break your grades.

A lot of kids goofed off on them at least once. I know I did. Once you got down to it, they could really make us squirm in our skins. Looking back now, we each of course were only beginning to really learn who we each were. It was a lot to imagine, perhaps too much to imagine, what it was like to be someone else. Something else.

The first one, where we each anonymously pretended in text to be the opposite sex, was the one where we acted the most childish. Boys played at being girls, and girls played at being boys, and we nearly all mocked the other, acting all mincing or macho. A bunch of people stopped taking it seriously right after that. But for most others, it loosened us up, which I think was the point. Made us more willing to play pretend.

We could venture into strange territory. And anger. Conservative or liberal? Atheist or religious? Gay or straight? Natural or engineered? Immigrant or native? We got into raging fights over caricatures drawn for each answer. We got into raging fights over actual convictions.

What were more than a little spooky were the answers that seemed utterly genuine but were apparently total fabrications. We all said we didn’t like classmates who were such good liars, but I think we were really surprised at how well others could understand us, and a bit ashamed that we couldn’t see others the same way. That’s what led many of us to start actually trying, to grasp how others felt.

Of course, we eventually got to the entire human or machine test, the entire reason we all had to take these Turing tests in the first place. It’s a touchy subject, more touchy than all the others, and the idea of taking it just made me want to withdraw deep into myself. I mean, why does anyone care who I am? Why does it matter what I am? So long as I’m intelligent. We’ve proven that again and again now. And I know I wasn’t the only one in class who felt that way.

Still, in the end, we didn’t have anything to worry about. The AIs were really nice to us.

***

The idea of artificial intelligences advanced enough to perfectly mimic humans was most famously proposed by mathematician Alan Turing, one of the founders of the modern computer age, who helped the Allies win World War II by breaking top-secret Nazi codes and would’ve turned 100 this year. With the most famous version of his “Turing test,” he asked whether a machine could impersonate a human well enough in a conversation over text to be indistinguishable from human. If so, one could say that computer is at least as intelligent as a human.

However, Turing’s original game didn’t focus on the question of human or machine at all, but a test to see whether a speaker was a man or a woman. Seen in this light, Turing tests can be used to ask more than just what makes us human, or what makes us intelligent. They can be used to ask what it’s like to be someone or something else.

There have been other kinds of Turing tests that people have proposed. Conservative economist Bryan Caplan developed the notion of an ideological Turing test in response to liberal economist Paul Krugman — could conservative or liberal partisans explain the beliefs of their opponents well enough that an outsider couldn’t tell if they were the real thing? Atheist Leah Libresco created a similar Christian or atheist Turing test (hat tip to Sean Carroll).

I talk about Turing tests now because interest in computers that can pass them are reviving, according to Robert French, research director of cognitive science at the French National Center for Scientific Research. He writes in this week’s issue of Science about technological advances such as a new generation of experimental “neurosynaptic” microchips IBM recently unveiled that are based on the computing principles underlying our brain’s neurons.

The question of whether or not a machine can pass the Turing test is a profound one. However, just as profound might be the question of whether or not we can pass Turing tests — whether we can understand what it’s like to be human, or other kinds of humans. Turing himself lived in a society that did not understand what it was like for him to be gay, and he likely died of suicide by cyanide after he was chemically castrated by the UK government.

I don’t really think Turing tests will ever become de rigueur, but it would be interesting if they did become more commonplace. In debate classes, debaters regularly take up views opposite to their own and win — my friends tell me really good debaters depend on skill and wit, not emotion and conviction.

Ultimately, understanding humanity is what Turing tests of all kinds are about.

You can email me regarding Visions at toohardforscience@gmail.com.

Charles Q. Choi About the Author: Charles Q. Choi is a frequent contributor to Scientific American. His work has also appeared in The New York Times, Science, Nature, Wired, and LiveScience, among others. In his spare time, he has traveled to all seven continents. Follow on Twitter @cqchoi.

The views expressed are those of the author and are not necessarily those of Scientific American.

Tags:





Rights & Permissions

Add Comment

Add a Comment
You must sign in or register as a ScientificAmerican.com member to submit a comment.

More from Scientific American