Thank you very much for posting this.
This article is number 12 in the Fallows series.
At the point of time you posted this, many might have been thinking "Does the California High-Speed Rail series of articles by James Fallows ever come to an end?",

but then again, many might think it's good to have some ongoing public debate about the project and its different aspects.
It seems like a few hours ago, article 13 was published:
http://www.theatlantic.com/business/archive/2014/10/california-high-speed-rail-lucky-no-13-lets-look-at-maglev/381638/?single_page=true
It is about alternatives to high-speed rail like self-driving cars and magnetic levitation trains. Some might be really happy to read the sentence "Elon Musk's 'Hyperloop' transport vision [...] is still hypothetical enough that I'll leave it for another time" - so that smokescreen anti-high-speed-rail campaign is finally seen as something similar to what it is.
Seeing Maglev as at least a theoretical alternative to high-speed rail, many might be able to understand that, still many might be surprised to see the theme of "self-driving cars are an alternative to high-speed rail" be picked up here - especially since self-driving cars might also still be "hypothetical enough" for many. Many might realize that the different "assisted drive" features offered today don't equal "self-driving". Most of all, many might think that self-driving cars in no way would even compete with with high-speed rail - one would have to speak of cars in general competing with high-speed rail, still many might realize that it doesn't make a difference if they are "self-driving" (which is likely not going to happen anytime soon out of liability issues, and "assisted drive" does not equal "self-driving") or not.
Despite the claim by the anonymous reader of James Fallows from the "advanced-research parts of the info-tech industry" that "Self-driving cars can be faster on highways because they can caravan", this seems to be absolutely counter-factual, as there is no indication at all yet that speed limits will be lifted to allow "self-driving" cars to speed by all the "slow" cars that are not "self-driving".
California high-speed rail will connect Los Angeles and San Francisco in less than 3 hours - will cars that are not "self-driving" connect Los Angeles and San Francisco in less than 3 hours? No, they won't. Will "self-driving" cars connect Los Angeles and San Francisco in less than 3 hours? No, they won't. Driving the distance, it is likely it will take double that time. One could go and mention other examples here, Fresno to San Jose, Bakersfield to Los Angeles etc., still the answer remains the same.
California high-speed rail will also be used be commuters a lot. To just cite one example, it will connect Palmdale to Los Angeles in less than 30 minutes. Will cars that are not "self-driving" connect Palmdale and Los Angeles within 30 minutes? No, they won't, it's likely to be double that time. Will "self-driving" cars connect Palmdale and Los Angeles within 30 minutes? No, they won't, it's likely to be double that time. Will cars that are not "self-driving" be stuck in traffic on US-101, CA-170 or I-5? Yes, they will be. Will "self-driving" cars be stuck in traffic on US-101, CA-170 or I-5? Yes, they will be. So to some it seems obvious that it doesn't matter if cars are "self-driving" or not. Many might think, it's still just cars. And many might think, cars with all the negative effects and externalities immanent to them.