Here's a freight angle on the AI boom that doesn't get enough attention: AI doesn't build itself. Somebody has to haul the material. According to Trucking Dive, Landstar said that 9 of its top 100 customers are now tied to data centers — roughly 12% of its revenue.
Data centers are massive physical builds. They need steel, concrete, transformers, switchgear, cooling equipment, generators, and racks — much of it oversize, heavy, or specialized. Every one of those components rides on a truck before it powers a single model. As the build-out accelerates, that's a structural source of demand for flatbed, heavy haul, and specialized carriers.
For an operator, the lesson isn't to chase a buzzword. It's to notice where durable, repeatable freight is forming. The AI infrastructure cycle is early, capital-intensive, and concentrated in specific regions. Carriers and brokers who position around that build-out — the equipment, the lanes, the project freight — are reading demand a step ahead of the load board.