The good news is that they'll still be making money a century from now, and there won't be any more competitors than there are now.

Mild disagreement. Self driving 18 wheelers, if successful at scale, will take some market share, depending on price and reliability. They won’t handle the really huge volumes that a railroad can, and there are certain segments that just don’t work as well (grains, coal, etc.) but for many commercial products, especially those going to retail, it’s a simple end-to-end solution rather than a load on the car, load off the car, transfer and drive to the distribution center, or even store to store within a distribution area.

It will all depend on relative price, of course, and it’s possible that self-driving trucks (whether diesel or electric) could be faster by eliminating a couple steps (and waits) in the distribution process. That could help some manufacturers where time-to-market is relevant to stay closer to trends and react more quickly. Maybe like fast-fashion? I’d guess some “really fast” bypass railroads altogether (iPhones, other tech) and it’s not likely they could get those in sufficient numbers to move the needle even with precision railroading or other improvements, but I’d hazard an opinion that there’s some level of product that could be vulnerable.

And self-driving trucks, while expensive, require nowhere near the capital investment that “starting a new railroad” (practicably impossible) or even retrofitting a current one would take. That new segment is pretty much a ‘buy one at a time’ and build out the network as you wish kind of thing. I expect it to grow pretty well once it proves itself.

Other than that, they are desperately vulnerable to the overall economy (isn’t everyone?) and if it’s slowing, it’s hard to imagine them ramping up profits, unless there are gross inefficiencies already that I’m not aware of.