The global AI infrastructure build-out is rapidly becoming the defining megaproject of this decade. But as momentum accelerates, so do its contradictions. Reporting onsite from PTC’26 in sunny Honolulu, RCR’s James Blackman captures a growing reality: The AI economy is not lifting all parts of the digital ecosystem equally.

Instead, a widening gap is emerging between hyperscalers and AI-native companies racing ahead and the network operators and infrastructure providers struggling to keep pace. The imbalance is striking. AI workloads depend on dense fiber, resilient transport networks, and massive power availability — yet telecom operators continue to face chronic underinvestment and limited monetization opportunities. Capital is flowing aggressively into compute and software, while the connectivity layer that makes AI possible is increasingly treated as a utility rather than a strategic asset.

As Blackman writes, this dynamic risks creating a new kind of digital divide — between the industries making money and those that aren’t — that will leave the broader AI ecosystem structurally unbalanced and telcos struggling “for a share of the pie.”