Skip to content | Skip to menu | Skip to search

Banbury Cross

a pillow for lost thoughts...

Post details: Rossum's Universal Robots

Rossum's Universal Robots

In 1920, at the age of 30, the Czech writer Karel Capek wrote a play about a future populated with intelligent androids. The play was titled "R.U.R", an acronym for the company manufacturing such creatures/machines, and premiered in Prague a year later. When searching for a good descriptive expression, his brother Josef suggested the word "robot" - derived from an old Czech word "robota" meaning "serf labor" (the forced work of feudal vassals toiling on their Lord's property).

Nearly a century later, the word robot found its way into many languages and robots themselves, albeit a bit simpler than Capek's nearly-human androids, are invading the world of manufacturing at a break neck speed. And apparently not just manufacturing.

Whenever I listen to our beloved politicians, I start thinking that their ranks have been infiltrated by an army of nearly indistinguishable androids. I watched a couple of political debates recently, and I could not help noticing that its participants often fell into two broad categories.

The first kind reacted like normal people. They listened to the question and then tried to answer it the best way they could. Sure, you might have disagreed with them, but they made sense - at least from their point of view. Tom Coburn, Elizabeth Warren, Ted Kaufman or Marcy Kaptur are examples I can think of from the top of my head. These people usually think in terms of complete sentences or whole paragraphs, not just simplistic and easily digestible sound bites or clearly ideological bullet points. They might rant on occasion, they might ramble here and there, but they will tell it like it is.

On the other side, there is a different kind of animal. Politicians who in their effort not to alienate any segment of society will scrub their thinking of any potential controversy, and hence also of any potential meaning. These are politicians who - when you hear their answers - make you immediately wonder if they actually heard the question. Lindsey Graham, Nancy Pelosi, Michele Bachman, or Chris Dodd are the first examples to come to my mind. To a human ear their answers will always seem a bit incoherent, intellectually stiff, formulaic and even evasive.

A typical conversation goes like this:

Reporter: "Senator, our viewers would love to know what you had for breakfast today.

Politician: Breakfast is a very important segment of every healthy individual's daily regimen.

Reporter: That is definitely true, but I am sure you must have a favorite meal.

Politician: I think all meals have significant nutritional value and my office would be happy to provide some specific details.

Reporter: Are you trying to make a point that we should pay more attention to the quality of our nutrition?

Politician: Ummmm, again, I think all meals have significant nutritional value and I am sure my staff would be happy to give you specifics.

In this kind of dialogue, something doesn't quite add up, does it? You quickly get the sense that the two parties are talking past each other. Almost as if politicians were some kind of pre-programmed automatons. Sometimes, you can track this suspicion step by step: they digest the verbal input, turn it into a stream of salient keywords, then scan the bank of hundreds of thousands of possible responses, find the closest match, run it through a politically correct filter and finally send it to the output. In other words, they seem to be driven by a complex but purely mechanical meta-algorithm in lieu of a normal human brain - you know, that old fashioned and finicky medium that gave us such anachronisms as beliefs, visions, principles, leadership and integrity. That is something a Natural Language Processor - no matter how sophisticated - will never fully emulate (or at least not in the foreseeable future).

See, we are spending all this money on the research in Artificial Intelligence, and yet all that time the answers may be lying at our doorstep. Or more precisely, at the doorstep of the Capitol Hill. So why don't we just send a team of our top notch researchers to the DC area, let them catch one of those political androids, find their Central Processing Unit, open it and see how they are wired. I think this would further the development of robotics at least by 50 years.

And save the taxpayers untold millions, too.


No Comments for this post yet...

Comments are closed for this post.

This site works better with web standards! Original skin design courtesy of Tristan NITOT.