Page 149 - Robotics and AI class 10
P. 149

• Adaptability:  Humans  can  quickly  adapt  to  new  situations  and  learn  from  different  domains,  whereas
              machines need instructions or retraining to handle new tasks.
               • Social Skills: Humans understand social dynamics, emotions, and build relationships, while machines can
              simulate human-like responses but don't truly understand or have emotions.
               • Senses: Humans have various senses like sight, hearing, touch, taste, and smell, which help gather information.
              Machines have limited sensors and need specific data inputs to make decisions.
               • Context: Humans understand the context of situations, considering cultural, historical, and situational factors.
              Machines require explicit programming to consider context and lack human-like understanding.


            Turing Test, What Is It?

            The Turing Test got its name from Alan Turing’s name who was a computer scientist, mathematician, cryptanalyst
            and theoretical biologist. It was originally called the “imitation game” by Alan Turing. He introduced it in 1950, in
            a paper "Computing Machinery and Intelligence" which he wrote while working at the University of Manchester.
            This test finds out a machine's ability to exhibit intelligent behaviour similar or equivalent to or indistinguishable
            with  that  of  a  human.  In  other  words  in  the  field  of  Artificial  Intelligence  it  is  a  way  to  find  out  whether  a
            computer is capable of thinking like a human being or not.


            Role of the Turing Test in AI

            There are several versions of the Turing Test. The simplest of all is the Imitation Game which originally had 3
            Terminals to determine if the computer has Artificial Intelligence and can think and replicate or respond like
            humans would. All the 3 terminals were physically separated from each other. Out of the 3 terminals, two were
            operated by Humans, and one by the Computer.

            Now let us understand the process in detail: In the test, one of the humans was a questioner. The second human
            and the computer were the respondents. The human questioner asked questions on specific subject areas, which
            were in a specified format and context, to the respondents. After a specific time and after a series of questions,
            the questioner would adjudicate, as which respondent was human and which was a computer. For example, a
            question may ask, ”What is the difference between a basketball or a football?”, or it may be “Which season gets
            a lot of snowfall?”, or it may be “What is the difference between a pencil and a pen?”

            This test was repeated multiple times, each time a human with the questioner would need to adjudicate which
            respondent is human and which is computer. After repeating the test, if the human with the questioner feels that
            answers from the computer were reasonably close to that of humans in most of the cases, it would deem that
            the computer has Artificial Intelligence.


            Uses of Turing Test

            Turing test finds its uses and applications in a number of places. Some of its are discussed below:
               • Humans can be biased in judgments but with the Turing test in controlled procedures and environment with
              hidden participants details proves to be unbiased and accurate. This will be in the form of a quiz where some
              abstract or emotional questions will be asked and the answers to these questions will often be judged based
              on the type of the answer given.
               • CAPTCHA is one type of Turing Test to find out that the website which is required to be accessed is by a human
              or a computer. In this test a person is supposed to answer a question based on the set of images displayed.






                                                       Machine Intelligence and Cybersecurity in Computing  147
   144   145   146   147   148   149   150   151   152   153   154