But is it a bubble? It doesn’t appear that it is if you compare it to past industries. Although it is moving faster than any technology from the past it still has a long way to go. Near-future investments in AI and related industries (semiconductors, data centers, energy infrastructure) are projected to meet and potentially exceed, as a percentage of the U.S. economy, many past fast-growing industries like the early phases of automotive and aviation. Forecasts indicate this will occur due to an unprecedented velocity and concentration of capital.
Projected Investment Growth
Dominant Force: Analysts at Bank of America and Vanguard believe AI will remain "the place to be" for investment in the coming years, citing its ability to drive massive productivity gains across all sectors.
Massive Scale: McKinsey predicts a staggering $5.2 trillion will be invested in chips and data centers over the next five years. Global AI spending was nearly $1.5 trillion in 2025 and is expected to exceed $2 trillion in 2026.
Share of GDP: U.S. capital spending on AI infrastructure currently accounts for around 1.2% of GDP in 2025, which is already set to rise to a projected 5.6% of GDP by 2030. This would surpass the historical peak investment share of the telecommunications (dot-com) boom and many other industrial revolutions.
Historical Comparison of Investment Scale
Current investments in AI and its supporting infrastructure (data centers and energy) are significantly larger in absolute terms and faster in velocity than the early years of aviation, computing, or the automotive industry, though they have yet to reach the historical peak of the railroad boom as a percentage of GDP.
Railroads (The Peak Benchmark): Railroad investment in the late 19th century remains the gold standard for "general purpose" infrastructure booms, peaking at 3.5% to 6.0% of U.S. GDP. Some estimates even place its long-term total at 20% of GDP over decades. Currently, AI capital expenditures (capex) are roughly 20% of that railroad peak.
Aviation & Automotive: In their first decades, these industries were highly fragmented and lacked the concentrated, massive capital injections seen today from "hyperscale" tech companies. For example, the aviation industry's current projected IT spend of $46 billion by 2026 is dwarfed by the $490 billion to be spent on AI infrastructure in 2026 alone.
The Computer Revolution: Current AI investment has already surpassed the peak of the dot-com bubble and 5G/telecom spending (which peaked at ~1.5% of GDP). One analysis notes that AI leader Nvidia is capturing the highest share of market-wide capital spending since IBM's peak in 1969 at the dawn of the mainframe age.
Unique Characteristics of the AI Investment Cycle
Concentrated Capital: Unlike past revolutions led by thousands of small players, five large cloud companies (Amazon, Alphabet, Microsoft, Meta, and Oracle) are projected to invest $602 billion in 2026, with roughly 75% going directly into AI infrastructure.
Speed of Impact: AI-related capex is already contributing significantly to national numbers, accounting for roughly 40% of U.S. GDP growth in 2025/2026. In past cycles, like the 1990s tech boom or early 20th-century electrification, productivity gains typically took 4–5 years to meaningfully show up in GDP data.
Physical Infrastructure Demand: The current cycle is unusual because digital AI software is driving an immediate, massive demand for physical infrastructure. Some individual data centers now under construction have "city-scale" power demands (up to 1 GW), a rate of industrial growth not seen in decades.
Comparison with Past Industries
Historically, the most comparable investment booms to the current AI surge in terms of their share of the U.S. economy were the railroads in the 19th century, which represented a much higher percentage of the GDP, and the late 1990s IT/Internet boom, which AI is approaching.
AI (2026) vs. Past Industries
Railroads (mid-19th Century)
3.5% – 6.0% Peak Investment (Share of U.S. GDP)
Still the historical benchmark, but AI is rapidly approaching this level.
Electrification (late 19th-early 20th Century)
Unquantified but massive private capital share
Widely regarded as on par with railways in scale
Automotive (early 20th Century)
Up to 4.9% Peak Investment (Share of U.S. GDP)
AI's investment velocity is much faster due to highly concentrated capital sources (Big Tech). Higher in overall share of output, but in a different investment era.
Computers (1960s)
Data not directly comparable
AI chip share matches IBM's 1969 peak
IT/Computers (1990s)
~1.5% Peak Investment (Share of U.S. GDP) IT/telecom combined
AI investment has already surpassed the telecom peak, but not necessarily the overall IT boom share yet
Telecommunications (peak 2000)
~1.2% Peak Investment (Share of U.S. GDP)
AI data center spending has already matched this level and is forecast to grow significantly beyond it. Comparable to current AI infrastructure
Current AI Infrastructure (2026)
~1.2% - 1.7% (Share of U.S. GDP)
Expected to grow rapidly toward the 5.6% mark by 2030, potentially exceeding most past booms.
The main difference lies in the speed and concentration of capital flow. A handful of large tech companies are pouring hundreds of billions into AI infrastructure annually, a dynamic not seen in earlier, more fragmented industrial revolutions.
Railroads as the Benchmark: The railroad boom of the 1880s is widely considered the largest single industry investment wave in U.S. history relative to the size of the overall economy at the time.
Speed of the AI Boom: While AI's share of GDP is not yet at the railroad peak, its velocity is unprecedented. The current pace of capital expenditures for data centers and AI is set to exceed the telecommunications buildout of the early 2000s.
AI Growth is Not Limited to One Industry
While AI is the primary driver of current U.S. capital investment, the Energy and Semiconductor manufacturing sectors are receiving comparable or even larger long-term financial commitments to support the AI infrastructure buildout.
Industry-wide investment trends for 2025 and 2026 show:
1. Chipmaking (Semiconductors)
Total Commitment: As of late 2025, private sector commitments to U.S. chipmaking exceed $500 billion.
Scale: This effort aims to triple domestic production capacity by 2032, fueled by both the CHIPPS Act and private demand for high-performance processors.
2. Energy and Utilities
Rapid Growth: The utilities sector is currently the fastest-growing area for capital expenditure, as providers race to modernize the grid and add generation capacity to meet massive power demands from new data centers.
Renewables: In 2025, clean energy stocks significantly outperformed the "Magnificent Seven" tech stocks, returning 47% compared to the tech group's 25% average.
New Capacity: Approximately 88% of new electrical generating capacity in the first eight months of 2025 came from wind and solar.
3. Data Centers and AI Infrastructure
Concentrated Spending: U.S. data center investment alone is projected to approach $500 billion in 2026.
"Hyperscaler" Lead: Large cloud providers (Meta, Alphabet, Microsoft, Amazon, and Oracle) are projected to allocate $342 billion to capital expenditures in 2025—a 62% increase over previous years.
4. Advanced Manufacturing and Reshoring
Shift in Focus: There is building momentum in industrials, specifically in aerospace, defense, and manufacturing reshoring, supported by new trade policies and supply-chain realignment.
Investment Intent: Roughly 80% of manufacturing executives plan to invest at least 20% of their improvement budgets into "smart manufacturing" (automation and cloud analytics) throughout 2026.
Summary of Scale: While AI categories represented roughly 37% of real GDP growth in the first nine months of 2025, the physical infrastructure—power plants, the electrical grid, and semiconductor fabs—is now attracting the massive, multi-hundred-billion-dollar investments necessary for the next phase of the digital economy.
The AI Economy Has Arrived
Taken together, the scale, speed, and concentration of investment make it clear that AI is not a speculative bubble but a once-in-a-century industrial transformation. Unlike past revolutions, AI is simultaneously a digital productivity engine and a catalyst for massive physical infrastructure buildouts across energy, semiconductors, data centers, and advanced manufacturing. While its current share of GDP has not yet reached historical peaks set by railroads or electrification, its unprecedented velocity suggests it may surpass most prior booms within the decade. As capital, talent, and infrastructure continue to converge around AI, it is increasingly evident that artificial intelligence will define economic growth, industrial strategy, and competitive advantage throughout the 21st century—much as railroads and electricity did in eras past.