I awake in a hotel bed 5,000 miles from residence in Seattle. After breakfast, I venture out of the sea and explore the coastline of the nation that created my language. However, I'm unable to understand most of the signs I encounter in my travels: "LET AGREED," the one that is prominently written in large font but it doesn't mean anything to me. I stop, and stare at the water in a trance for a few seconds, trying to understand and rearranging the message. Usually, these types of cultural and linguistic gaps are fascinating to me. Today they're generally a source of concern. In about two hours, I'll sit on a laptop and conduct a series five-minute instant message chats with strangers. On the other side of these conversations is a psychologist an linguist, researcher in the field of computer science, as well as host of a renowned British tech show. Together, they will form the judging panel, which will evaluate my abilities to perform one of the most bizarre things I've ever had to perform.
The test's name is derived from Turing, the British mathematical genius Alan Turing, one of the pioneers of computer science. He in 1950 tried to solve one of the first questions: could machines be thought of as intelligent? That is the question: is it feasible to create an advanced computer that it could be considered to be thinking, intelligent, or to possess the ability to think? If indeed the possibility of this kind of machine, what would we know? Instead of arguing about this issue on only theoretical grounds, Turing proposed an experiment. Judges each ask questions, using a computer terminals, to a number of pair of unidentified corespondents each one of whom is computer program and one a human "confederate," the other computer program, in order to identify which is the other. The conversations can range from casual conversation to naive questions, from gossip from celebrities to serious philosophical debates -- the entire gamut of conversation between humans. Turing predicted that by 2000 computers could be capable of fooling 30 percent of human judges within five minutes of conversations, and as a consequence the user will "be able to speak of machines thinking without expecting to be contradicted. Turing's prediction did not proved to be true, However, in the 2008 competition the computer program with the highest score was unable to meet the mark by one vote. After reading the news I was aware that the 2009 test at Brighton may be the most decisive one. I'd never been to the event however I was convinced that I was required to go, not only as a mere spectator however, but to be an element of human defence. The thought of going head-to-head (head-to-motherboard?) against the top AI programs made me feel awestruck. an idea of a dream that, as a confederate, would be fighting for humanity like Garry Kasparov's chess game with Deep Blue. During the contest the judges will have the conversation with the contestants for five mins, and then the next, and finally will be given 10 minutes to think about and determine which judge is the most human. The judges will also grade all contestants, and this can be used to break the tie. The program that receives the most votes and the highest score from judge (regardless of whether or not it passes the Turing Test by fooling 30 percent of them) is given the title of the most Human Computer. This is the title the teams of researchers are trying to win and the one that comes with a cash prize (usually around $3,000) and the prize that everyone who participates in the contest is worried. However, there is interestingly, a different title which is given for members of the confederatewho has the most convincing argument most convincing: it's the Most Human Human award. A winner of this award in 1994 was the science-fiction journalist and author Charles Platt. How'd he do it? by "being moody, irritable, and obnoxious," according to what he wrote within Wired magazine. I find it not only funny and dark and bleak, but also, in a more fundamental sense, a plea to be a better person to ask how, actually can we become the most human that we can be, not only within the limitations in the testing, but also in the real world?
Kommentare