Research develops new Turing test
27
Nov
Alan Turing a pioneer of Computing, which developed a theoretical model called after Turing machine, completed after mathematician Alonzo Church in 1938, which gave rise to all computing algorithms, the basic principle stated:
“The algorithm consists of a finite set of simple and precise instructions that are described by a finite number of symbols.”
About this idea was proposed a test, called the Turing Test, that if something was proposed to a person and a machine and we could not decide which was the response of a machine, would have been an “intelligence” machine.
Now a professor at Georgia Tech, Mark Riedi claimed to have developed an algorithm that describe it as a more accurate way to evaluate the potential of a smart machine than the famous Turing test, as the news in Georgia Tech site.
The new Riedi test, called Artificial Intelligence Creativity Test and the Lovelace 2.0 are based on Lovelace test, created in 2001, testing the production of a work of art or other creative enough human labor.
Ried believes that the original test was very subjective and created specific parameters for your test to be followed by a human reviewer. This allows the evaluator to assess the creative work of a machine without making subjective judgments of value.
Riedi will be presenting its new test with the Beyond the Turing Test, in Workshop of the Association for the Advancement of Artificial Intelligence (AAAI) in January 2015, in Austin, Texas.
