In this blog post, we will look at whether the day will come when artificial intelligence can think and converse naturally like humans.
From “Star Wars” to “Transformers,” many science fiction films feature robots as a major theme. In fact, many robots have been developed to make our lives more convenient by helping us clean or enabling sophisticated surgery. However, robots in movies are very different from robots in real life. For example, in “Transformers,” there are frequent scenes in which the main character and the robot talk to each other, and most robot movies include scenes in which the robot talks to the human. To have a conversation, you must be able to understand what the other person is saying and respond accordingly.
A: “Have you eaten?”
B: “I’m on my way to school.”
In the above example, “A” and “B” are talking to each other, but they are not actually having a conversation. A conversation is not simply a back-and-forth of words, but a process of properly conveying one’s thoughts to the other person. The reason why there are so many scenes like this in movies is that they reflect the dream of humans to be able to talk to robots. The academic discipline needed to create a “thinking machine” that can communicate with humans is artificial intelligence.
Most people have heard of the term artificial intelligence, but not many people know its exact meaning. Artificial intelligence refers to intelligence created by humans or intelligent beings, that is, artificial intelligence. Artificial intelligence is broadly divided into strong artificial intelligence and weak artificial intelligence. Both are computer-based artificial intelligence, but strong artificial intelligence can actually think and solve problems, while weak artificial intelligence cannot. In this article, I will explain the strong artificial intelligence needed to create robots that can talk to us.
In discussions about strong artificial intelligence, philosophers have debated from a philosophical perspective whether it is possible to implement human intelligence or consciousness in machines. The Turing Test is a representative experiment that supports the possibility of strong artificial intelligence, and the Chinese Room Experiment is a counterargument to it.
The Turing test was designed to answer the question “Can a machine think?” and is based on the claim that “if a computer’s response cannot be distinguished from a human’s, then the computer can think.” In other words, if a machine’s response is indistinguishable from a human’s, then the machine should be considered intelligent.
On the other hand, the “Chinese Room” experiment shows that a computer’s response does not necessarily indicate that it has gone through a thought process. To put it simply, a person who does not know Chinese is put in a room, and the room is equipped with a writing tool and a list of pre-written Chinese questions and answers. When someone writes a question in Chinese outside the room, the person inside the room writes the answer according to the prepared list and sends it outside. The person outside may feel that they have communicated with the person inside the room, but in fact, the person inside the room has just written down the prepared list without thinking. This leads to the conclusion that even if a computer can answer a question, it cannot be concluded that the answer is the result of intelligent thought. In other words, it is difficult to determine whether a machine can actually think using the Turing test alone. The debate over the possibility of machines thinking has been active for a long time.
To discuss whether machines can think, we must first understand the meaning of the question. Machine learning is the technology that makes this possible. Machine learning is literally giving computers the ability to learn and perform actions that are not specified in code. Through learning, humans can expand their knowledge, understand new situations, and adapt to change. Machine learning is an attempt to implement human learning ability in computers. In order for a computer to understand a sentence, it first needs a huge amount of data. A large number of data containing appropriate answers to each question are stored in a database, and statistical rules or patterns are found in large-scale data using big data and data mining techniques. Through this process, a computer can answer a given question.
It is possible for a computer to have a conversation with a human based on pre-prepared data in this way. However, if you ask whether a computer can “think,” opinions may differ depending on the definition, but it is likely that it is not possible at this time. This is because it is difficult to process data that does not exist. Babies sometimes use the expression “it’s there” by combining “it’s gone” and “it’s here,” and this creative expression is an area that computers cannot do. Of course, with current technology, it is capable of beating the world champion of chess or winning a quiz show. However, chess plays by storing all possible cases in a database and using them, and a quiz show is also conducted by analyzing questions and searching for relevant information in the database to find the answer. In other words, it is not that machines think like humans, but that they find answers based on vast amounts of data. However, just as the days when flip phones were novel have passed and smartphones have become commonplace in just a few years, we expect that the day will come when we will be able to talk naturally with robots.
