Building on Quicksand

The accelerating deployment of AI reveals three critical challenges: our infrastructure is inadequate, monopolization threatens resilience, and we're eliminating entry-level career pathways. But the same maintenance backlogs that companies defer also represent an overlooked opportunity—if we choose to use AI to address them rather than eliminate the workers who would learn from them.

We're deploying artificial intelligence at breakneck speed, but we're doing it on infrastructure designed for a different era—one that prioritized velocity over security, market capture over resilience. As AI systems become deeply embedded in critical operations worldwide, three uncomfortable truths emerge: our foundational infrastructure is dangerously inadequate, corporate concentration is creating systemic risk, and we're eliminating the very entry points that build organizational knowledge.

The path forward isn't predetermined. It requires confronting these realities now, before the decisions we're deferring become crises we can't contain. And it offers concrete actions for those already using AI—choices that shape whether we build toward resilience or fragility, distribution or concentration.

This article draws on research current as of January 2026.