Part of it is that guys who get into trucking do so because they don't want an office job with all the office stuff that goes along with it. Then all the technology that they do have to deal with is pushed on them, so they don't want to go in any further than they absolutely need to, like hiring an IT guy or heaven forbid an IT department. Most of their custom software ends up being written in Access or Excel by some hotshot kid they had working on the loading dock two years ago.
Someday in the next five to ten years the big companies (Walmart, Amazon, UPS) they will get there. They're going to have autonomous trucks running point to point routes between distribution centers, particularly on open highway areas in the plains states. It will have to come from those places because of the initial costs in developing self-driving software and handling the legal labyrinth of getting permission to put those trucks on the road is immense, and only the big companies will have the infrastructure to support the planning software needed to link all of those logistics moves together.
Trucking companies will still have a need for drivers for years on out past this, but when they do start losing market share to the robots, they probably won't be in position to do anything about it but close up shop.
My understanding is that AI is limited, period. There is no amount of money or research that will develop a thinking machine, ever. That is my understanding of what they are finding out about AI.
All AI can do is take over mundane tasks and that it can save industry billions of dollars in eliminating human help and allow people to do more. If a human, on a long stretch of the road, can catch a short nap and be awoken by the truck when human intelligence is required, that alone can eliminate millions of jobs. That’s pretty valuable to the industry.