[Correction: Mr. Scocca informs me via email that the dialog in his piece was an actual transcript of a session with Eliza. So nothing “mock” or “contrived” about it. All the more interesting, read on.]
Over at Slate, Tom Scocca has staged a mock dialogue with Eliza to good, even if somewhat contrived, effect. Eliza, in case you’re wondering, is a computer program designed in the 1960’s to simulate human interaction based on certain scripts (such as DOCTOR which simulates a particular style of psychotherapy).
About half way through the dialogue we get these few lines which more or less encapsulate Scocca’s main concern:
TS: On Twitter, I am having a hard time telling the bots from the humans.
Eliza: How long have you been having a hard time telling the bots from the humans?
TS: Weeks and weeks now.
Eliza: Please go on.
TS: It is not because the bots are fluent or facile.
Eliza: Do any other reasons not come to mind?
TS: It is because so much of the human conversation has come down to bot level.
And a few lines later he adds,
TS: That’s not really someone who likes my blog post. It’s a pharma ad-bot that presumably grabs anything with “Obama” in it.
Eliza: Why do you say your blog post its a pharma adbot that presumably grabs anything with obama in it?
TS: Well, of course, by now I have been trained to put “Obama” into the headlines myself, for SEO purposes.
Eliza: Can you elaborate on that?
TS: My actual transactions with human readers are so machine-mediated, it’s necessary for me to keep machine-logic in mind while I’m writing.
I’m taking these observations as rather useful illustrations of how the language (or, logic) of a digital media platform shapes our communication to fit within its own limitations. Borrowing linguist Roman Jakobson’s maxim regarding languages, I suggested a few posts down that, “Languages of digital media platforms differ essentially in what they cannot (or, encourage us not to) convey and not in what they may convey.” In other words, we shape our communication to fit the constraints of the medium. The follow up question then becomes, “do we adapt to these limitations and carry them over into other fields of discourse?” Scocca provocatively suggests that if a computer ends up passing the Turing Test, it will not be because of an advance in computer language capability, but because of a retrogression in the way humans use language.
Keep in mind that you don’t have to be a professional writer working for a popular web magazine to experience machine mediated communication. In fact, my guess is that a great deal, perhaps the majority, of our interaction with other people is routinely machine mediated, and in this sense we are already living in post-human age.
The mock dialog also suggests yet another adaptation of Jackobson’s principle, this time focused on the economic conditions at play within a digital media platform. Tracking more closely with Jackobson’s original formulation, this adaptation might go something like this: the languages of digital media platforms differ essentially in what their economic environment dictates they must convey. In the case of Scocca, he has been trained to mention Obama for the purposes of search engine optimization, and this, of course, to drive traffic to his blog because traffic generates advertising revenue. Not only do the constraints of the platform shape the content of communication, the logic of the wider economic system disciplines the writing as well.
None of this is, strictly speaking, necessary. It is quite possible to creatively, and even aesthetically communicate within the constraints of a given digital media platform. Any medium imposes certain constraints; what we do within those constraints remains the question. Some media, it is true, impose more stringent constraints on human communication than others; the telegraph, for example, comes to mind. But the wonder of human creativity is that it finds ways of flourishing within constraints; within limitations we manage to be ingenious, creative, humorous, artistic, etc. Artistry, humor, creativity and all the rest wouldn’t even be possible without certain constraints to work with and against.
Yet aspiring to robust, playful, aesthetic, and meaningful communication is the path of greater resistance. It is easier to fall into thoughtless and artless patterns of communication that uncritically bow to the constraints of a medium thus reducing and inhibiting the possibilities of human expression. Without any studies or statistics to prove the point, it seems that the path of least resistance is our default for digital communication. A little intentionality and subversiveness, however, may help us flourish as fully human beings in our computer-mediated, post-human times.
Besides, it would be much more interesting if a computer passed the Turing Test without any concessions on our part.
Oh, and sorry for the title, just trying to optimize my search engine results.
2 thoughts on “Obama Talks With A Computer”
Interesting post. Thanks for that.
Did you happen to see the upcoming human vs. machine “Jeopardy!” match that is akin to Kasparov vs. Deep Blue? http://www.bloomberg.com/news/2010-12-14/ibm-computer-to-match-wits-with-jeopardy-champs-in-1-million-showdown.html?cmpid=yhoo
Interesting, take a look here too: http://futurisms.thenewatlantis.com/2010/12/progress-in-robotics-and-ai-coming.html