4 ai systems will never really solve the problem of

This preview shows page 390 - 392 out of 408 pages.

4) AI systems will never really solve the problem of language understanding because they cannot deal effectively with the meaning (i.e. semantics) of language. Answer: The correct answer is (4) Refer: Studyguide pg. 104 - Chapter 13 Reasoning: Just like the man in the Chinese room, a computer program also merely manipulates symbols according to a set of rules. It executes instructions to manipulate symbols and strings of symbols, but has no understanding of what these symbols mean.
391 PYC3703 Cognition: Thinking, Memory and Problem Solving QUESTION 24 Select the best description of the Turing test from the explanations below: 1) John: “The Turing test is used to evaluate computer expertise in specific areas of knowledge. The test is performed to see how well an artificially intelligent program compares with a human expert in solving problems relating to the area of knowledge.” 2) Thande: “The Turing test involves a computer interacting with a Chinese speaker in order to determine whether the computer can translate English sentences correctly into their Chinese equivalents.” 3) Mariannne: “The goal of the Turing test is to determine whether a computer can emulate a human so well that a human conversing with the computer would believe that he or she is talking t o a human being and not a computer.” 4) Joseph: “Allen Turing devised the Turing test in order to establish whether a computer can be programmed to produce grammatically correct sentences, and to simulate a psychotherapist.” Answer: The correct answer is (3) Refer: Studyguide pg. 100 Chapter 13 Reasoning: The Turing test was developed by Alan Turing because he maintained that it was not productive to try to answer ill- formed questions such as “What is intelligence?” and “Is a machine intelligent?”. He sugg ested instead that a simple behavioural test for computer intelligence be employed. 2014 Exam October-November QUESTION 25 Which theory of intelligence is based on the hypothesis that intelligence involves a hierarchical structuring of cognitive abilities? 1) Carroll Three-stratum Model of Intelligence 2) Gardner Theory of Multiple Intelligences 3) Sternberg The Triarchic Theory 4) None of these theories is based on such a view of intelligence. Answer: The correct answer is (1) Refer: Chapter 13 Reasoning: Gardner's theory exemplifies modularity because it is based on the notion that intelligence involves eight different types of abilities that are separate from one another. Gardner argues that these intelligences are distinct, reflecting different abilities, and that different brain structures are associated with them. He therefore proposes a modular theory of intelligence. Caroll and Sternberg both reject the notion that intelligence is a unitary construct, however they do not present theories that are specifically modular in approach. Thus, Carrol’s theory is hierarchical , and although Sternberg’s triarchic theory is based on the assumption that intelligence derives from different components, he does not emphasize modularity as such, and also does not claim that different brain structures underlie these components.

  • Left Quote Icon

    Student Picture

  • Left Quote Icon

    Student Picture

  • Left Quote Icon

    Student Picture