Goldman Sachs CEO Solomon: AI "Job Apocalypse" is an Overhyped Panic

Goldman Sachs CEO Solomon published an opinion article in The New York Times, offering a different perspective on the market panic that AI will trigger an "employment apocalypse." Solomon pointed out that AI automation is more likely to follow the path of past technological revolutions, eliminating some jobs while expanding others. Goldman Sachs economists predict that in the next decade, AI could automate 25% of current work hours, and the $700 billion capital expenditure by large cloud service providers has already driven growth in emerging employment sectors such as construction.
(Background summary: Bitcoin consolidates at $76k, ETFs lose $1.26 billion for six consecutive days, market awaits the first FOMC meeting on 6/17)
(Additional background: Breaking news! Foreign media reports that the US-Iran ceasefire agreement has been extended by 60 days! Iran promises to clear mines within 30 days to reopen the Hormuz Strait)

Table of Contents

Toggle

• The historical trajectory of AI’s impact on employment
• Specific data forecasts: 25% of work hours will be automated
• AI capital expenditure driving new employment markets
• The self-regulating capacity of the labor market
• Cross-country comparison: Impact of AI on Taiwan’s job market
• Policy recommendations: Investing in human capital
• Conclusion

Goldman Sachs CEO Solomon published an opinion article in The New York Times, offering a different perspective on the market panic that AI will trigger an "employment apocalypse" and mass unemployment. Solomon believes that the impact of AI on the job market is not as severe as market fears suggest; AI is more likely to follow the path of previous technological changes, eliminating some jobs while expanding others.

The historical trajectory of AI’s impact on employment

Solomon pointed out that technological revolutions over the past centuries have followed similar trajectories. During the Industrial Revolution, horses were replaced by cars, but driver and mechanic jobs emerged. During the computer revolution, typists were replaced by word processing software, but programmer and data analyst roles increased. During the Internet revolution, physical retailers were replaced by e-commerce, but logistics and digital marketing jobs expanded.

"AI is very likely to repeat the previous technological revolution’s path, eliminating some jobs while expanding others."

Specific data forecasts: 25% of work hours will be automated

According to the latest forecasts from Goldman Sachs economists, AI could automate 25% of current work hours over the next decade. This figure is not as alarming as imagined. White-collar industries will be most affected: banking, legal, accounting, software development, and customer service will face the greatest impact. Blue-collar sectors are relatively stable: automation in construction, manufacturing, and service industries will proceed more slowly. New roles will emerge: AI system management, implementation, validation, and regulation will continue to grow.

AI capital expenditure driving new employment markets

Solomon provided a concrete example to support his view. Just this year, major cloud service companies plan to invest $700 billion in capital expenditure, which has already spurred a surge in construction jobs. This $700 billion includes data center construction, chip factory expansion, infrastructure upgrades, and energy supply systems.

The self-regulating capacity of the labor market

Solomon emphasized that the U.S. economy has successfully absorbed multiple waves of technological shocks. During the Agricultural Revolution, employment in agriculture dropped from 70% in 1800 to less than 3% today. During the Manufacturing Revolution, it peaked at 33% and has fallen to 8%. During the Service Expansion period, employment grew from 30% to 80%.

"Overall employment and living standards continue to improve."

Cross-country comparison: Impact of AI on Taiwan’s job market

As a global center for semiconductors and technology manufacturing, Taiwan’s AI employment market will be significantly affected. In the semiconductor industry, leading companies like TSMC are actively integrating AI for process optimization, driving global investment in AI chips. In tech startups, Taiwanese companies are rapidly adopting AI tools to boost productivity, making software development and data science popular careers. In manufacturing, Taiwan is transforming from a "manufacturing powerhouse" to "smart manufacturing," with widespread use of robotic processes and machine learning applications.

Policy recommendations: Investing in human capital

Based on this analysis, Solomon proposed the following policy suggestions: education reform to strengthen STEM (science, technology, engineering, mathematics) education; vocational training to establish a "lifelong learning" system; expanding social safety nets, including unemployment benefits and health insurance; infrastructure investments to leverage AI for improving public service efficiency.

Conclusion

Solomon’s perspective offers an optimistic view: AI will not be the "end of employment," but rather a key to new "opportunities." The key lies in investing in human capital, enabling workers to smoothly transition into new roles. As AI technology continues to develop, the labor market will keep evolving, but history shows that humans possess remarkable adaptability.