Using the Imitation Game to Determine if Machines can Think
The interrogator must figure out which of the two people is a man and which is a woman by asking questions and analyzing the answers A and B give to the questions. In addition, A is suppose to trick C and make him or her make a wrong identification, while B tells the truth. The terms for the test are to remain constant so that the interrogator does not rely on things such as voice tone or handwriting to distinguish person A and B. Therefore all answers are typewritten and the interrogator must sit in a separate room so he cannot hear the answers that are given, but only read them.
Turing takes this “imitation game” further in order to evaluate machines and their ability to think. Rather than using two humans as A and B, Turing uses one human and a machine, with the interrogator being a human. The new question then arises, “What will happen if a machine takes the part of A in this game?” The evaluation of this question depends on whether the interrogator makes the wrong guess more often with a machine than he does with a man.
To further look at this situation, it is important to evaluate the machines that are concerned in the game. The machine that is to be used is a “digital computer” because it is intended to carry out human operations more so than any other machine. To compare the make-up of the computer to the human, one can break the digital computer up into three parts. The store, which is used to store information, it is like the paper humans use to do calculations. Second, the executive unit, which carries out the operations used in calculations, and finally, the control, which is used to make sure that instructions that the computer is given are followed correctly and in order. These three parts are used to mimic the actions of humans. The programming of the computer is the instruction table and it is described as the rules in which the computer must follow. These machines are described as the discrete state machines because they move in sudden jumps to each state; also, an important characteristic is that these machines have finite storage.
There is a great sense of universality of digital computers because there are many machines that could perform these operations. In addition, the technology that exists allows for the “perfect computer” for the game to become more of a reality. Turing mentions that the storage of the machine is incredibly important in its ability to mimic a human, and that the future will allow for more storage, making computers follow the human mind more closely.
Turing targets many opposing views to his theory that computers can think. The first argument mentioned is the “Mathematical Objection.” This argument sates that there is a mathematical limit to the power of the discrete state machine. The second is the “Argument from Consciousness” which believes that no machine could have feelings. Meaning that no machine could feel the pleasure of its successes or the sadness of its failures. This argument could be taken to the extreme and argued that one can not be sure a machine thinks, unless a one is a machine and can feel itself thinking. This extreme view of the argument causes many of the argument’s supporters to support the “imitation game” because the extreme view leaves little conclusion and many do not want to be forced by argument into supporting this view. The “Arguments from various Disabilities” view uses scientific induction to describe the preconceived idea of a machine that humans have and the expectations of the machine that this entails.
This argues a machine cannot have much diversity where humans do, at this point in time, there is not enough storage capacity to characterize the diversity of thoughts, and beliefs have into the computer. Another argument is the “Lady Lovelace’s Objection” which states that machines cannot go beyond human intellect because they can only know what humans can program them to know. The “Informality of Behavior” arguments believes that humans have different rules of conduct and machines have one, leading one to conclude that machines cannot mimic the many diversities of human belief and thought. The final opposing argument mentioned is “The Argument from Extra -Sensory Perception” which according to Turing is a strong argument because there is strong statistical evidence supporting telepathy.
Turing admits that he has few convincing arguments to support his views. He attempts to use analogies for support, but these do little for the argument other than state that the “real mind” humans have, which differs them from machines is nonexistent. Turing’s main argument is that an increase in technology will increase the storage of machines and better support his idea that, “machines can think”. Machines can store more information that will allow them to mimic humans more closely.
Turing begins to evaluate human beings when he mentions, “Learning Machines” He states that there are three characteristics in the adult human mind, the initial state of mind, the education to which one had been subjected, and other experiences outside of education. These characteristics help better explain the learning machine and the ideal human to use if evaluating a machine on the same level. He believes that rather than simulating an adult; it is best to use child because a child has a pure initial state of mind with little experiences to reshape his or her conclusions on things, comparable to a computer. In addition, if one were subjected to the education program of a child, one would have a better chance of achieving a better brain as an adult. To better solve the problem, Turing divides it into two parts, consisting of the education process and the child-program. He then analyzes the learning process by stating that it is important to acknowledge that the teacher knows little of what goes on inside of the pupil’s brain, but can still predict the pupil’s behavior to a certain extent. Also, that the behavior of an individual should be characterized by only a little change in computation from the disciplined one of computation. The more predictable a person’s behavior is the easier it is for a computer to imitate it and appear to be able to think.
Finally, it seems that Turing puts a lot of emphasis on the future and what it could offer in the ways of machines and their ability to improve on the basis of mimicking man. He gives little support of his own views and concentrates heavily on those that oppose his own. Rather than depending so much on an unknown, the future, Turing should have given more information and support of his view in order for the reader to better evaluate and understand his point.
