AI infrastructure is being built faster than near-term demand can absorb, echoing the late-1990s fiber overbuild, meaning today’s speculative rush in data centers and GPUs may ultimately pave the way for a later wave of AI applications that benefit from cheaper, abundant compute.
This article comes at the perfect time, and I really appreciated the insightful parallel with the 'dark fiber' era. Do you think future innovators will similary capitalize on this 'compute glut'?
This article comes at the perfect time, and I really appreciated the insightful parallel with the 'dark fiber' era. Do you think future innovators will similary capitalize on this 'compute glut'?