Friday, November 10, 2006

Robot thinks "think we taste like bacon": will robots acquire a taste for us?

By this time, you will be aware of the cute robot, designed by NEC System technologies and Mie University, which has been described as "a metal gastronome" and "an electromechanical sommelier". It is supposedly capable of identifying wines, cheeses, meats and hors d'oeuvres. Upon being given a sample, it will produce in a childlike voice an identification of what it has just been given. The idea is that restaurants and individuals, for example, can be told if a wine is authentic without even opening the bottle. However, when a reporter placed his hand in the robot's omnivorous clanking jaw, he was identified as bacon. A cameraman then tried his hand and was identified as prosciutto.

Reactions to this story have varied from dire predictions of robots acquiring a taste for human flesh and therefore to a "robot holocaust", and through "I told you, robots will never be intelligent enough to take over the world", to stories about all the ills that might result from robot mistakes or malfunctions.

But before we become too terror-striken, let's just pause a moment. Are we sure that this was a mistake or malfunction? Note that the robot did not identify both the humans as the same, but one of them as "bacon", and the other as "prosciutto".

We know that humans often start smelling of what they eat (which is why, for example, some people don't like to consume garlic). Is it possible that the characteristics of human flesh also change, depending on the sort of food that one most eats? Further, is it possible that the reporter was particularly fond of bacon, while the cameraman particularly fond of prosciutto? Or that they had just consumed a quantity respectively of those meats before the experiment?

If so, I wonder what the robot would have pronounced if I had placed my hand there. My guess is that, since I am a vegetarian, it might have identified me as a vegetable!

Any prizes for guessing which one? Sphere: Related Content

No comments: