Agenda

• Moderator:
  Li-Hsuan Liang, Deputy CEO (NIIEPA)
• Panelists:
  –Hung-Yue Suen, Associate Professor (Department of Technology Application and Human Resource Development, National Taiwan Normal University, ) -online
  – Seraphim Ma, Managing Partner  (Baker McKenzie Taipei)
  – Cherng-Tay Hsueh, Retired Professor (Social Policy Research Center, National Taiwan  University) -online
  –Tsung-Che Wei, Associate Research Fellow (Introduction of The Third Research Division, Chung-Hua Institution for Economic Research) -online

15:45-16:00  Q&A

Meeting Minutes

Professor Hung-Yue Suen shares recent research data from several academic institutions. Firstly, a study conducted at the Massachusetts Institute of Technology in the United States indicates that ChatGPT has the most noticeable impact on improving the efficiency or quality of work for entry-level white-collar workers. However, it is also likely that artificial intelligence technologies, including ChatGPT, may replace tasks performed by white-collar workers, potentially reducing the demand for human labor by employers. In contrast, the impact on blue-collar workers is relatively smaller.

Research from OpenAI and the University of Pennsylvania suggests that 80% of workers will be affected by AI in some way, with job types most profoundly affected including translation, text-based tasks, data analysis, accounting, administrative work, creative professions, journalism, and software engineering, among others. Furthermore, AI is expected to reduce many high-paying white-collar job opportunities, posing a particular risk to those with advanced degrees. The International Labour Organization predicts that jobs related to administration and clerical work will be most affected by AI, as while AI can automate tasks and improve efficiency, it also increases job uncertainty.

AI will gradually reshape the methods and content of work, meaning that the ability to operate AI proficiently will become a necessary skill in the future. Even entry-level employees with AI assistance can perform high-level professional tasks. Considering cost and efficiency, businesses may increasingly prefer using AI to replace mid to high-level project-based positions. According to a report from the McKinsey Global Institute, by 2030, the U.S. labor market will undergo significant changes due to AI. Approximately 30% of jobs, including logistics, customer service, and sales, will be replaced by AI and automation systems. Lower-paying positions such as clerical and retail jobs are expected to decrease by more than 600,000, with these roles likely being replaced by AI, reducing opportunities for career transitions. However, high-earning knowledge workers may still be affected, as generative AI has the potential to play a role in customer management, marketing, software engineering, and other fields, presenting greater challenges for those with higher education levels.

Lawyer Seraphim Ma believes that AI will reshape the overall social and industrial ecosystem, impacting various fields such as technology, media, law, research, finance, and more. Jobs involving simple data collection, organization, and output are particularly vulnerable to AI replacement. The beneficiaries of AI are likely to be concentrated among technical elites and capital owners, potentially widening income disparities and increasing social and economic risks for low-skilled laborers and vulnerable groups.

Potential legal disputes arising from the impact of AI on labor include scenarios where employers use AI to predict employee turnover or violations, biometric identification and monitoring of employee health, remote monitoring, and the use of AI to assess Key Performance Indicators (KPIs) and employee performance. Given the difficulty for employees to refuse such scenarios, it is recommended that future labor laws establish external mechanisms to prevent the misuse of AI tools.

Lawyer Ma also emphasizes that, according to the current interpretation of labor standards by the Supreme Court, if AI can replace a specific employee’s job (meeting the criteria for a change in the nature of the business), companies have legal grounds for termination. She believes that the use of AI in businesses should be regulated. For example, AI algorithms used by multinational corporations for job interviews may contain human biases and discrimination. It is suggested that biases be detected and corrected, and disclosure practices be implemented to eliminate AI bias. Ma also advises domestic companies to exercise caution when implementing AI tools to avoid violating the Employment Service Act due to AI bias.

AI is currently widely used in areas closely related to people’s lives, such as transportation, employment, justice, and public services, posing high risks and requiring regulation. Lawyer Ma cites examples such as the U.S. AI Bill of Rights Blueprint and the EU AI Act and suggests that Taiwan can establish regulatory mechanisms, including close cooperation between law enforcement teams in network security, IoT (Internet of Things) information security, widespread adoption of digital evidence, the establishment of dedicated agencies to research related issues, and the creation of mechanisms for ongoing discussions.

Lastly, Lawyer Ma advises future newcomers to the workforce that AI will not replace human jobs, but it will replace those who do not use AI. Having a habit and discipline of lifelong learning will be essential for competitiveness in the future job market. Currently, AI lacks a strong focus on thinking and strategy, and jobs with a strong social aspect remain beyond the reach of AI.

Professor Cherng-Tay Hsueh’s presentation focuses on the impact of AI on “human society,” and he believes that the definition of “work” will undergo significant changes under the influence of AI. He begins by referencing the book “AI is Coming” by Kai-Fu Lee and Yong-gang Wang, where they describe that if you condense 6,000 years of human history into one day, AI technology appeared in the last 10 seconds of that day, and generative AI appeared in the last five seconds. Professor Xue believes that these final five seconds will bring about polarized and profound effects on human society, leading either towards extinction or a new future.

From a sociological perspective, Professor Hsueh analyzes AI and identifies three key characteristics: data storage surpassing human memory, computational speed surpassing human intelligence, and self-learning surpassing human inertia. Human expectations of AI include hoping that it can serve as assistants, pets, butlers, and more, efficiently completing tasks without negative emotions or resistance.

Professor Hsueh further examines the short-term impacts (in the next 20 years) of AI technology, including the rapid transformation of electronic communication and transportation modes, disrupting current learning methods and potentially replacing the writing and thinking roles of university professors. The powerful analytical capabilities of AI may render conferences unnecessary and have an impact on essential social organizations such as governments, banks, and businesses, leading to their bubble-like behavior. Even the functionality of families may face challenges.

Professor Hsueh also issues a caution for the future beyond 20 years, suggesting that humans may begin to worry about the possibility of AI developing consciousness and a soul, which could pose an existential threat to humanity.

Dr. Tsung-Che Wei’s analysis primarily focuses on the economic and industrial aspects of AI. He acknowledges that AI has indeed begun to change human work, with AI assistance leading to increased work efficiency, primarily resulting in positive developments at this stage.

Dr. Wei proceeds to trace the development of AI technology. He explains that AI technology has evolved from being a simple tool that receives human instructions in the 1960s to becoming an expert knowledge repository that assists in professional decision-making. By the 2010s, AI started to mimic human behavior, and today it can even collaborate with humans in creative endeavors, with its imagination surpassing that of humans.

As of now, AI and robots have started assisting in physically demanding and injury-prone professions, such as heavy labor and caregiving. In an aging society with a decreasing workforce, AI’s assistance in labor and healthcare contributes to maintaining social stability and economic growth. The introduction and utilization of AI technology in industries can address issues such as labor shortages, excessive workloads, and low efficiency, thus having a positive impact. However, it may also lead to an increase in unemployment in the short term.

Dr. Wei believes that new advancements in AI technology will create opportunities for talent transformation in industries. Talent transformation takes time and may come with some challenges. The restructuring of industries due to AI is likely to result in a short-term increase in the unemployed population. However, over time, most of the workforce will gradually adapt, aligning themselves with the demands of the job market and helping to alleviate the social problems associated with widespread unemployment.

He also suggests that businesses should proactively plan for talent transformation. After the active implementation of AI technology in enterprises, without corresponding talent transformation practices, there is a risk of top-notch research and development talents leaving the organization, mid-level management talents facing reduced salaries, and frontline operational talents facing unemployment. However, with concurrent efforts in talent transformation policies, organizations can gradually realize a vision of enhancing both talent creation capabilities and salaries. This will also reduce the negative impact on the job market caused by AI.

Dr. Wei concludes by noting that the new developments in AI technology will likely lead to issues such as lower wages and increased unemployment. However, by cultivating the ability to apply AI among the workforce and improving the environment for human-AI collaboration, repetitive and labor-intensive tasks can be delegated to AI, while core strategic work remains the responsibility of humans. This approach can enhance the added value of human work.

Presentation Download <Provided with the consent of the speaker>

• –Hung-Yue Suen, Associate Professor (Department of Technology Application and Human Resource Development, National Taiwan Normal University, ) -online –Presentation Download
• Tsung-Che Wei, Associate Research Fellow (Introduction of The Third Research Division, Chung-Hua Institution for Economic Research) – Presentation Download