Deep Dive: How AI Infrastructure Spending Is Reshaping Big Tech Valuations — And What It Means for Investors
The largest capital expenditure boom in corporate history is underway, and it is being driven by a single technology: artificial intelligence. In 2025 alone, four companies — Microsoft, Alphabet, Meta, and Amazon — spent a combined $357.5 billion on capital expenditures, more than 2.5 times the $140.4 billion they spent just two years earlier. That spending surge, concentrated on data centres, GPU clusters, and networking infrastructure, is fundamentally reshaping how investors value the biggest companies on the planet. The stakes are enormous. NVIDIA, the primary beneficiary of this spending wave, has grown into a $4.6 trillion company — the world's most valuable — on the strength of AI chip demand that shows no sign of slowing. Yet the companies writing those checks face a harder question: will the revenue generated by AI models, cloud services, and enterprise tools justify hundreds of billions in upfront investment? OpenAI recently told investors it expects industry-wide compute spending to reach $600 billion by 2030, a figure that would dwarf even the dot-com era's infrastructure build-out. For investors, understanding this capex cycle is essential. It determines which companies are building durable competitive advantages, which are destroying free cash flow, and whether today's valuations reflect rational expectations or speculative excess. This explainer breaks down the numbers, maps historical parallels, and examines what the data actually shows about big tech's AI bet.