I want to be upfront that I went into this conversation with a chip on my shoulder about the existing coverage. The truck driver angle gets pulled out constantly in pieces about automation — as shorthand, almost. "And what about the three million truck drivers?" It's usually a rhetorical question. The number gets used, the drivers don't.
This person drives long-haul routes across the Midwest and has been doing it for about fourteen years. He agreed to talk partly, he said, because he was tired of reading articles where truckers were "the example" but nobody had actually asked any of them what they thought.
His sentence:
"The robot trucks aren't coming for my routes yet, but the routing software already changed the job more than anything else in the past decade."
He wanted to unpack that. The highway autonomy stuff — the self-driving long-haul pilots — is real, he said, and he's not dismissing it. But it's not what's changed his work week. What's changed his work week is that dispatchers now use AI-assisted routing that is genuinely better than the old system at optimizing for fuel and time. Which sounds like a win, and in some ways it is.
But the old system had slack in it. Routes got built around driver knowledge — who knew which weigh stations were slow on certain days, which rest stops had decent food, where you could make up time. That knowledge lived in the drivers. The new routing assumes optimal conditions and builds schedules accordingly. When the optimal conditions don't materialize, which is often, drivers eat the difference. They're making the same decisions they always made; they're just making them against a schedule that wasn't built with their experience in mind.
He said this without a lot of bitterness. More like someone explaining a system to someone who doesn't know it well, which I appreciated. He wasn't saying the software was bad. He was saying it was designed by people who didn't know the roads.
On the autonomy question, he was measured in a way I didn't expect. "It'll do the highway miles eventually," he said. "The terminal work, the local routes, the backing into docks in bad weather — that's a different thing. I'm not sure they're as close as the coverage makes it sound." He said he was watching it. He wasn't moving his kids across the country over it right now.
What I came away with: the story about drivers and automation is not primarily a story about self-driving trucks. It's a story about software that's already here, already mediocre in specific ways, already shifting whose expertise counts. That's less dramatic than "three million jobs gone." But it's more accurate, and it's the story the drivers I've talked to are actually living.