We all know that what makes humans entirely different from other organisms is our rational being, which means we can reason and think and not just work on with our instincts like other animals do. If we aim to really be able to make human-like androids, this should be the very first thing we have to make a study on. Can machines really think like humans do? This question has been one of the most intriguing ones that philosophers and computer scientists alike are trying to answer.

A program named ELIZA written by Joseph Weizenbaum in the mid-90s asserts to emulate the responses of a psychiatrist during consultation by a patient. With ELIZA’s responses, one may be convinced that machines do think. However, there is still no certainty to say so because as aforementioned above ELIZA satisfies the Turing test in a limited domain.

According to Lister, examination of the program shows that ELIZA’s responses are only sophisticated echoes of the input from the patient. A script containing a number of templates is the basis of the program used for ELIZA. Each template would consist of the constant statement, and a variable which is replaced by the statements given by the patient. An example would be a template containing ‘What makes you think xxx’, where xxx represents part of the sentence which is to be completed by material from the preceding input. First the input statement is inspected for presence of keywords such as I, you, mother, hate, which are chosen for their probability of occurrence in a patient’s remarks. There is then a set of rules associated with each keyword for transforming a sentence containing the keyword into material which can be used in response, like ‘you are’ into ‘I am’. ELIZA does not really understand the conversation it is engaged in because it only runs and follows the program built on it. Thus ELIZA does not necessarily ‘think’.

Without doubt, answers to this question are still not plausible. The controversy is still going on, and answers are still very hard to clear up. However, it is to be believed that machines will be able to do more than what we think they can. That we will not miss.