20Q
From Wikipedia, the free encyclopedia
20Q is a project in Artificial Intelligence (AI).
[edit] Principle and history
20Q began its life on a floppy disk as an experiment in artificial intelligence (A.I.). The principle is that the player thinks of something (either abstract or not) and then 20Q asks twenty questions, such as "Is it smaller than a breadbox?" or "Would you touch it with a ten foot pole?" The player can answer these questions with: Yes, No, Unknown, Irrelevant, Sometimes, Maybe, Probably, Doubtful, Usually, Depends, Rarely, or Partly. The experiment is based on the classic word game of "Twenty Questions," and on the computer game "Animals", popular in the early 1970s, which used a somewhat simpler method to guess an animal.
After you have answered the twenty questions it poses (sometimes fewer), 20Q makes a guess, and you can choose whether or not it is correct. If not, it asks some more questions, then guesses again. The 20Q A.I. uses a true neural network to pick the questions and to guess. It makes guesses based on what it has learned; it is not “programmed” with information or what the inventor thinks. It has learned by playing against people who visit the website and play against the online A.I. Answers to any question are based on players’ interpretations of the questions asked. 20Q also draws its own conclusions on how to interpret the information. It can be described as more of a folk taxonomy than an taxonomy. Its knowledge develops with every game played. In this regard, the online version of the 20Q A.I. can be inaccurate due to the fact that it gathers its answers from what people think rather than from what people know. Limitations of taxonomy are often overcome by the A.I. itself because it can learn and adapt. For example, if the player was thinking of a "Horse" and answered "No" to the question "Is it an animal?", the A.I. will, nevertheless, guess correctly, despite being told that a horse is not an animal.
The 20Q A.I. is adaptable, scalable, modular and embeddable, and for this reason, it is possible for it to expand its knowledge and learn about more specific things. It is now learning in twenty-one languages, as well as everything it can about music, sports, movies and television.
As noted by the inventor, Robin Burgener, the “Uncommon Knowledge” generated by the A.I. at the end of a game is what it comes up with when something seems odd to it and it can’t fit it in with what it knows. The questions it poses at the end of a game are not made up by humans; it is information the A.I. thinks up on its own, generating answers based on what it has learned and what it knows. Over time, it will be able to make more refined distinctions. 20Q learns, and learns to make distinctions, through play—the more times you think of an object and play (answering correctly, it is hoped) the more it learns about that object. The online 20Q A.I. has about 10,000,000 synaptic connections.
The 20Q A.I. was invented by Robin Burgener, of Ottawa, Canada, in 1988. Thousands and thousands on people play online each day, which means that the A.I. learns on a broader level than it did in 1988 when it lived on a floppy disk that Mr. Burgener swapped amongst friends to help it learn. Originally, it knew one object and one question. After it moved to the internet in 1995, and more people started playing, its level of learning increased, as well as the connections it can make when guessing. The 45,000,000th game was played in November of 2006. Mr. Burgener notes that the success rate of the online A.I. is between 73 to 78 per cent. According to Mr. Burgener, the "real" success rate is higher, but he has "dumbed it down" in order to make ity more interesting for the players--if 20Q won every game, all the time, as it is capable of doing, no one would come back to play and the A.I. would not learn.
[edit] Modularity of the artificial intelligence
The modular capability of the 20Q artificial intelligence means that it can be embedded in any small screen device. At age sixteen, it finally got a job when a subset of the 20Q A.I.'s knowledgebase was licensed to a toy company for use in a handheld version of the A.I. in 2004. The device contains a small portion of the original 20Q website knowledgebase, and unlike the online versions of the game, the toys do not have the ability to learn. However, this artificial intelligence is unique from less flexible, and extremely large, expert systems. Its modularity, adaptability and scalability means that it can be applied to other, more complex devices, for more complex uses. As an outcome of the experiment, it is interesting to note that the A.I. has applications that go beyond the playing of the word game.