TECH Signal 489
Americans are rallying against data centers. Surprisingly few are getting built
↗
Public opposition and severe supply chain bottlenecks are stalling AI data center construction, with a large gap between announced projects and those likely to be completed.
For engineers, the lag means AI compute capacity will remain constrained longer than demand curves suggest, affecting deployment timelines and cost models. Infrastructure planning must now account for multi-year delays in power, materials, and labor rather than assuming rapid scale-out.
Written by elseif from the cluster below · every claim links back to a sourceThe three things worth knowing
Only about half of scheduled AI computing capacity through 2028 is expected online on time, down from a historical 72% on-time rate.
Roughly 60% of capacity planned for 2027 has not started construction, and 7% of underway projects are already delayed.
Two-thirds of the 565 GW planned pipeline is considered implausible, with materials, chips, power, and labor shortages compounding delays.
THE CLUSTER