Friday, February 25, 2011

Ayo Technology!


# Ja tvoi sluga,
# ja tvoi Rabotnik

# Ja tvoi sluga,
# ja tvoi Rabotnik

We geeks (yes, geeks) already had a good month with IBM's Watson winning Jeopardy, beating the crap out of two human champions. Artificial inteligence is fascinating. But now I got news from a carnivorous robot and Thursday the first two-legged robot marathon kicked off in Osaka.

Run robot, run!

picture taken from Geekologie
Thursday, five knee-high androids have crossed the starting line in the world's first marathon for two-legged robots. The race is being organised by local authorities and Vstone, a local robotics firm, which hopes the marathon will become an international event in the future.
The contestants are expected to take four days to complete the course of 42 kilometers, which involves 423 laps of an indoor track. Operators are allowed to change the robots' batteries and motors but if the machines fall over they must get up by themselves.

Organisers say they expect the race to be won not by the fastest robot but by the one that can withstand the most wear and tear. The winner (Vstone's own entry: the Robovie PC) took only 54 hours 57 minutes 50 seconds to finish the race [1, 2]. One second later, the robot with (in theory) the fastest walking speed crossed the line, it was the radio-controlled robot entered by the Center for Robotics.

So now that I read "radio-controlled", I understand that pure artificial intelligence was not a requirement. I have to admit that five, knee-high robots walking a track for more than two days seemed kind of a cute to me, it reminded me of my old robot toy. But when I saw that "if the machines fall over they must get up by themselves", I had to surpress the mind image of an army of Terminators, crawling in my direction and getting up by themselves any time they got shot. Brrrr...

More:

Eat robot, eat!

In some kind of eighties-nostalgia, robots walking a marathon could be considered cute, but the idea of a "clock that keeps itself in motion by catching and eating flies" just sends shivers down my spine.


The designers themselves, James Auger and Jimmy Loizeau, put it rather nicely and call it "making a domestic robot self sufficient": eight flies will make the clock tick for twelve days.
Well hey hey hey, you will never have to buy batteries again and you finally get rid of all those flies, what a bargain!

From a science point of view, this is very interesting, but personally it freaks me out. Before this news, I was already wondering how far we could go in the research of artificial intelligence while still keeping it ethical. How moral is it to re-invent the brain?
Technologically speaking: brilliant and purely fascinating, but how far can we go? And I am not talking about "robot armageddon" or so (don't worry, mankind will have destroyed itself long time before we get the robots fed up with us), it is purely the question whether we can treat our own, dear memories as pure data. Are all our decisions really only a matter of 1100110's?

As a computer scientist, reaching that perfect (imitation of? artificial?) intelligence, is the Holy Grail. But as a human being, I find it unethical to be able to completely map out (my) brain. Or maybe find a way to alter memories or even create false ones.

I am on a slippery slope here, because a decade ago I might have said the same about trying to map out the chemical DNA-structure of a human being. Despite me having my reservations about the topic, mapping out the DNA-structure hasn't really brought us in a state of Sodom and Gomorrah.
Of course it is not true that every (advanced) knowledge will bring us in an immediate state of moral decline, but we have to bear in mind that with power comes great responsibility. And without a "moral framework", things would definitely derail.

So from DNA-structure and AI back to machines that can eat. Okay, to some extent you could say that cars run on petrol and petrol comes from oil which is merely compacted dead bodies from centuries ago. But we are not confronted with some peaceful grass grazing metal cows at this moment, what the clock is doing is catching (mwah, more like "trapping") a living entity and digesting it to be able to run.
If it stays with using vermin for fuel, I guess that I can live with that, but for me we are close to my "edge of ethically tolerable".

But as I type, I am sitting in an open air office in the tropics, the night has fallen and because of the desklight and the reflection from the laptopscreen, insects are just buzzing around my head, making me crazy. At this moment, that clock would be very welcome indeed.

More:

2 comments:

  1. Documentary about the promise, problems and ethics of AI and robotics: Plug & Pray (Wikipedia, IMDB)

    I believe. Do you?

    ReplyDelete
  2. Also: http://cleverbot.com/ and http://www.existor.com/

    ReplyDelete

Related Posts Plugin for WordPress, Blogger...