In the 1950s, programmers began speculating about the idea of intelligent computers. Inventor Alan Turing declared that the first major milestone of this effort would come when there is no clear distinction between a conversation with a computer and with a human. Turing wrote, “I believe that at the end of the century the use of words and general educated opinion will have altered so much that one will be able to speak of machines thinking without expecting to be contradicted.”
Turing’s Al litmus test requires that at least 30% of humans be fooled into thinking that the robot is human. On Sun., Jun. 6, at the Royal Society in London, a computer program managed to convince 33% of a judge panel that it was 13 year old Ukrainian boy. In this competition, each of the programs was required to participate in a series of five minute, text based conversations. This was the first time that a computer has passed this test in history.
The program is called “Eugene Goostman” and it is open for public experiment online. However, because of the immense traffic it has recently caused, the site may not be accessible for a few days. Eugene’s makers described the process of making this program, saying, “You don’t write a program, you write a novel. You think up a life for your character from scratch – starting with childhood – endowing him with opinions, thoughts, fears, quirks.”
This momentous achievement has raised many, heavy questions that people have been asking for years. Principally, the question of what qualifies as “intelligence.” The programmers of Eugene Goostman managed to make the program’s intellectual abilities resemble that of a real human. One of the programmers mentioned, “Our main idea was that he can claim that he knows anything, but his age also makes it perfectly reasonable that he doesn’t know everything. We spent a lot of time developing a character with a believable personality.”
Nevertheless, it is difficult to say whether making the chatterbox a young boy is a breakthrough or a cop out. The issue of the turing test’s foundation then arises. It seems like a rather outdated and ineffective means of determining the presence of true intelligence in a computer.
This is why there is a new standard for the modern turing tests. Inventor Hugh Loebner outlined these rules in 1990. According to Loebner’s standards, the conversation must last a duration of 25 minutes with four separate judges. In this instance, the computer must fool at least half of the judges.
Another test, the most difficult version of the turing, was constructed by Mitch Kapor, Lotus founder, and Ray Kurzweil, futurologist. In their spin on this test, the robot must have two hour long discussions with three judges and ultimately convince ⅔ that it is human. Additionally, it must be ranked as being “more human” than at least two of the human competitors. Kapor bet that no robot could pass this test before 2029 and the scientists have laid $20 thousand on the bet.
Either way, the most important quality of the product is how useful it will be. It must be acknowledged that Eugene Goostman is not a super-computer, it is not even a computer, but rather a chatter-box. The main significance here is not necessarily of artificial intelligence, but rather the growing difficulty in distinguishing chatterboxes from humans.