Why the AI Boom Can’t Be Compared to the Dotcom Bubble

Why It Matters

The shift from speculative over‑building to demand‑driven investment signals sustainable growth for AI, but concentrated capital and compressed software multiples introduce new vulnerability points for investors.

Key Takeaways

• AI revenues exceed $20 bn annually, showing real demand.
• 90% of new data‑center capacity pre‑committed before build.
• Hyperscalers to spend ~$700 bn on infrastructure in 2026.
• ChatGPT reached ~1 billion MAUs, dwarfing early internet.
• Software valuation multiples fell to ~4×, lowest in years.

Pulse Analysis

The dot‑com era was defined by cheap, speculative fiber over‑build, leaving telecoms with under‑utilized assets and debt‑laden balance sheets. Today’s AI surge is anchored in tangible demand: enterprises are committing to data‑center capacity before construction, and power‑intensive facilities cannot be expanded without firm customer contracts. This fundamental shift reduces the likelihood of a classic over‑capacity crash and aligns capital deployment with actual usage patterns.

Infrastructure spending underscores the new reality. Hyperscalers such as Microsoft, Amazon, and Google are projected to pour roughly $700 bn into servers, networking, and power systems in 2026, a figure pulled forward by contracts from AI‑driven workloads. Meanwhile, AI leaders like OpenAI and Anthropic have crossed the $20 bn annual recurring revenue threshold, demonstrating that the market is monetizing advanced models at scale. The combination of pre‑committed capacity and robust revenue streams creates a feedback loop that fuels further investment while keeping utilization rates high.

For investors, the narrative is nuanced. Software valuation multiples have slipped to about four times earnings, reflecting market caution after the “SaaSpocalypse” and signaling tighter pricing discipline. Yet, capital is now concentrated among a handful of globally scaled, profitable AI and infrastructure firms, concentrating risk if spending slows or consolidation accelerates. Understanding these dynamics—real demand, physical constraints, and valuation compression—is essential for navigating the AI landscape as it matures beyond hype into a sustainable growth engine.