Can a machine know?

In certain cases a machine can know. Indeed, if we choose specific aspects of the meaning of “to know”, a machine can know things through different methods. Like a human being, it can reason, solve problems (mathematical equations for example, calculations using specific programmes such as Excel…). A machine is good with logical problems. It can also translate words and sentences because it “knows” the translation of the words. A machine knows what a human put into its electronic system. But only specific machines can do that. Indeed, coffee machine could not translate a word! Computers, calculators, translators are capable of solving problems or, in the case of the latter, translating a text. Computers are the machines that are most likely to be considered as “knower machines”. When we agree that a machine can know, we have to talk about artificial intelligence which is an important contemporary science. It is a multidisciplinary field encompassing computer science, neuroscience, philosophy, psychology, robotics and linguistics, and devoted to the reproduction of the methods or results of human reasoning and brain activity. There is the conventional artificial intelligence which is a branch of artificial intelligence mainly dealing with symbolic problems and the computational artificial intelligence which its research aims to use learning, adaptive, or evolutionary computation to create programs that are, in some sense, intelligent. Computational intelligence research either explicitly rejects statistics or tacitly ignores them. So those machines are imitating the human brain. Some of them know more that certain persons, but only because humans gave them information.

Humans have access to knowledge through language. Indeed, this is a way of communication. However, machines do not talk or interact with each other as humans do. As Hans Reichenbach said: “If you can’t say it, you don’t know it”. Computers can communicate between each others through the internet but other machines such as cars, vacuum cleaners, or televisions cannot. To illustrate the language as a way of knowing, the Chinese room experiment is very relevant. It was done by John Searle in 1980. In this experiment, Searle asks us to suppose that he is sitting inside a computer which is in a small locked room. In other words, he is in a small room in which he receives Chinese characters, consults a rule book, and returns the Chinese characters that the rules dictate. Searle notes that he does not, of course, understand a word of Chinese. Furthermore, he argues that his lack of understanding goes to show that computers do not understand Chinese either, because they are in the same situation as he is. They are mindless “manipulators of symbols”, just as he is, and they do not understand what they are 'saying', just as he does not.

There is no point in knowing something if you cannot understand it and then use it. That was what Goethe meant when he said: “Knowing is not enough, we must apply.”

We know that machines (apart from computers, cars and specific machines…) have only one or two functions programmed by humans: for example a coffee machine makes coffee, a camera takes pictures. Being programmed is not the same thing as knowing. A machine cannot break into the abstract from its primary programming: it must serve its function and nothing more. For a machine to be able to know, something would mean it is aware and conscious, and for something to be aware and conscious on that level, then it is not a machine anymore, and it has become a life form with free will and free imagination. As Rabelais said: “Science without conscience is only ruin of the soul”. It does not have the ability to change things it learnt to make it better. It can only restore what has been given to it. Plato’s definition of knowledge was “a justified true belief”. A belief means nothing for a machine because it cannot think; it cannot have a judgment or an opinion about anything.

A human being cannot know something if he does not believe it.  Indeed, he cannot know a fact if he does not believe that the fact is true. But a machine is not capable of believing because it doesn’t have faith in anything. It differs really from human knowledge which has a faith in something, like religion for example. Moreover, machines have no concept of “what is right” and “what is wrong”. Ethics are not a part of the machines’ system. When we answer the question “Can a machine know?” we have to talk about Turing’s experiment that proved that a machine cannot fool a human even if Turing predicted that it would be able to fool 30% of human judges during a five minutes’ test. (It was a written conversation between a human and a computer). It had a little success but not as much as Turing’s prediction. Humans have different ways of knowing which a machine cannot share with them. Let’s take the example of a car accident. You crash the car into a tree: the car will be broken but will not cry or feel any pain: indeed, a machine does not have feelings or emotions which are very important for a human to know. Also a machine has neither perception nor intuition. These elements are what constitute the soul. So we can say that a machine does not have a soul.

Personally, I do not think that a machine can know because of its lack of soul, mind, perception, emotions, intuition and feelings. It is just a human tool and it obeys to its master. There could be some exceptions like the computers that can sometimes “imitate” human behaviour. By that, I mean that they can do things that humans are used to doing such as translations, calculations… Otherwise they are just tools. However, it maybe possible that, in the future, humans will be the tools of machines…

Bibliography:

∙        25 Key words of Philosophy, Hervé Boilot, Succès du livre, Sartre (France) March 2004

∙        Theory of knowledge, Nicholas Alchin, IB Diploma programme( second edition) , Hodder Murray, London, 2006

∙        

 

---