Understanding AI Datacentres: Are We Facing a Telecom Crash Repeat?
As the AI datacentre boom unfolds, comparisons to the 2000s telecom crash abound. However, the fundamentals reveal critical differences that could reshape our perspective.
Key Comparisons:
-
Supply Dynamics:
- Telecoms: Exponential capacity increases led to massive overcapacity.
- AI: Slowing supply improvements caution against underestimating future demand.
-
Demand Patterns:
- Telecoms: Demand was grossly overestimated by 4x.
- AI: Rapidly growing demand, especially from agent adoption, may still be underestimated.
-
Investment Trends:
- Datacentre CapEx is expected to balloon from $127B in 2023 to over $255B by 2025. This cyclical growth reflects robust AI activity rather than mere hype.
The Real Risk: Timing over direction. Should businesses adopt AI agents slower than anticipated, overcapacity could lead to financial strain, but this differs from obsolete infrastructure.
Let’s Explore Together! Think critically about these trends and engage with your thoughts below. And if this summary intrigues you, share it with your network to foster discussion!
