Hi everyone. This is Prof. Yongjin Yoon from KAIST. From now, we're going to study about Math for AI Beginner Part 1 which is about linear algebra first. This is our week first and first session. Let me introduce myself first. I'm Yongjin Yoon in KAIST and I'm the Professor in Mechanical Engineering. Also I'm very interested in the incubating international or startup such as some matching platform and some tech-based entrepreneurship based startup. Also, I'm doing a lot of some consulting about some AI-based startup together with tech-based startup activity. From this course, we are going to study about what is AI, which is a very popular topic in these days and how we learn and we're getting close to the AI, and then, to understand AI, we try to learn the basic mathematics such as linear algebra and the second part would be the vector calculus. In this course, we are going to study about the linear algebra to understand basic artificial intelligence. What is artificial intelligence and math for that? To understand this course, we need to study or understand about the basic of the artificial intelligence. As you know, the artificial intelligence is emerging technology and I assume that you already hear a lot of AI from the media and also a lot of some startups, however, many people is afraid of the AI because of that they cannot understand what is AI exactly. It's very like their strange phenomena because we already very get used to the use of AI in our daily life such as, for example, I see that you already are guided by some GPS and a lot of some mapping system when you drive, and also, you sometimes use a voice recognition device such as Hey Google or some other like service. However, many of us, we really don't know what's the inside of the algorithm for the AI and what's the potential threat from the AI. The major reason why we feel some strange between when we just use AI and when we are trying to understand AI because many people still think AI is only for technology, only for the expert and to understand AI, we need to understand very difficult topic and difficult technology so that we can understand AI. That's why we're afraid of AI, so I think, it's very valuable to learn about what is AI and what's the inside of AI algorithm. Actually, there is a very simple algorithm inside of AI. There is four main question when I encounter with our students and normal people who doesn't know about any AI background. Those are four major occasions. First is what is AI? People still don't know what is AI. They are very getting used to or use for AI, but they don't know about AI. Question 2 is why AI comes out recently. Those are also very big question and the third question is show that how this artificial intelligence is connected with the Fourth Industrial Revolution and in the end, the other question is when I say, if you know the addition and multiplication and divide, then that will be enough for learning the AI. Then everybody has a question about the. Why we need to learn about this paging mathematics to understand AI? Those are four main questions. Let me explain those four questions before we really study what is the linear algebra meaning for the AI. First question, what is AI? May I ask you what is AI? What is artificial intelligence? For example, today morning, I also talked with AI speaker, say, "Hey, Google. Can you play some music for me in the morning." Like this, AI get close to our daily life. The other example, the AI, is maybe you heard about the Lee se-dol and then the play Go match with Lee se-dol with Alpago. That was very shocked to the everybody because most of the people expect that the master of Go, who is a Korean Go master, Lee se-dol would be the winner of that game. However, the result is that the most of the game the Lee se-dol is failed to win the Alpago. At that time, many people really feel the threaten of the AI. How come AI is so smart so that they very easily win over the Lee se-dol. In line with this, many people talk about the singularity, the era of singularity. Singularity means that, the certain point, the machine and computer, their smartness, exceed over the human smartness. The machine learning or the artificial intelligence will rule over the human being. Then they take all our current job so that everybody, including me, may have threaten or may lose their job in the future. In the last image, you can see in the movie a lot, it's Terminator. The robot features of AI, they really destroy human society and human civilization. Is this AI really threaten to us and really dangerous technology for us? Is it really? Still there is a lot of discussion, what would be our future living together with AI? There's a lot of some opinion comes out. We should stop to develop AI or we just learn, use AI for our convenience. This is just open discussion topic. In our class, let us think how the AI affect our daily life. Before that, those strategies and those urgency of this AI comes out. I think comes out because we don't understand about AI. Once you understand AI, we can have more clear opinion and you can have more benefit maybe from the AI. That's what I believe. Let's understand AI together with me. What is AI? From the Wikipedia, you can search that the AI. Artificial intelligence is an intelligence demonstrate by the machines, which is computers, and often used to describe machines or computer that mimic cognitive function that human associate with the human mind. Here, the keyword is mimic human's cognitive functions, such as learning and problem-solving. Those are the definition of AI. However, I think this definition you serve already very difficult to understand. I summarize in a simple way what is AI. AI is the machine or computer which mimic human cognitive function. Those are the simple definition of AI. AI is a machine or computer which mimic human cognitive function. Second is, or AI is a technology for solving problems and achieving goals which is similar to the human cognitive. Here, as you see, there's two keywords, there's computer and technology and human cognitive. Those are the main concept of AI. Don't need to be very confused about the Wikipedia definition. You just remember that those, human cognitive, mimic, and computer, those are three keywords when you try to understand AI. Followed by my definition about AI, I highlighted important part, mimic human cognitive function, and similar to the human cognitive function. Let me introduce some example for us to understand easily about AI. Let me bring out the example about AI speaker. There are three major part when you talk about AI. First part is cognition. Cognition is like sight, hearing, or touching, smelling, or tasting. Those are our five senses. Those external stimulus can be obtained from the sensors. Then based on that data from, we call this IoT, Internet of Things data, and then with that data, well, the computer analyze to understand and learning and analyze what is the meaning of the data from cognition, and then respond to that result. This three-step is very similar to the human brain function. For example, when you ask order to the AI speaker, for example, "Please, play cafe music for this morning," and then AI speaker cognition is hearing. They hear those please play cafe music. Second is, from the AI machine, they understand what is the ordering from the owner, and then respond is okay. The AI speaker say, "Okay, I play number 1 popular cafe music for you right there." I assume you already experienced this AI. Alan Turing is known as a father of AI. He had been working on encryptizing some phrase from the enemy side which change their code in due time. It's very difficult to solve that encryptizing code. So that he bring out some idea, can machine have an intelligence like a human? Then he bring out, how can you define the machine has an intelligence? He bring out the definition for the AI. At that time he didn't call this one as an AI but he say, "If the machine can communicate with human, then machine has an intelligence." That's Alan Turing's definition of AI. When you talk about this machine has artificial intelligence, there is Turing test, it was developed in 1950. Turing test, for example, there is one questioner and the other side, there is a robot or a machine and dear human being. Those two groups cannot see each other. They cover with a bar right here. Then the left side, a human being asks some normal question, "What is your name?," for example. Then the answer from the one side is maybe, "Name is for calling someone." Then the other side, "I am Yongjin Yoon." Like that. Then the left-hand side, right human being can recognize that the person who answered like I am Yongjin Yoon is a human being. Then the other side would be not human. During five-minute conversation, the questioner tried to find who is human and who is machine. Then this Turing test is, for example, there are three questioners, and then ask, among three person at least one person designate the machine as a human, then AI passes the Turing test. This is the basic concept of a Turing test. With the 33 percent of the correctness for confusing, questioner confused the machine as a human, then those machine passed the Turing test. This is the end of the first session. Let me introduce more details about example of a Turing Test, and then the other question in the next session. Thank you very much.
