Good Essay About Other Tests
Over the years, artificial intelligence has evolved and has led to the creation of machines that behave like human beings. The main philosophical question of Artificial Intelligence is whether machines created are similar to human beings. In my opinion, based on various tests by different philosophers, machines are not similar to human beings. Machines can perform some functions better than human beings, but there are some cognitive processes that they cannot exhibit. A machine can only perform tasks programmed by the human brain. It can easily outperform a human being’s brain at any computational activity but when a human brain is calculating a mathematics problem, it is also doing much more than that computation. Machines still remain to be human creations and may simulate human actions as close as possible but at no point will they be equal to human beings.
According to Descartes, human beings are not machines because they have two abilities that machines do not have. The ability to talk intelligently about a variety of topics and the ability to act intelligently in different situations. Machines can only act intelligently in situations they are programmed or designed to handle. Human beings can act intelligently even in situations they have not faced before which makes them more than machines. Some people in the field of cognitive science do not agree with Descartes’ way of assessing artificial intelligence. Most people agree that assessing the ability to use language and solve problems are the best in indicating intelligence. Another test to differentiate the human mind and machine mind exist and some check the ability to think or existence of a conscience (“The Mind-Body Problem” 77).
According to a philosopher named Gottfried Wilhelm von Leibniz, thinking is not a mechanical concept. It is not possible to mechanically explain the ability of a machine to think. Leibniz explained this by suggesting, if there exists a machine that can think, feel and perceive. If it were possible to enlarge the machine in terms of proportions and enter it as a mill, then someone entering it would observe parts moving and pushing each other and not an explanation for the perception. Unlike Descartes, Leibniz believes it is not impossible to create or construct thinking machines. However, if machines can think it is not possible to have mechanical explanations of thinking (“The Mind-Body Problem” 78).
Allan Turing introduced the Turing Test that was a way to compare human beings and machines’ abilities. Turing created a party game known as the imitation game that was meant to show the existence of intelligence in machines. The game involves a man and a woman who would answer questions that the contestant would use to determine who was male and who was female. The game is arranged such that their voices, shape, and sizes would not be revealed. If the contestant correctly identified the genders, he or she would win. The Turing Test can be carried out in many different ways. For example, the machine can be substituted for the male participant. In this case, the ability of the machine to show it is human would be evidence of its intelligence (Fetzer).
Harnard suggested the conception of Total Turing Test. According to him, nonverbal behavior is important as verbal behavior in determining the words used. Objects can be labeled, identified, sorted, and classified based on their sensory projections. The ability to categorize depends on the capacity to isolate the different ways things appear. Total Turing Test was introduced to test non-symbolic and symbolic behavior where symbols are the verbal and nonverbal behavior displayed by a system. Like Descartes, Harnard identified mentality not simply as the ability to think but based on thinking is always conscious. A machine might display verbal and nonverbal behavior similar to a human being, but it will not possess the mentality (Fetzer).
In 2011, scientist in IBM created a computer called Watson to compete in one of the best American games known as jeopardy. In 1997, IBM also developed a supercomputer called Deep Blue that competed with the world chess champion at that time, Garry Kasparov and defeated him. Playing chess requires logical analysis of the possible moves that can be made in the next game. Winning jeopardy, on the other hand, requires understanding of the question posed in natural language and accessing a huge database and answering as fast as possible. The intelligence required in both games is different. Chess is more like a vertical sort of intelligence while in jeopardy is a horizontal thought process. The Artificial Intelligence community is beginning to question whether it should be so focused on recreating human intelligence. The emphasis is now shifting to creating machines with intelligence unique to other machines (Sautoy).
Films on Artificial Intelligence
Blade Runner is a science fiction film set around the year 2020. Deckard, who is a retired law enforcement officer, goes back to service for a special mission. A number of enslaved replicants have fled to another planet and Deckard is supposed to hunt them down and kill them. Deckard meets Racheal, who is a replicant too, he gets to know her and advises her to flee. Eventually, he tracks all the rebel replicants except Racheal, who becomes his lover. The film raised questions on whether it is possible to distinguish real human beings from artificial replicants and how a machine can completely act like a human being (Philosophical Films).
‘Her’ is a futuristic film about a computer that can compose music, carry on conversations seamlessly with people and even fall in love. The film shows the evolution of Samantha operating system from an assistant to an ideal girlfriend and to an entity that loses interest in people because they are unsatisfying companions. The system not only recognizes speech but also understands natural language, reasons, generates speech, plans and learns. According to today’s intelligent systems, the unstructured information on the web is inaccessible. Programs such as Never Ending Language Learner that can read the information on the web and populate a knowledge base indicate promise in this field (Nuance).
Research in artificial intelligence has progressed, and it is now possible for machines to function in ways that are almost similar to a person. If we were to communicate with a machine, would the answers given show genuine personhood. It is possible to differentiate whether machines are thinking and behaving rationally and whether they appear to be doing so. If one can predict the questions that a computer will be asked, the answers can be pre-programmed. Preprogramming the questions is not a guarantee that the computer will understand the questions and answers. It is, for this reason, possible to build a machine that behaves like a person. However, the philosophical question is whether we can make the robot be a person and not just seem like a person (Pruss).
There are various questions on identity that a person can ask themselves. If there was another two-year-old that grew up to be person A, would the person be the same as person A herself? Is Queen Victoria the person the same as the corpse? If a mother gives birth to Siamese twins, is that one person or two? If someone’s mind was completely erased and the resulting amnesiac tortured, who would be feeling the pain? Some people claim the persistence of the body or brain shows a person’s identity. Some claim memories guarantee identity. Others claim the presence of something beyond the body guarantees a person’s identity (Pruss).
Weak and Strong AI
Artificial Intelligence is the concept that a machine can think. There are two types of Artificial Intelligence namely strong and weak. Weak AI claims that a computer can simulate a cognitive process, but it is not a cognitive process itself. Just like a computer simulation of a hurricane is not the hurricane itself. Strong AI asserts that a computer can get programmed to be a mind to understand, perceive and exhibit cognitive states similar to human beings (University of Texas at Austin 2). In my opinion, there can only exist AI. It is possible for the best developers to create a machine that is almost similar to a human being, but the machine will never be equal to human beings in terms of the cognitive process it exhibits.
My argument against strong AI is based on the Searle’s Chinese room theory. Suppose an English speaking human being who knows Chinese was put inside a room to simulate the execution of a computer program operating in Chinese characters that he or she does not understand. Suppose the program being executed is an artificial intelligence program receiving natural language questions in Chinese and responding correctly to Chinese sentences. Generation of natural language responses makes the claim more reasonable, they are indistinguishable from those a Chinese speaker would generate. There is no ‘understanding’ of the symbols only meaningless symbols are getting generated (University of Texas at Austin 2).
Fetzer, James H. “Minds and Machines: Behaviorism, Dualism, and Beyond.” Constructions of the mind SEHR 4.2 (1995). Web. 27 Jan 2015. Retrieved from http://web.stanford.edu/group/SHR/4-2/text/fetzer.html
Nuance, Vlad S. “Can we Build ‘Her’?: What Samantha Tells us About the Future of AI.” Wired 19 February 2014. Web. 27 Jan 2015. Retrieved from http://www.wired.com/2014/02/can-build-samantha-tells-us-future-ai/
Philosophical Films. Blade Runner (1982), 2015. Web. 27 Jan 2015. Retrieved from http://www.philfilms.utm.edu/1/blade.htm
Pruss, Alexander R. Artificial Intelligence and Personal Identity, 2015. Web. 27 Jan 2015. Retrieved from https://bearspace.baylor.edu/Alexander_Pruss/www/papers/AIAndIdentity.html
Sautoy, Marcus. “AI robot: How Machine Intelligence is Evolving.”. The Guardian 1 Aprill 2012. Web. 27 Jan 2015. Retrieved from http://www.theguardian.com/technology/2012/apr/01/ai-artificial-intelligence-robots-sautoy
The Mind-Body Problem. Capter 2. Web. 27 Jan 2015. Retrieved from http://highered.mheducation.com/sites/dl/free/0073386685/741316/ch02.pdf