A Brief Overview of the Turing Test: A Test designed for testing the intelligence of machines
- Jan 02, 2018
Developed by Alan Turing in 1950, the Turing Test was designed to test a machine’s ability to show intelligent behavior.
The test dictates that a machine has to imitate a human being not just in terms of looks/appearance but also on the basis of intelligence. In simpler terms, the machine has to disguise itself as a human being. If it successfully disguises itself as a human, it passes the test.
The Turing test will take place between a machine and a human being in the presence of a human evaluator.
The evaluator will know only one thing- s/he is testing a machine and a human being but s/he won’t be aware of the fact that which one among the two is a machine and which one’s a human.
The evaluator will then converse with the two of them in the form of text. Textual conversation is preferred, so the language and speech recognition don’t act as the barrier.
The entire conversation should be limited to the screen only.
The machine has to behave like a human. The goal of the machine should be to disguise itself in such a way that the evaluator fails to realize that s/he is conversing with a machine. If the evaluator makes a wrong choice or fails to detect the machine among the two, the test is a success.
Image: The standard interpretation of Turing Test
According to the diagram up above, A’s a machine and B’s a human being. C’s the human evaluator. C is aware that s/he is evaluating a machine and a human, but he doesn’t know the fact that A (specifically) is a machine and B (specifically) is a human. The evaluator’s task is to determine which one among the two (A and B) is a human and which a machine through textual conversation. If the machine misguides the evaluator by disguising itself as a human being, it passes the test.
Turing Test: Strengths
The main strength of the Turing test is its simplicity. Anything simple is beautiful, and Turing test is nothing different.
The definition of this term “intelligence” is very hard to conceive from the machines’ point of view. Turing test might not be perfect from all angles, but it can definitely do the job reasonably well in comparison to the rest.
Breadth of the subject content
The format of the Turing test aids the interrogator by providing him/her with a number of options to test the capability of a machine in performing intellectual tasks.
The test can be taken in natural languages. This is a huge positive because of the total absence of language barrier which would have created a huge test inconvenience.
The absence of voice
Since there’s no involvement of voice, the interrogator will not be able to determine which is a human and which is a machine from voice itself.
Turing Test: Weaknesses
The Turing Test doesn’t test for general machine intelligence. It only checks whether the machine is equally intelligent to that of a human or not.
There’s one thing that cannot be denied that Turing Test’s simplicity makes it a stand out candidate among all the methods used for testing AI. With that, we’ll sign off finally for the day. Hope you had an excellent read.
Learnpick is an online marketplace for all kinds of tutoring and learning services. We provide a platform to help learners find home tutors, online tutors, expert trainers,tutorial centers, training institutes and online tutoring schools. Learners can also find local classes, workshops, online courses and tutorials on a huge variety of topics.
Latest posts by LearnPick (see all)
- CHED advised public universities to start classes in August - Apr 18, 2019
- Schools should stay as open communities said Ong Ye Kung - Apr 17, 2019
- Indonesians vote today, to choose the next President - Apr 17, 2019
- Australian University Signs a $100m China Deal - Apr 17, 2019
- UNICEF praises Thailand for promoting early childhood education - Apr 16, 2019