Nowadays, AI is developing quickly. It is no exaggeration that the AI is already penetrated into our lives and has become an inseparable part of service (Huang, 2017). In March 2016, AlphaGo, an AI program created by Google DeepMind played Go against Lee Se-Dol who is famous professional Go player in Korea. This match has attracted attention. Many people expected that Lee Se-Dol would win this match and no one doubted it. However, against expectations, the result was that AlphaGo had overwhelming victory with a score of 4-1. It was the moment when a new Go champion appeared. It is proven that AI has increased to a level, where it is impossible for a human to defeat AI in calculation and strategy. Watching the AlphaGo beat him, many people came to realize the boundless potential of artificial intelligence. With the advent of the high level of AI technology, AI is being combined with various industry.
Nowadays, varied unimaginable products are being appeared. For example, there is an image translator which recognize words in the picture, robo-advisor which offers advice or automatically invest client assets and symptom checker which helps people to get the check-up in home. The one thing in common that should be noted about these systems is that these replace human tasks. Many people are looking forward to AI will do simple or complex tasks for humans and trend also has been changing this way. There is no doubt that artificial intelligence is helpful and useful to humans in many ways. The PWC, a professional services firm in the UK, estimated that AI could be contributed to 14% global GDP until 2030 same as 15 trillion dollars at today’s values (Berriman, R., 2017). AI can significantly improve work efficiency and save labor costs. This essay will determine the trend of AI, analyze the extent that AI systems will replace jobs in the future and discuss the advantages of AI and misunderstanding about AI.
First of all, look at the history of AI, In 1956, John McCarthy came up with the term of Artificial intelligence (AI). It is the simulation of human intelligence processed by machines, especially computer systems. It is designed to learn themselves from big data and solve a complex problem. In the past, automation was structured formulaic tasks based on explicit rules. Autor (2013), a labor economist at MIT University in the U.S., suggested an analysis of the labor market impact of automation. He believed that all jobs consisted of bundles of tasks and the more repetitive tasks followed certain rules, the more likely jobs were to be automated by computer programs. automatical tasks included not only mental labor, such as book-recording but also alternative physical labor. On the other hand, tasks that did not follow explicit rules were defined as unstructured tasks and were considered difficult to automate such as driving a car or preparing legal documents. These tasks were considered to have the characteristics to enhance skills through human experience and training and they thought that these tasks would be difficult to replace with a computer program.
The Polanyi's paradox can be one of the reasons that describe unstructured tasks are hard to automate. If one sentence sums up Polanyi's paradox, it is that people know more than words can tell. What is difficult to put into words means that it is difficult to teach computer using explicit rule. The Polanyi's Paradox has long been a drag on artificial intelligence researchers. However, today's artificial intelligence technology is breaking the boundaries of automated work by circumventing Polanyi's paradox. As computers learn from the various information and improve their own algorithms by deep learning, they can acquire knowledge through experience just like humans. An article about the education of AI described that Deep learning is a key to solve progressively complicated assignment with improving accuracy as time passes (Peters, 2017). In the case of AlphaGo, instead of coding how to play well a game of go, it was designed by inducing computers to learn the winning strategy from the match data. It was able to accumulate skills that surpassed human by deep learning playing with itself million times. This approach is being combined with virtual environments like gaming, as well as with robots interacting in real life, producing results in areas ranging from collaborative robots to self-driving cars. As the development of an artificial intelligence capable of self-aware and judging the situation is accelerating, applications are spreading to IT services, home appliances, cars, and factories. Attention is being focused on the changes that artificial intelligence will bring to the labor market accordingly. Moreover, it can be divided into 3 steps. The first one is Artificial Narrow Intelligence(ANI), the second one is Artificial General Intelligence(AGI) and the Last one is Artificial Super Intelligence(ASI) (Hussain, 2018).
ANI is the a fundamental step where users simply communicate with AI. It includes virtual assistant in mobile such as SIRI by APPLE and Bixby by Samsung. AGI can perform human-level tasks without control such as a self-driving car, robot workers in the factory. ASI is a step referring to the time when AI is superior to human intelligence. Most experts would not agree that societies have reached this level. IBM Watson is considered a supercomputer which participated to the quiz show. Watson and AlphaGO were closest to ASI. The interest of AI is rapidly increasing around the world. According to MGI report (Bughin et al., 2017), global companies invested between 26 to 39 billion dollars in the development of AI technology in 2016. This investment growth is three times more than in 2013. In may this year, the Korean government announced that it would invest 2 billion dollars for the development of AI technology until 2022 (The Medium, 2018). In addition, the China government are planning to build the world's largest AI research park near Beijing (Kharpa, 2018). They said it would be completed within 5 years. AI can bring us many profits.
Firstly, it considerably releases us from performing “three D’s”, that is jobs that are dull, dirty, or dangerous (Lin et al., 2014). For instance, AI robots obediently carry out numerous tasks which people are reluctant such as mindless and repetitive assembly in the factory, clearing the dirty drain and removing dangerous mines. Moreover, Robots can also be put into delicate and accurate tasks. It can take up the scalpel instead of the human, take care of old people and children and identify enemies quickly in combat. For example, Phalanx CIWS, a U.S. Navy weapon system can exactly detect enemy missiles that approaching on a ship and immediately destroy it. In a way, robots can be faster and more efficient than humans. AI robots will work silently without a complaint if people ensure a stable supply of electricity to them.
Secondly, It classified the ways in which AI can make benefits in four areas: Project, produce, Promote and Provide (Bughin et al., 2017). Companies enable better project and predict to anticipate demand, optimize R&D. They can increase the ability to produce products and services using lower cost and good quality with an attractive advertisement. It is able to help promote goods at the appropriate price. It allowed them to provide plentiful, personal, and memorable experiences. Examining the AI-related business effect from recent real application case of manufacturing, it increased 30% of the material delivery time and 13% earnings before interest and taxes. Furthermore, it allowed saving 12% of fuel costs by using a machine learning for manufacturers and customers.
Thirdly, AI can considerably promote work efficiency and save labor costs by replacing tasks that humans have to do manually. AirAsia, a Malaysia airline saved about 9 million dollars by introducing AI technology called Flight Efficiency Services, which suggest an optimal path by analyzing data collected from an airplane (Data and Analytics driving Success at AirAsia, 2018). It also allowed the pilot to only focus on control. In Nagasaki, Japan a hotel operated by robot workers based on AI technology has attracted attention. This hotel reduced labor costs by using AI robots replacing 70% of human tasks such as check-in, guiding to the room and carrying luggage.
In contrast, many people are concerned about job losses. In the future, AI expected to replace various jobs. In the recent article (Berriman, 2017), it found that approximately 30% of UK jobs have the high probability which can be replaced by AI until the early 2030s, lower than the USA at 38% or Germany at 35%, but higher than Japan at 21%. Especially, the four key industry sectors such as wholesale and retail trade (44%), manufacturing (46%), administrative and support services (37%) and transportation and storage (56%) were researched at potential the high risk of automation based on AI, but the health and social welfare sectors is the lowest at 17%. However, there is no need to worry about it for now. Most companies did not introduce an AI system in their business yet recently. According to a survey of 3,000 executives in 14 sectors in 10 countries, only 20 percent use artificial intelligence technologies related to their businesses (Bughin, 2017). As a review of more than 160 cases, only 12 % of all cases were commercially deployed. The reason why is that many firms said that they were uncertain of the business case of AI and expected income. They thought AI is premature and far too early to adopt. The field of artificial intelligence also will create new jobs. Wilson, Daugherty and Bianzino’s study (2017) mentioned that companies that deal with advanced AI systems will need employees who can describe complex AI algorithms to non-specialist. They explained three categories of jobs: Trainers, Explainers, and Sustainers. Explainers will act as a bridge between technologists and businessman and will help in understanding opaqueness systems what they need. Trainers will need workers to illustrate AI systems how they use and solve errors. Sustainers will help guarantee that AI systems are running as designed and compensate for a sudden accident.
Looking at the newest research (Manyika et al., 2017), 250 million jobs connected with AI will be created until 2030 by increasing incomes. In addition, research of firm Gartner (Poitevin et al., 2017) said more jobs will appear than lost. They analyzed that AI will transform into a positive job motivator, creating 2.3 million jobs whereas reducing 1.8 million jobs. In 2021, it will be expected to that AI will generate 2.9 trillion dollars in business value and create 6.2 billion dollars hours of worker productivity. Besides, some people think that AI can be a threat. They even imagine human will be dominated by AI in the near future. Stephen Hawking, the world-famous physicist, even warned that artificial intelligence could lead to human extinction. However, these ideas are groundless fantasies. Demis Hassabis, representative of deep-mind said that artificial intelligence should be used as a research assistant to help scientists. Artificial intelligence is just a smart technology just as a cog or spring cannot magically turn itself into a murderous killing robot. our smart software embedded within their products cannot turn itself into a malevolent AI (Bentley, 2018). Looking back on history, mankind has always faced a new invention with fear. Erik Brynjolfsson, an MIT economist who wrote The Second Machine Age, he thinks that movement against AI corresponds with 21st century Luddite riots. In the 19th century, The Luddite movement in Europe demonstrated fears about the new invention of the machine. They called the machine a monster that steals a job and smashed machines, but nowadays new jobs were created and mankind is living much better than before by the benefit of a monster. Like this, AI is still not familiar to us yet, but it must be a turning point for a new era.
In conclusion, the wave of automation by artificial intelligence is still in its infancy. Although there are signs of gradual commercialization of artificial intelligence throughout the industry, some experts say that there is still a long way to go. However, considering the development speed of artificial intelligence in recent years, it cannot be ignored. The utilization of machine learning algorithms that analyze vast amounts of data can recommend products for customers and match demand and supply. The introduction of artificial intelligence will spread across industries while artificial intelligence is combined with advances in image, voice recognition and robot technology. There are many advantages of artificial intelligence. AI technology can significantly enhance work efficiency and reduce labor costs by replacing tasks that humans have to do manually. Furthermore, the diverse tasks of three D’s which people are reluctant are replaced by an AI robot. An analysis of the effects of artificial intelligence on jobs in several countries found that on average about 30 percent of jobs can be replaced by artificial intelligence. transportation and storage (56%) are estimated at potential the high risk of automation based on AI. the risks of automation by artificial intelligence on a particular job, industry or class can be a significant challenge in the future. This is because massive structural unemployment could occur in certain jobs and industries. Even when new industries and decent jobs are created, social costs can be increased as unemployment and polarization issues are highlighted. On the other hand, more jobs are expected to be created by artificial intelligence. Despite the development of artificial intelligence, survived unique jobs will become more valuable resources in the future.
To successfully accept AI technology, people should discuss and prepare for any possibilities that may arise from the development of AI. With the spread of artificial intelligence becoming increasingly visible, individuals need to develop vocational skills that can use artificial intelligence as a complement to their work. Companies need to create flexible organizational structures that can cope with the AI age. Companies that accumulate vast amounts of data that will be used for machine learning sometimes enter other industries. The contrast between companies that can increase productivity and create new value for customers can become more apparent. It is necessary to organize and deploy personnel who can effectively automate tasks based on artificial intelligence. Michael Dell, the founder of the Dell, said that if artificial intelligence is a rocket, data is the fuel that pushes forward. Therefore, consideration of collect data is essential for utilizing artificial intelligence and should also be given first priority. The government should take steps to enhance the flexibility of the labor market to prepare drastic changes in the employment environment. To deal with a change of industry, they should arrange diversification of forms of employment and flexible labor market. At the same time, to protect the vulnerable social group, they should reinforce community safety nets such as re-education, outplacement support program and unemployment insurance. For responding to the basic income, robot tax, and innovations in techniques, they should mitigate the uncertainty of job losses through examination, a policy experiment. Moreover, to take action against the retrenchment of the knowledge life cycle, they should tighten the lifelong learning system.
