The Dark Side of AI Investment: White-Collar Layoffs and the Structural Destruction of the Labor Market [EN]

"The massive movement of capital into AI technology is not a simple infrastructure investment. 

It is a structural transition that permanently replaces corporate operating expenses (OPEX) with capital expenditures (CAPEX)."

— Goldman Sachs Asset Management (2026)

Prologue: The Market Observer's Perspective

In the spring of 2026, the most frequently heard word from Silicon Valley to Wall Street is 'efficiency.' Behind this word hides a massive systemic shift. While global Big Tech companies pour hundreds of billions of dollars into building AI data centers and GPU clusters, paradoxically, tens of thousands of high-wage knowledge workers are packing their bags at their headquarters. This shows that the white-collar class is following the exact trajectory of factory automation experienced by blue-collar workers in the past. Capital is no longer investing in the human brain; it is heading toward the computational power of silicon chips. How will this macroeconomic inflection point, where the value of knowledge labor falls and the monopoly of capital deepens, ultimately collapse the foundation of the global consumer market?

EXECUTIVE SUMMARY

As of 2026, AI-related capital expenditures (CapEx) of the four major US Big Tech companies (Microsoft, Amazon, Alphabet, Meta) are projected to exceed $650 billion to $700 billion. This massive hardware investment is accompanied by a structural destruction within companies, replacing labor-centric operating expenses (OPEX) with AI algorithm-centric capital expenditures. The IMF and major employment statistics clearly indicate the extinction of white-collar jobs and the polarization of the labor market, which acts as a massive systemic risk, triggering a decline in the disposable income of the middle class and eroding the domestic consumption fundamentals of the macroeconomy in the long run.

01. The Historical Expansion of AI Infrastructure Capital Expenditures (CapEx)

└ Big Tech's $700 Billion Bet and the Concentration of Capital

Currently, the liquidity of the global capital market is being sucked into artificial intelligence hardware infrastructure like a black hole. According to a March 2026 Goldman Sachs Asset Management report, the total CapEx of global hyperscalers in 2026 is projected to be a staggering $737 billion, which is nearly five times the 2023 figure ($160 billion).[¹] In particular, Amazon allocated $200 billion and Alphabet up to $185 billion for 2026 alone.[⁵] This is an unprecedented concentration of capital into a single sector in history.

└ The Structure of OPEX to CAPEX Substitution

The essence of this massive infrastructure investment is not mere service expansion. Companies are reducing operating expenses (OPEX) previously consumed to maintain human resources, converting them into capital expenditures (CAPEX) to secure computing resources. In other words, it is a structural macro-transformation where the labor costs of middle-tier white-collar workers in charge of coding, translation, research, and data analysis are being completely replaced by the depreciation and power costs of GPU servers.

02. Structural Cause Analysis: The Redundancy of the White-Collar Workforce

└ Penetration of Generative AI into the Middle-Class

While past robotics and automation technologies replaced physical labor-centric blue-collar jobs, current Large Language Models (LLMs) and generative AI are squarely targeting cognitive knowledge labor. For mid-level knowledge processing tasks that do not require complex reasoning or high-level creativity, AI provides instant outputs at a fraction of the cost. As a result, the office worker group that acted as the 'waist' within corporations is being rendered systemically redundant.

└ The Zeroing of Marginal Costs in Knowledge Labor

Once an AI model is built, the marginal cost of generating additional text or writing code practically converges to zero (0). This means that a company's productivity can be expanded infinitely without hiring additional labor, fundamentally extinguishing the economic incentive for capitalists and management to employ white-collar workers.

03. Data and Statistical Verification: Polarization Indicators in the Labor Market

└ Visualization of AI-Driven Layoff Data

Cracks in the labor market are already appearing in clear figures. According to a January 2026 report by global insight firm Challenger, Gray & Christmas, US employers announced a total of 1,206,374 layoffs throughout 2025, a 58% surge from the previous year and the highest level since 2020.[²] Notably, over 150,000 layoffs occurred in the tech sector, and since 2023, there have been approximately 72,000 cases where companies explicitly cited 'AI adoption' as the direct reason for job cuts.[²]

└ Warnings from the IMF and Global Research Institutions

Through its 'Bridging Skill Gaps for the Future: New Jobs Creation in the AI Age' report published in 2026, the International Monetary Fund (IMF) formalized the polarization of the labor market in advanced economies.[³] The IMF warned that while AI technology increases the wages of highly skilled top-tier workers, it directly reduces employment in professions with low AI complementarity, causing a shrinking of the middle class.

04. Systemic Ripple Effects: The Risk of the Macroeconomy's Consumption Base Collapsing

└ Middle-Class Collapse and the Entrenchment of a K-Shaped Consumption Market

The loss of white-collar jobs goes beyond a simple rise in the unemployment rate to become a fatal detonator for the macroeconomy. When the disposable income of the middle class, the backbone of the economy, decreases, the 'mass consumption' system that has driven global economic growth takes a hit. The market is already fracturing into a strict K-shaped consumption structure, forcing a structural deterioration in earnings for domestic demand-focused companies.

└ Cracks in the Corporate and Income Tax Revenue Structure

When AI replaces human labor, the 'earned income tax' base, which is the core of national finances, shrinks. Conversely, as capital and profits concentrate in a handful of Big Tech companies, national-level tax collection becomes more difficult. This has a high potential to deepen the fiscal deficits of various governments and trigger a chain collapse of social welfare systems.

05. Historical Analogy Comparison: Blue-Collar Workers of the Industrial Revolution vs. White-Collar Workers of the AI Revolution

└ The Parallel Theory of Mechanized Large-Scale Industry and Artificial General Intelligence (AGI)

The introduction of power looms during the 19th-century British Industrial Revolution evaporated the jobs of skilled artisans (Luddites). While extreme unemployment occurred in the short term, new forms of jobs were created in the long term. However, the AI revolution of 2026 differs fundamentally in that it universally replaces human 'cognitive ability' itself, not the 'physical body'. Whether new labor domains where humans can hold a comparative advantage over AI will be created on a massive scale as in the past remains an unproven macroeconomic gamble.

06. Variables and Limitations: Labor Union Resistance and Government Regulations

└ The Rise of the 'AI Tax' (Robot Tax) Debate

As the employment shock materializes, discussions around a so-called 'Robot Tax'—levying taxes on AI infrastructure to fund Universal Basic Income (UBI) for displaced white-collar workers—could rapidly emerge, primarily centered in the European Union (EU). This is a structural risk that exerts severe downward pressure on the AI Return on Investment (ROI) models of Big Tech companies.

└ The Possibility of a Prolonged Productivity Paradox

Despite massive CapEx investments, US macroeconomic indicators do not immediately show a nationwide economic growth (GDP) enhancement resulting from AI adoption.[⁶] If a 'Productivity Paradox' persists—where efficiencies are achieved within specific companies but fail to translate into overall macroeconomic growth—the risk of the capital market's AI investment frenzy rapidly cooling remains ever-present.

Macro Scenario: Probabilistic Future Trajectories

Scenario A (Base Case): Continuous Contraction of White-Collar Jobs and the Gig Economy Transition

AI infrastructure investments are executed as planned, and middle-management jobs in large corporations gradually disappear by 3-5% annually. Laid-off white-collar workers move downwards into platform-based gig work or low-wage service jobs, and income polarization and a K-shaped consumption market fully settle as the economic New Normal.

Scenario B (Structural Shift Case): Bursting of the AI CapEx Bubble and Capital Market Correction

Trigger: If, in 2026-2027, Big Tech companies that invested hundreds of billions of dollars fail to prove sufficient AI monetization in the B2B market that exceeds the savings in white-collar labor costs.
Result: The 'AI investment payback delay risk' originating from Silicon Valley is highlighted, triggering intense de-leveraging centered on tech stocks. The entire hardware value chain, including data center construction, faces a rapid negative growth phase.

Scenario C (Tail Risk Case): Mass Unemployment and the Clash of Populist Regulations

Trigger: If the advent of high-performance agentic AI simultaneously replaces high-income professional groups such as accounting, legal, and medical assistants on a massive scale.
Result: Unemployment spikes in the short term, sparking political riots and white-collar solidarity strikes. Governments worldwide impose extreme 'AI adoption quotas' or punitive taxes, forcibly halting the pace of innovation.

Implications from an Investor's Perspective

Short-Term (1-2 years from the writing date)

The $700 billion Big Tech CapEx confirmed by market institutions like Goldman Sachs guarantees unprecedented, definitive demand in the physical hardware value chain, including semiconductors (GPUs), server power grids, and cooling systems. The infrastructure build-out cycle is the safest investment haven in the short term.

Mid-Term (3-5 years from the writing date)

The profit margins of non-tech blue-chip companies that successfully slim down their organizations and reduce labor costs (OPEX) due to AI adoption will structurally improve. Investors must select companies in finance, healthcare, etc., that have successfully internalized AI.

Portfolio Perspective

Reflecting the risks of middle-class collapse and reduced disposable income, a structural hedge is essential: reducing exposure to mass-market consumer goods and building a barbell strategy focusing either on the luxury industry targeting the ultra-wealthy or the completely ultra-low-cost (discount) retail industry.

Conclusion

The explosive expansion of AI capital expenditures is not merely a boom in the tech sector. It is a massive structural reform of humanity's traditional capitalist system, where wealth was distributed in exchange for labor. Behind the hundreds of billions of dollars in CapEx spent by Big Tech companies looms the shadow of the permanent income loss of the tens of thousands of white-collar workers they have laid off. Market participants must coolly observe whether the current AI revolution is a blessing that enhances economic productivity, or a self-contradictory system error that extinguishes the 'consumer,' the ultimate engine of the macroeconomy. The ruins of the labor market left behind after the infrastructure investment boom ends will ultimately be billed as a massive cost to the macroeconomy.

※ Disclaimer

This report does not recommend the purchase or sale of any specific assets, nor does it support or criticize any specific regime, government, or politician. It is an article of macroeconomic system analysis based on publicly disclosed data and historical indicators. It is impossible to predict all market variables, and the responsibility for all judgments and subsequent consequences rests entirely with the reader. While the author (Neutral Observer) makes every effort to ensure the reliability of the analysis, the flawless accuracy of the provided information is not guaranteed.

Sources and References

 
• [¹] Goldman Sachs Asset Management, Learnings from Earnings (2026.03) — Estimated data for hyperscalers' 2026 CapEx of $737 billion
 
• [²] Challenger, Gray & Christmas, 2025 Year-End Challenger Report (2026.01) — Total 1.2 million layoffs in the US in 2025 and AI-related job cut indicators
 
• [³] International Monetary Fund (IMF), Bridging Skill Gaps for the Future: New Jobs Creation in the AI Age (2026.01) — Analysis of white-collar job polarization and middle-class collapse risk in advanced economies
 
• [⁴] World Economic Forum (WEF), Future of Jobs Report 2025 (2025) — Structural forecast of global job group declines due to AI technology advancement
 
• [⁵] MLQ.ai Research, Big Tech's $650-700 Billion AI Infrastructure Push Reshapes Cash Flow Dynamics (2026.02) — Detailed CapEx guidance statistics for the top 4 Big Tech companies including Amazon and Alphabet
 
• [⁶] Reddit / Goldman Sachs Research Cites, Massive investment in AI contributed basically zero to US economic growth last year (2026.03) — Limitations of short-term productivity contribution (Net-Zero) of AI investment and macroeconomic productivity paradox