Signal Over Noise: What US History Tells Us About AI Capex

This report analyzes the current AI capital expenditure cycle by drawing lessons from historical innovation cycles like railroads and the internet boom. It concludes that while competitive overspending is present, AI infrastructure remains an attractive investment opportunity.

Key Takeaways

• 1.AI capital expenditure cycles tend to follow historical patterns of innovation cycles, including phases of euphoria, overcapacity, and potential market consolidation.
• 2.Competitive pressure in the AI sector currently resembles historical cycles like the 19th-century railroads and the 1990s internet boom, where companies overspend to prevent rivals from gaining an advantage.
• 3.UBS maintains a positive outlook on AI infrastructure investments, predicting AI CapEx will reach 3% of US GDP next year, supported by more durable infrastructure and elastic demand from autonomous agents.

Table of Contents

• CapEx lesson #1—CapEx cycles rhyme
• Lesson #2—Competitive necessity drives overspending
• Lesson #3—Lasting wealth creation happens in the application layer

Document Preview

Page 1 of 5

Subscribe for full accessStart Free Trial