JARDIM DIGITAL W4LKER

NOTAS E IDEIAS

170120231159

criado em: 11:59 2023-01-17

Relacionado

The Turing test: AI still hasn’t passed the “imitation game”

A computer that could decidedly pass Alan Turing's test would represent a major step toward artificial general intelligence.

fonte big think

Alan Turing posed the question "Can machines think?" in his 1950 paper "Computing Machinery and Intelligence" and proposed the Turing test as a way to measure the sophistication of computers. The test involves an interrogator asking written questions to two players, one human and one computer, and trying to determine which player is the computer by evaluating the "humanness" of their responses. The test was not designed to determine if a computer can think, but rather to establish a measurable standard for assessing computer sophistication and sidestepping philosophical questions about defining intelligence and thinking.

The Turing test is a measure of a computer's ability to mimic human intelligence, proposed by Alan Turing in 1950. To date, no computer has passed the test conclusively, but some have come close. In 1966, a chatbot called ELIZA was able to fool some human interrogators, but the rules and designs of the test have been debated among computer scientists. In 2014, a computer algorithm called Eugene Goostman convinced one-third of human judges that it was human, but it had the advantage of claiming to be a 13-year-old boy from Ukraine. In 2018, Google's virtual assistant Duplex was able to successfully book an appointment over the phone without the person on the other end realizing they were talking to a computer.

Alan Turing, in his 1950 paper "Computing Machinery and Intelligence", posed the question "Can machines think?" and proposed the Turing test as a way to measure a computer's ability to mimic human intelligence. The test involves a human interrogator asking written questions to a human and a computer, and trying to determine which player is the computer by evaluating the "humanness" of the responses. To date, no computer has passed the test conclusively, but some have come close, such as the chatbot ELIZA in 1966 and a computer algorithm called Eugene Goostman in 2014. The Turing test is highly debated among computer scientists due to the ambiguity of the rules and varying designs of the test.