$NOK CEO Says AI Infrastructure Buildout Is Still in Its Early Stages

$NOK CEO Justin Hotard says demand for AI data center infrastructure remains strong and the broader buildout is still in its early stages.
According to Hotard, customers would be willing to expand capacity at roughly twice the current pace if memory and power constraints were less severe.
That suggests the main bottleneck is not necessarily demand.
Instead, access to electricity, memory and other infrastructure components is limiting how quickly new AI capacity can come online.
Hotard also argues that the investment cycle does not depend on constant releases of new frontier AI models.
Existing models can continue driving years of data center expansion as more companies deploy AI into production and inference workloads scale.
For $NOK, the thesis is that networking demand can remain strong even if model progress temporarily slows.
The bigger opportunity is the continued buildout of the physical infrastructure needed to connect and run increasingly large AI systems.