Nokia CEO Says Supply Shortages Are Holding Back a Twice-as-Fast AI Buildout

AI data-center customers would probably build twice as fast if supplies allowed, Nokia CEO Justin Hotard told CNBC in an interview published October 5. Memory-chip and energy shortages are constraining the sector. His argument: the pace of construction reflects those limits, not how quickly customers would choose to add capacity without them.

Speaking on CNBC’s The Tech Download podcast, Hotard pushed back on the idea that the industry is already building too much. He pointed to customers’ appetite for faster construction as a reason for confidence that the AI infrastructure buildout remains in its early days.

Hotard’s twice-as-fast statement describes a hypothetical world without today’s supply constraints; it is not an announced construction target.

A case for deploying what already exists

Hotard also challenged the assumption that infrastructure demand depends on a steady stream of more advanced models. He described deployment as still being at an early stage and argued that the capabilities already available leave substantial work to do. In his view, growth need not wait for another breakthrough.

He put that argument in a concrete hypothetical: even if no new frontier model—the most advanced class of AI model—were released for three years, companies could probably make tremendous progress deploying today’s technology. That is a view about the remaining opportunity to use existing models, not a prediction that model releases will stop.