January 27, 2016 6:08 pm
Ex Machina is a 2015 British sci-fi psychological thriller, written and directed by Alex Garland and recently nominated for Academy Awards in the Best Original Screenplay and the Best Visual Effects categories. Starring Domhnall Gleeson, Alicia Vikander and Oscar Isaac, it tells the story of programmer Caleb Smith (Gleeson), who is invited by the eccentric billionaire Nathan Bateman (Isaac) to administer the Turing test to a female android with artificial intelligence (Vikander).
For those of you who don’t know what it is, the so-called Turing test was developed by Alan Turing in 1950. It should help determine “a machine’s ability to exhibit intelligent behaviour equivalent to, or indistinguishable from, that of a human”. What’s most important, however, Turing’s main criteria to pass the test is that a human evaluator would judge natural language conversations between another human and the machine which is designed to generate human-like responses.
As it is shortly explained in today’s videoclip too, it is of course assumed that “the evaluator would be aware that one of the two partners in conversation is a machine, and all participants would be separated from one another. The conversation would be limited to a text-only channel so that the result would not be dependent on the machine’s ability to render words as speech. If the evaluator cannot reliably tell the machine from the human, the machine is said to have passed the test.”
On the other hand, over the years since the creation of the Turing test, researchers in computational linguistics were able to note the importance of arbitrary world knowledge in even the simplest natural-language processing tasks as might be required for machine translation.
For example, Bar Hillel stated that there are some extremely simple sentences in English which, within certain linguistic contexts, would be “uniquely and unambiguously translated into any other language by anyone with a sufficient knowledge of the two languages involved”. Yet, not a single machine would come up with the same unique rendering unless by a completely arbitrary – meaning ‘not mechanic’ – procedure.
A sentence of this kind could be “The box was in the pen.” The linguistic context from which this sentence is taken is, let’s say, “Little John was looking for his toy box. Finally, he found it. The box was in the pen.” Let’s assume that pen in English has only two meanings, which are 1) a writing utensil, 2) an enclosure where small children can play. It is then possible to say that no electronic computer would determine that the word pen in the given sentence and within the given context has the second meaning, whereas every reader with a sufficient knowledge of English will do this spontaneously.
Ever since Bar Hillel expressed this quite pessimist view, there has been wide acknowledgement that natural-language-processing tasks, even simple ones, require “knowledge and reasoning of essentially arbitrary complexity”.
- The Turing Test: Verbal Behaviour as the Hallmark of Intelligence
- Wikipedia: The Turing Test
Written by Eva Barros Campelli
Communication trainee at TermCoord
2,412 total views, 4 views today