Exactly.
There's another possibility that nobody really likes to think about: automated trucking won't even start for decades or more. The railway industry is a good example of this. Trains run on dedicated pathways with hardly any traffic, they're the perfect example of a system that should be entirely computer controllable.
Yet the world still has many, many train drivers. Computers could have and arguably should have replaced their jobs long before we even started thinking about the infinitely harder problem of cars and trucking, but outside of a few well known projects it hasn't happened because:
1) There are very few computer programmers in the world who can develop and lead ultra-high reliability automated driving software projects.
2) Of those programmers, many aren't interested in rail projects, partly because so many programmers are in the USA and Americans don't use rail much.
3) The companies that need to buy the tech tend to only have civil engineering expertise at the top and no software engineering expertise, making it impossible for them to select the best providers in the market, meaning the market doesn't force evolution towards better technology and project failures are frequent (markets need informed customers and there aren't enough informed customers in the tiny railway industry for this to work well).
4) Drivers unions are very powerful and railways are often state owned or at least heavily regulated, so there's little commercial force of will to pick fights with the workforce.
The London Underground is a good example of these dynamics in action.
Everyone here is assuming that the trucking industry will immediately adopt auto-driving technology if it's available. But that's making some big assumptions.