ETRI's artificial intelligence expert, Kim Ki-hoon alumni

As always, the year has passed and I have come to a time to look back on this year and vow the next year. If you promise next year, you will naturally worry about the far future. However, things will not be easily solved just by worrying. So in this article, I would like to give you a special opportunity to overcome career anxiety. With the increasing interest and demand for artificial intelligence(ai) around the world, we met your alumni Kim ki-hoon who is actually working in ai.

 

1. Nice too meet you. I appreciate your time. Please introduce yourself briefly.

Hello, I am Kim ki-hoon from ETRI. I am in charge of artificial intelligence in KSB fusion research group. I am an alumnus of yours, and have received master's degree and finished Doctor's course in KAIST. I am now conducting research on the development of artificial intelligence common framework and edge computing in ETRI.

 

2. It seems that there are lots of EE students who understand ETRI as a place only for the research. Please explain what ETRI is for students. How did you choose your current carrer path?

When you graduate from KAIST, you will usually go to a national research institute, industry employee or university professor. After graduation, I first worked as a researcher in the telecom industry called LG Dacom to develop a VOIP system. Then I moved to ETRI after then. ETRI is a national research institute supported by the government. It aims to contribute to national economic and social development by predicting and researching future science and technology and ICT (information and communication technology) which are essential for the country. Based on this philosophy, the most important characteristic of ETRI is to study the source technology that benefits the entire nation's industry. The difference with the general enterprise research institutes is that while they mainly focus on the short-term research which is profitable for the company, ETRI conducts the mid to long-term research for 5 ~ 10 years rather than profitability. The goal I grew up as a researcher was to help people live a  convenient and happy life. I have been studying science and technology and ICT that naturally seek the convenience of people, and I wanted to make people happy with the ideas I conceived. I thought that my goal fits well with the characteristics of ETRI mentioned above, and I chose ETRI to do research that brings convenience to people in the mid and long-term, not with a short-term profit.

 

3. What are some of the difficulties you have experienced while working at the ETRI? Who are suitable for being a research? 

In ETRI, it is usual to go home late or lack sleep. Since you are sitting on the desk for a long time, it is important to do exercise. I do not think there is anything required for physically to do as a researcher, so I think anyone who is curious about new things can consider it. Especially if you like to use new products and like to touch new machines, I strongly recommend them. If you want to change the world like Bill Gates and Steve Jobs, it's good to start a research at institute like ETRI.

 

4. What moment do you feel most rewarding when you work at ETRI?

The rewards I feel can be divided into two parts: company and individual. In terms of the company, I feel rewarded for developing technology such as memory semiconductors and LTE so that Korean companies such as Samsung Electronics and LG Electronics become the best in the world for 35 years. On the personal side, I felt very rewarding when I was recognized as a national representative of Korea's information and communication field.

 

5. I'm impressed that you told me that ETRI is conducting research on artificial intelligence. Recently, artificial intelligence has attracted a great deal of attention from people. When did artificial intelligence begin?

In 1956, the term "artificial intelligence" (AI) began to be used for the first time through the Dartmouth conference held by Professor John McCarthy at the University of Dartmouth in the United States. At the time, the core of artificial intelligence research was inference and navigation. Research in artificial intelligence that can solve problems and think like humans has been active until the 1970s. However, Artificial intelligence has been frustrated by the challenge of solving more complex problems beyond simple problems. In other words, the artificial intelligence research, which was once active, was followed by a rapid glacial period. However, with recent advances in hardware, we have achieved outstanding performance, and research on artificial intelligence has been revitalized, resulting in the present large market.

 

6. If artificial intelligence is used in real life in the future, what are you expecting to be used in?

First of all, artificial intelligence can be used in medicine. Since you can operate correctly without tremors, artificial intelligence will be more accurate than a person when operating. In the diagnosis of disease, artificial intelligence can make an accurate judgment based on a lot of medical evidence and data. Also, it will be used for autonomous driving. It can utilize the image recognition function to judge what traffic situation is like as if the person is looking at it. This will enable us to respond quickly and accurately in hazardous situations. Finally, it will be widely used in search engines. By using AI in search engines, you can identify and learn about the characteristics of the site through keywords used in your website. This will allow you to optimize the Internet by separating low-quality sites or harmful sites, and users will be able to access the Internet more comfortably.

 

7. Recently, we have seen that research on artificial intelligence is becoming more active.  What is the current position of artificial intelligence?

Do you remember last year's GO competition between human and computer? Surprisingly, the Alpha-GO beat the human best at  4: 1. This year, the Alpha Go even won the world's top three at 3: 0. Go is a brain game, where the number of cases so immeasurable that it was considered impossible to be conducted with computer programming. That's why we focus our attention on Alpha-Go's victory. In recent years, there have been lots of attempts to apply artificial intelligence to fields other than GO that humans have deemed impossible. The possibilities of artificial intelligence are endless. Speech recognition, image recognition and translation have already surpassed human abilities. Future development is expected more.

 

8. According to the trend, interest in artificial intelligence has increased greatly in electronics and there are many students who dream of future artificial intelligence experts. Please advise these students as seniors.

In order to become an AI expert, you must first study hard. Especially you should be good at math. It is also good to expand your insights through reading or do a lot of activities. Usually AI research can only be conducted in computer engineering department, but EE department can also consider. Nowadays, the field of artificial intelligence utilizes hardware as well, so if you take advantage of EE's in these areas and try to work with what I have mentioned above, you will be a good artificial intelligence expert. Professor Shin Jin Woo, Professor Kim Jun Mo and Professor Kyung Soo Park are leading professors who study artificial intelligence in KAIST. 

One more tip is that you have to spend a lot of time studying software. If you do not know which of many programming languages you should study, I would recommend to study Python rather than C language in artificial intelligence. In order to process large amount of data, we need SPARK program.

Through the interview, I am afraid that I gave advice focusing only on the artificial intelligence field. I hope you to do your best in the fields that you want!

Show List
메뉴닫기