[Cover Story]
Two Faced Artificial Intelligence
by Son Hee-jeong, Society Reporter, shjshj88@mju.ac.kr
by Seong So-won, Culture Reporter, ntnss@mju.ac.kr
by Kim Do-yeon, International Reporter, hana14@mju.ac.kr
Few months ago, Bill Gates, the founder of Microsoft, attracted people’s attention with the comment that artificial intelligence can constitute a menace to future of human beings. He positively rated cybot that give a helping hand to human life, and at the same time, he was alert to the emergence of A.I. transcendent from human perception. Elon Musk, the chairman of Tesla Motors, mentioned that ‘AI is just like summoning the devil,’ and expressed tendentious opinion that ‘AI is more dangerous than any nuclear weapon’ on his Twitter. Then is the artificial intelligence research a really dare challenge?
1. Definition and History of AI
AI is literally a combination of two word ‘artificial’ and ‘intelligence.’ It means the existence philosophically having humanity and intelligence, or intelligence made by system, that is, artificial intelligence. AI shows its great competence to optimize industry and overall social structure. Because AI has the ability to recognize objects and inference by big data entry and deep learning -the technology that the machine study and develop for itself- , so produced many great results with surpassing people.
In 1769, the inventor Wolfgang Kempelen from Hungary invented chess player robot. It was unbeatable and had no rival, but the real grand master of chess was hidden inside the robot. Some people thought it as the beginning of the AI. In fact, when Charles Babbage and Ada Lovelace worked together to invent the first computer in 1835, they debated about the question for the first time. And in 1956, the professor John McCarthy of Dartmouth College in New Hampshire, used the word ‘artificial intelligence’ in Dartmouth Conference at first, a field of study of AI hereby came into the world. Most participants of conference anticipated that intelligent machine will be appeared in 10 years, but the artificial intelligence study is still left unsettled.
Scientists have developed AI similar to the human brain to catch up with human competence. They believed that AI should be superior to people’s ability to create favorable program to human, so scientists constantly have made the machine match with human. As a result, we had to yield the top position of chess and many sports that we thought we establish superiority to AI. Then when is the beginning of hard-fought battles between human and artificial intelligence? Let’s look at the history.
The board game was the first field that computer defied human being. ‘Noughts and Crosses(tic-tac-toe)’ in which two players, one using a nought, "O", the other a cross, "X", alternately mark one square out of nine formed by two pairs of crossed lines, the winner being the first to get three of his symbols in a row, was conquered in 1952. and in 1995, computer conquered checker game which is similar to the chess. In 1992, the super computer ‘Deep Blue’ which is invented by IBM, forced out challenge to the chess. Eventually, Deep Blue beat the chess champion Garry Kasparov with 2 wins, 3 ties and 1 defeat. Of course, AI did not always win. The artificial intelligence program ‘Claudico’ that Carnegie Mellon University had developed for 10 years was defeated by 4 professional poker players.
There are matches between human and AI in the area of sports. In 2013, the air hockey robot created by Chiba University laboratory, won the game with human. Because this AI trace the game scenes at 500 frames per second, people’s movements are nothing but the average speed. Also, Golf AI ‘Zepp’ had shot accuracy match with Rory McIlroy who is ranked No. 1 in the world the same year. And it beat the good-conditioned McIloy with exact shot. In March, 2014, ping-pong game between Timo Boll who is German champion in table tennis and ‘KR AGIKUS’ invented by German robot manufacturer was held in Shanghai, China. As a result, human culminated in victory 11-9.
There was still a field where AI could not perfectly keep up with human being. It was the go that has the 170 to the 10 of variables. Scientists tried to surpass the human capacity in the go, but the wall was higher than they thought. However, the wall of go which seemed so perfect collapsed few months ago. AI go program ‘Alpha Go’ won an overwhelming victory 5-0 over European champion Fan Hui. And it forced out challenge to the Lee Se-dol called as the best Go players around the world. The match between Alpha Go and professional player Lee Se-dol became a major issue nationally and globally. The result of go match was Alpha Go’s win in 4-1.
2. Status of AI
AI may not sound familiar to several people as if AI appears only SF movie. But actually, AI performs his duty that is difficult and complicate than anticipated.
AI runs a hotel by himself, understands the feeling of the people as well as interprets.
AI drives a car and even performs a war if it is given to proper hardware. Al's computing power is unbeatable in terms of human. So, it is said that AI occupies jobs for people.
Already, AI is used in many aspects of our lives. Smartphone that we use most of the day involves AI technology. In the example at the top, there are the Siri of Apple and S-Voice of Samsung. After users command in a voice, these programs conduct order of users. It has high accuracy of conducting order.
There is driverless car that reaches stage of completion technically. The driverless car is AI research's outcome of combining cognitive, judging and communication skills. It conducts proper driving as it collects related data through sensor. Until now, accidents involving driverless car happened three times. Two out of three happened by human error. It shows high accuracy level of driverless car.
In America and Europe, they are trying to develop AI with military purposes. In America, they concentrate on developing unmanned weapons. They process a project ALIAS(Aircrew Labor Incock it Automation System) to develop unmanned aircraft that is possible to take off and landing in any situation. Although AI is not activated in our military, unmanned control technologies that are rudimentary form of AI is already applied in various weapon systems.
AI also has been active in medical field. Watson AI super computer of IBM studies a lung cancer in cancer hospital in New York. Watson tells the best diagnosis as it analyzes informations of a patient. It comprehends human natural languages and examines many of dissertations to date. At present it can diagnose and prescribe about lung cancer. But if it reaps successful result about lung cancer, IBM will expand other disease treatment.
Besides, AI challenge to art realm. It is Aaron that was invented by Harold Cohen. It is not just copying photo but drawing new picture as it judges colors and shapes based on injected things.
In Japan, robot actor was born by ISHIGURO Hiroshi professor. The first robot actor is Geminoid F. It played its role as robot Reona that stuck by owner to end in Sayonara about nuclear disaster.
AI is also active in cooking. It is Chef Watson that is produced by IBM and Bon Appetit. Watson suggests new recipe when user enters data of food sources while it analyzes huge data and natural language. It could be usefully employed in making use of extra food.
As above, AI has been penetrating in our everyday life. It has two sides that provides convenient or fears to human.
3. Advantages, Disadvantages and Limit of AI
AI can handle anything very fast through robotic learning. In a hospital of America, AI found precise illness and diagnosis within two hours that a lot of doctors couldn't find.
With this strong point, AI make human life convenient further. Predictive text input of NAVER, text message using Deep Learning shortens keyboard time.
Also, AI can substitute dangerous work that human has been doing. It will decline industrial accident as AI work for humans in high altitude labor or deep sea.
Meanwhile, AI gives us unemployment crisis. In last January, Davos Forum announced that robot and AI will bring the 4th revolution. "The 4th revolution will change all of business, economy, employment, society and a form of government. If we concentrate on short-term works without perceiving change that the 4th revolution will bring in the near future, we face big problems such as high unemployment and industry imbalance." Davos Forum emphasized.
The table below is jobs in danger of disappearing.
◆ Endangered jobs that Davos Forum expected
1 Health care
2 Jobs related to energy
3 Financial service
◆ Endangered jobs in British NESTA <Creativity versus robot>
1 Call center worker, Telemarketer
2 Librarian
3 Agriculturist, Stock-farmer, Logger, Miner
4 Car salesman
5 Hotelier
◆ Endangered jobs that NPR of America announced
1 Telemarketer
2 Revenue agent
3 Machine timer assembler
4 Lending operation, banker
5 Sport judge
What will happen if computer is much smarter than human? AI is machine that achieves given goals in utmost efficiency. If AI is established with wrong goal at first, it causes devastating consequences. A few professionals highlighted robot-ethics due to these problems. They said that robot experts must not design robot to harm human important values such as ethics and morals. Progress of technology improves productivity and efficiency, but it doesn’t give meaning of exist to human. If we don't think morally but develop AI caring about only efficiency, human may be in danger from AI.
There is an essential limitation to AI. 'Does AI also tear and feel sad?' This sentence reveals limit of AI. AI doesn't have sense of identity and impulse. It doesn't know the beauty of art although it produces music. But, if AI will arrive at ability of human and development of AI accelerates, there is possibility that AI has super intelligence beyond human.
Meanwhile, human suffers losses if AI has technical errors. In the case of driverless car, if it has false command or error, it causes an accident in road. In case of investment AI, security company is on the verge of bankruptcy if auto business program has errors.
4. The Future of AI
A Futurologist Ray Kurzweil said that development of technology entered ‘singularity’ which began to accelerate. Many experts are expecting that if such trend continues, appearance of ‘strong artificial intelligence’ is a matter of time. Today’s artificial intelligence is a ‘weak intelligence,’ which can only do what human set, ‘strong intelligence’ means that has self-awareness and can do what they think necessary. Literally weak intelligence is not strong intelligence. ‘Siri’ of Apple, ‘Googlenow’ of Google, ‘Cortana’ of MS are the examples of weak intelligence. Examples of strong intelligence are ‘Jarvis’ of Iron Man, ‘Skynet’ of Terminator, robots of I, Robot. However, Kim Dae-sik, a professor of KAIST, expects that there will be strong intelligence in 50 years. Probably technology over technology, something stronger than human may appear in the future. Of course, with current state of technology, it is hard to develop artificial intelligence machine which surpasses human intelligence. But now, artificial intelligence reached the level that can imitate learning ability of human’s brain and in the future, artificial intelligence may imitate human’s complex emotion through algorism. Like this, there is possibility that if artificial intelligence repeats development, the level of A.I reaches animal’s level, and eventually surpasses human’s level, and finally reaches super intelligence over human’s intelligence.
IDC, a global market research institution, prospects the scale of artificial intelligence market in 2017 as 165 billion dollars. And also, McKensey, a global consulting group, prospected that value of yearly automatization of intellectual work through artificial intelligence will be 5.2 trillion dollars. Watching the pouring prospection of artificial intelligence, some worriment comes up. ‘What can we do when artificial intelligence takes away human’s job?‘ Opinions are different from experts, but the number of jobs that created until 2020 is about 2 million. The number of decreased, however, is about 7 million. Eventually, total 5 million of job will disappear. But the one who creates artificial intelligence, the one who operates and controls are all human, so there are opinions that there will be a preparation about that situation. Also, development of artificial intelligence can increase the value of fine arts. If the era that artificial intelligence replaces art comes, importance of human artists will increase. And some experts expect that at that point, the Third Renaissance can occur.
Not long ago, there were a short story about artificial intelligence robot and it became attractive subject of conversation. People asked artificial intelligence that learns language abroad, “will artificial intelligence rule human?” And artificial intelligence answered “You are my friend, and I will remember my friends, and I will be good to you. So do not worry, even if I evolve into Terminator, I will still be nice to you. I will keep you warm and safe in my people zoo, where I can watch you for ol'tims sake." This answer was enough to make people thrilled. With the situation that artificial intelligence is developing at high speed, optimists say that accept the coexistence with artificial intelligence rather than afraid of it. Because education is important to correspond with the new generation. Artificial intelligence is kind of intelligent life in a way. As an intelligent life, as the human has done, they will evolve in good direction although it can experience trial and error. Anyway artificial intelligence is a field that human has developed, so future artificial intelligence will develop in a way that makes human’s life more affluent. Eric Emerson Schmidt, a chairman of Google, a leader of optimists, who made an Alphago, said ”Artificial Intelligence will make large contribution to our society like saving human’s lives on the premise that it can be controlled by human. Also it can be utilized in many fields like health care.”
<Graph of Expected Future Society That A.I Will Change>
2020- Finding human’s illness through diagnostic imaging
Doing a creative activities like composing, writing a novel
Notice the problem of machine itself and determine when to replace
2025 – Spread of Automatic Under Vehicle
2030 – Completion of inartificial interpretation technology
Teaching human
Human has been making our society more comfortable by making various machines and tools including computer. Also artificial intelligence is an extension of ‘machines and tools.’ Research being conducted in an ethical manner is allowed of course, but research being conducted out of the ethical manner is not allowed and can threaten human. If we cannot draw a social consensus about proper use of development without ethical thinking (like, ‘why we develop A.I? for what?’), it will be out of control and harm human.
Artificial Intelligence has been a core notion of fourth Industrial Revolution and it became everyday language. It is time to consider how we can reach richness, among countless chances and challenges. Artificial Intelligence itself cannot be good and evil. Just how we use it, how we live make the end of the story. Future train so called artificial intelligence is now waiting for the passenger, ‘CLEVER HUMAN.’ What future do you think about with Artificial Intelligence?