AI Is Rewriting Freight. Here Is the Part That Touches Your Car

There is a great deal of noise about artificial intelligence in freight at the moment, and most of it is aimed at other freight companies rather than at anyone actually shipping something. Route optimisation, yard management, predictive maintenance -- genuinely useful, entirely invisible to you. If you are moving one car, none of it touches your life.

A smaller part of it does, and it is worth separating out, because it is the part that changes what you pay.

The useful application is matching. When you request a quote, the job is to work out which carriers are plausibly interested in your specific shipment -- this vehicle, this route, this timeframe -- and get it in front of them quickly. Historically that was a dispatcher with a phone and a memory for who ran what. That worked, but it scaled to however many numbers one person could dial in an afternoon, and it was only as good as that person's recollection.

“Software is very good at finding the carrier who wants your route. It is no good at all at inventing one who does not exist.”

Doing the same job in software means your shipment reaches a far wider pool in minutes rather than hours, weighted toward the carriers whose existing routes actually fit it. That is a real improvement and it shows up in a way you can measure: more carriers see your job, so more of them bid, and more bids on the same car means a wider spread to choose from.

What it cannot do -- and this is where the marketing gets dishonest -- is create capacity. If four trucks run your corridor this week, no amount of clever software turns that into nine. In a tight year, automation finds you the best of a small field faster. It does not enlarge the field. Anyone implying otherwise is selling you something.

Where the difference is largest is on the awkward jobs. A sedan going Los Angeles to Phoenix will find carriers regardless of how anyone looks for them; that lane is thick with trucks. A non-running pickup coming out of rural Montana with a two-day window is a genuinely hard matching problem, and it is exactly the kind of problem where searching thousands of carriers beats calling fifteen. The customers who benefit most from any of this are usually the ones who were hardest to serve before it existed.

One thing worth being properly sceptical about is instant pricing. A great many sites will now hand you a number the moment you type in two ZIP codes, and it feels like progress. Often it is not. A price generated without any carrier having agreed to it is a forecast, and forecasts get revised. The revision tends to arrive after you have committed, which is the worst possible moment. An estimate drawn from real completed shipments on your actual route is useful; treat it as the estimate it is, and judge the real cost by what carriers actually bid.

We would apply the same scepticism to the language around all of this, including ours. The question to ask any company leaning on the technology is simple and boring: how many carriers actually bid on my shipment? That number is checkable. It is the only thing in this entire subject that directly determines whether you got a good price, and it is conspicuously absent from most of the marketing.

The honest summary is that automation has made the middle of this business faster and more transparent, and left the physical part exactly where it was. Trucks still drive at the same speed. Drivers still need somewhere to sleep. Weather still closes passes. What has genuinely changed is that finding the right truck for an unusual job went from a phone-call problem to a search problem, and search problems are the thing computers are actually good at.

If you want to see it working rather than described, request a quote and count the bids. That number tells you more than any article about artificial intelligence, including this one.