Translator: Nirmin Alhashoum Auditor: Ahmad Adel What is awareness? Can an artificial machine really think? Is the mind composed of nerve cells in the brain only? Or is there an intangible spark at its core? For a lot of people, these were basic considerations For a future of artificial intelligence But British computer scientist Alan Turing decided to overlook these questions With a much simpler question: Can a computer talk like a human? This question leads to a concept for measuring artificial intelligence Which will become very well known as the Turing test. In a 1950 paper, "Calculators and Intelligence" Turing put forward the following game: Judge Al-Bushra has a text chat with participants he does not see It evaluates their responses. The computer must be able to replace one of the limbs to pass the test Without actually changing the results. In other words, the computer will be considered intelligent If it is difficult to distinguish his conversation from human conversation. Turing predicted that by the year 2000, Machines with 100MB memory will easily pass its test.
But he may have anticipated events. Although today's computers have much more memory Except that few passed the test, And the ones that performed well She focused on finding clever ways to deceive the rulers Most of its reliance on great computing power. Although it has never undergone a real test, However, the reason for the success of the first program was called "Elisa". Using only quite simple and short text He managed to mislead many people by copying him to a psychiatrist, And encourage them to talk more, And do the opposite of their questions to put them to them. Another text called "Barry" took the opposite approach By imitating the schizophrenic and insane patient The one who kept directing the conversation to his preprogrammed thought.
One weakness of the test was highlighted by their success in fooling people. People usually attribute intelligence to a whole host of things They are not really smart. However, the annual competitions were held as the "Leubner Prize." By making the test more formal Putting rulers know in advance Some of their interlocutors are machines. But although the quality has improved, However, many chat programmers used strategies similar to "ELIZA" and "Barry". 1997 Winner Catherine managed, Of having an amazingly witty and focused conversation, But it is mostly if the judge wants to talk about "Bill Clinton." And another new winner was given, "Eugene Gustman", Character of a 13-year-old Ukrainian boy, The judges interpreted the inconsistencies and strange grammar rules As linguistic and cultural barriers.
While other programs such as "Cleverbot" have taken a false path, Statistical analysis of huge databases of real conversations To determine the best responses. While some also stored memories of previous conversations In order to improve it over time. But while the responses of "cleverbot", each separately appear to be the responses of humans to a large extent, However, her lack of character constancy And its inability to deal with new topics, The disadvantages are obvious.
Who would have expected in the Turing days that today's computers? She will be able to pilot spacecraft, And performing delicate surgeries, And solve the colossal equations, But she still faces difficulty in the most basic and simpler conversations? And it turns out that human language is an incredibly complex phenomenon of dasha As it cannot be enumerated in even the largest dictionary. Chatbots may be disturbed by simple breaks such as "mmm". Or questions that do not have correct answers. It might require a simple conversation sentence like, I took the juice from the refrigerator and gave it to it, But I forgot to check the history. " A wealth of tacit knowledge and intuition to analyze it. It turns out to be a simulation of human conversation It requires more than just an increase in memory and processing capacity. As we approach the "Turing" goal, Perhaps we should finally deal with all of those big consciousness questions..
