In any case, poor people who don't own cars rely on busses and subways--which this boondoggle does nothing for.
[1] $77 billion would also create a perpetual fund that would be able to buy a $10,000 used car for every household in poverty every 5 years.
Anyway, your last sentence is wrong. Much of the money under the HSR project is going to local transportation projects. So far they have spent or committed seven billion dollars for Caltrain, BART, Muni, Sacramento RT, Capitol Corridor, ACE, LA Metro, Metrolink, San Diego Blue Line, and Coaster, among others.
You can see this play out with the Acela on the east coast. It's almost all business travelers or nicely dressed professional looking people. People who can't afford $100+ per ticket take the bus ($35).
Likewise, even lower income people traveling alone aren't going to spend $86 for a ticket on HSR (if it's not more like $100+ by the time it's finally open). They'll pay $35 for a Bolt bus.
The supermajority of American families cannot even afford an unexpected $500 expense; they sure as hell aren't going to blow hundreds to drive to LA.
(Note: those poor people need to be able to do 30-40kg groceries, at a distance of, say 5-10 kilometers, when they're 50 years old)
So does building the same capacity in roads. And, even moreso, so does building that plus the extra capacity required for the increased load required to handle on-demand driverless vehicles.
Although the parent comment said "rail line", since TFA is, specifically, about HSR, that's a unique rail line in an existing system, which has something of an all-or-nothing construction problem. Having it half finished won't reduce the train travel time between LA and the SFBA for anyone.
Conversely, road capacity can be added incrementally, and, in many cases, when it is, it has an immediate effect for that section [1], providing relief as soon as the money and/or construction resources are available. In rural California, that's more likely to be months and fractions of a billion, not years and billions. The urban parts are slower and more expensive, but only proprotionally so (i.e. HSR is yet slower and even more expensive there, too).
You haven't presented any evidence for your assumption that on-demand driverless vehicles would require extra capacity, especially down a long-haul corridor. At best, that's some very heavy speculation on the operating cost (dominated by tires, suspension, electricty, I'd expect) versus capital cost of keeping enough spares in each metro.
[1] Yes, I'm aware that, at times, it can have a paradoxically negative effect on congestion, but I'm pretty sure that doesn't happen for capacity increases on what are essentially point-to-point corridors.
Driverless cars increase congestion, which increases cost of human travel, once the taxes needed to pay for road construction and maintenance are factored in.
Or taking a bus, no? Seems like you can get a trip for under $15.
The cost of HSR suddenly sounds so much more reasonable if you frame it in terms of the road budget. HSR's all-in cost for right-of-way, tracks, stations, and trains is about four years worth of California's current road budget.