A next-generation electric Atlas prototype is raising its hand in greeting at the Hyundai Motor Group press conference held in Las Vegas, Nevada, USA on January 5 (local time), a day before the opening of ‘CES 2026’. Yonhap News
[주간경향] #Scene 1. Hyundai Motor Group introduced physical AI ‘Atlas’ at ‘CES 2026’, an international consumer electronics and IT exhibition held in Las Vegas, USA on January 5 (local time). This robot, which can perform delicate tasks like humans through hands equipped with tactile sensors and can lift a weight of up to 50 kg, is scheduled to be used in automobile production plants in the United States and other countries starting in 2028. Initially, they will work on relatively simple processes such as parts classification, but from 2030, they plan to expand the scope of work to include parts assembly. To this end, we decided to establish a system capable of mass producing 30,000 units by 2028. It is said that when Atlas is put into a factory, humans will take on the role of learning and managing the robot so that it operates well. Amid expectations for a ‘productivity revolution‘, Hyundai Motor’s stock price soared by 80% (as of January 21st closing price) in the new year alone.
#Scene 2. It was the labor union that responded after the market. The Hyundai Motors union announced in a newsletter on January 22nd that “an employment shock is expected when the Atlas is mass-produced and put into production” and “not a single unit can be brought in without labor-management agreement.” The union also pointed out that one Atlas can be purchased for 2-3 years of a production worker’s annual salary, and that maintenance costs of approximately 14 million won are only required thereafter. Through this, they argued that Atlas threatens the employment stability of workers and serves to maximize profits for capitalists.
#Scene 3. The response to the union’s opposition was cold. Related articles included comments criticizing labor unions, such as “It’s only a matter of time before they are replaced by robots,” “21st century Luddites,” “Let’s replace everyone with robots and lower car prices,” and “You can’t go against the trend.” There were also many articles criticizing the union by mobilizing the union’s long-standing label of ‘noble union’ or linking it to Hyundai Motor Company’s stock price, which began to decline immediately after the union’s response. President Lee Jae-myung said on January 29, “It would be good to talk seriously about basic society,” and added, “There were movements to destroy machines in the past, but in the end, we could not avoid the huge cart that was flowing in. In the end, we have to quickly adapt to that society.”
Although companies, workers, and consumers seem to be making their own claims, the three scenes also have something in common. The idea of how the technology called Physical AI will be used is the same and clear. Whether they are in favor from a consumer perspective or against it from a worker perspective, everyone believes that ‘robots will enter factories and replace people.’
This ‘imagination’ of the Atlas robot reminds me of the technology and vision mentioned by Darren Acemoglu. In his 2023 book “Power and Progress,” Nobel Prize-winning economics professor Ashmoglu from MIT said, “We should not be too fooled by the monumental technological progress that humanity has achieved. A shared vision may trap us.” According to Prof. Acemoglu, historically technological advancements have not always raised everyone’s standard of living. Whether the fruits of technological development will be monopolized by a few or enjoyed by many depends entirely on society’s choice. What influences this choice is the imagination, or vision, of how the technology will be used and how the results will be shared.
Tesla CEO Elon Musk said, “Robotics and AI are the path to abundance for everyone,” but opinions differ. This is because it is impossible to know what kind of prosperity awaits workers who are immediately pushed out of the workplace.
We looked at the questions our society should ask in the controversy surrounding the ‘Atlas factory deployment’. It is necessary to consider whether the imagination of the ‘end of jobs’ is realistic, and if it is not realistic enough, is there anything to worry about? If the ‘future in which robots replace humans’ is truly inevitable, is society preparing for the future? There is no utopia that unfolds spontaneously. There is no dystopia that we have no choice but to accept. It is time for society to constantly ask questions to choose the direction of technological development that is better for everyone.
Job replacement or transition?
There are mixed views on how much jobs will be affected by the introduction of AI. Immediately after the advent of generative AI such as ChatGPT, the prevailing view was that jobs would be replaced. For example, looking at the forecast released by Goldman Sachs for 2023, it appears that one in four jobs in the U.S. and Europe is likely to be automated due to AI. Globally, it was estimated that about 300 million full-time jobs could be replaced.
As generative AI is used in the workplace, the outlook is gradually being revised. Recent forecasts emphasize that AI will change the way we work rather than replace jobs. According to the International Labor Organization (ILO)’s ‘Generative AI and Jobs (2025)’ report, one in four workers around the world said their work would be affected by the introduction of generative AI. Occupations such as general office workers and customer center counselors were classified as having a high possibility of having their work automated. Although it seems similar to Goldman Sachs’ forecast, the ILO predicted, “Because AI continues to require human input, jobs will change rather than be replaced.” This reflects the fact that complete automation is not as easy as initial expectations about AI.
Most of the changes that have changed our lives have come gradually, and it appears that AI and human jobs will do the same. For example, if you look at the case of a call center where AI was introduced for consultation, you can see how work is changing. On January 14, the Federation of Korean Trade Unions announced the results of a survey on the introduction of AI conducted among call center agents and customers. As AI took over primary consultation, the average number of consultations for call center agents decreased by 13.9%. However, the average call time per consultation increased from 6.95 minutes to 7.55 minutes. AI handles cases that require simple guidance, and human counselors handle complex cases, increasing the level of difficulty. Some counselors responded that the post-processing time to organize counseling cases and input information after counseling also increased. It is said that in some cases, counselors input counseling data to train AI.
In a survey of customers, only 18% responded that they were ‘satisfied’ with AI consultation, while 54.2% responded that they were ‘not satisfied’. Many responses said that ‘consultation time increased (43.8%)’ through AI consultation, and ‘I did not receive help in resolving problems (40.8%).’ The response that ‘I prefer human counselors’ was also overwhelmingly high at 87.5%. In summary, with the introduction of AI, customer satisfaction has decreased and the difficulty of counselors’ work has increased. Permanent job insecurity is a bonus. Beginning in 2023, the banking sector attempted to reduce the workforce of AI-introduced counselors, and last year, the Korea Scholarship Foundation’s call center virtually fired counselors.
This is what the call center example means. The dramatic situation where humans are replaced by AI has not yet unfolded. However, AI threatened some jobs and increased the difficulty of work for remaining workers. This may be a phenomenon that occurs because AI is in a transitional stage of development. However, there is no guarantee that when AI is fully evolved, the job situation for workers will be better than during this transition period.
A humanoid robot is shaking hands with a human. Senior Reporter Seo Seong-il
In the early 1800s in England, workers who were in danger of losing their jobs due to the introduction of textile machines destroyed the machines. The famous ‘Luddite movement’ is mainly used today to refer to people who are unable to adapt to the changing world and are opposed to the introduction of technology. Lee Sang-heon, ILO employment policy director, says in the book “Why Good Jobs Are Always Shortage” that we need to reexamine the entire Luddite incident. “What the workers were fighting was not technological innovation or the new machines themselves, but the lack of social support to help people adapt to the process of change they brought about. It was fighting poverty, not machines, and fighting the indifference of social and political forces.”
How is society preparing for the losers who suffer during the transition period and the ‘labor force replacement’ that may come in the later stages? Let’s assume that things go as predicted by technology optimists. Productivity will increase and jobs will decrease. Since there are no jobs, purchasing power decreases, and even if there are a lot of goods, there are few people who will buy them. To prepare for this situation, there are ideas presented by technology managers. The idea is to tax robots that steal jobs (robot tax) and use the tax to provide income to people (basic income). However, while robots are quickly approaching factories, the institutionalization of basic income seems to have a long way to go. And the greater the speed difference between the former and the latter, the greater someone’s suffering is bound to be.
The National Artificial Intelligence Strategy Committee under the President also has a social division to discuss these issues. However, at present, only a small number of job translations affected by the introduction of AI are being studied. It is said that retraining is being discussed as a measure to alleviate the employment shock. A member of the National Artificial Intelligence Strategy Committee said, “If AI’s entry into the industrial field is unavoidable, the social challenge is how to create a safety net. However, the speed at which AI is spreading does not match the speed at which countermeasures are discussed. There is no guarantee that AI will not develop and even replace those jobs while people are retraining for re-employment.”
This is also unrelated to the situation where AI optimism has become the dominant vision of society. This is because the center of gravity of the government, which must design social systems in line with new technologies, is focused on technological development rather than measures. Regarding the ‘AI Basic Act’, which is the minimum guideline regarding AI, the government said, “We will operate the promotion law with an emphasis on the development of the AI industry so as not to hinder the innovation of startups.” In fact, the Basic AI Act narrowly defines ‘high-impact AI’, in which safety management responsibility is given to operators, and ‘surveillance robots (AI)’ etc. are excluded from regulation.
Professor Lee Gwang-seok of Seoul National University of Science and Technology said, “Korean society’s technological optimism is very strong. There has been a sense of efficacy in state-led technology investment since the industrialization era. Even before physical AI, Korea was an overwhelming leading country in the industrial automation phase. There is a side where we will listen to opposing arguments, but we dismiss them as holding us back.”
There is no right answer to how to prepare for the future. However, there is a definite theory about which way to find your way. The ILO said that “social dialogue between the government and labor and management and consultation within the workplace are essential” to simultaneously improve working conditions and increase productivity due to AI. The Economic, Social and Labor Committee, a social dialogue organization in Korea, published a green paper containing 12 questions regarding changes in the labor market due to AI in November last year. As social dialogue in which all labor, management and government participates is not possible, they end up publishing a ‘green paper’ that only contains questions, rather than a ‘white paper’ that contains answers to questions.
Kim Byeong-kwon, director of the Green Transformation Research Institute, said, “The Hyundai Motor Company Atlas case became a problem when AI was introduced that ran counter to the interests of workers. If the Economic and Social Affairs Committee or the National AI Strategy Committee had worked properly and created a consensus (on how to use AI, etc.), a voice should have been raised to mediate conflicts when such problems arise. If it does not work properly, citizens, creators, and workers have no choice but to be pushed out unilaterally when there are conflicting interests related to AI.”
Regarding the declaration of war on Atlas, some people say, “Because it is Hyundai Motors’ union, they are raising the issue in a visible way.” This means that it is difficult for unions with weak bargaining power or workers without a union to raise their voices regarding the introduction of AI. In a survey conducted by the Federation of Korean Trade Unions on call center counselors, only 1.5% of respondents said that labor-management discussions were held regarding the introduction of AI counselors.
Song Gwan-cheol, a researcher at the Korea Labor and Society Research Institute who conducted a survey on call center counselors, said, “The union believes that it should be a subject of labor-management consultation because it affects working conditions and is linked to job insecurity. On the other hand, the company’s position is that consultation is necessary to introduce convenient work tools. However, looking at previous experience, there are things that come in as if they are not a big deal and then later replace jobs, so conflicts are bound to arise.”
In the end, the social solution depends not on criticizing the Hyundai Motors union, but on how well the voices of those with different interests from AI optimism are reflected in the policy-making process. Hani Kim, a member of the Digital Justice Network Steering Committee, said, “We need discussions involving people affected by AI. It is not a matter of how to foster technology and grow the industry, but a social consensus is needed on how to use this technology and how society can prepare for the side effects that come with it.” Director Kim Byeong-kwon said, “Today it’s a problem that occurred at the Hyundai Motors factory, but tomorrow it may happen in the Yeouido office and the day after tomorrow in a culture and art space. The government is in a situation where it needs to take action on how to prevent and protect the workplace from damage caused by AI. It is not desirable for the future AI era to even talk about Luddites and drive workers into being who go against the trend of the times.”
