$20 plus operating costs. That is, the $20 is just the additional cost to pay off the capital cost over $20 years, and doesn't include any of the amount necessary to pay the operating costs.

Airline tickets would be pretty cheap if all they were paying was the 20-year amortized capital cost of the airplanes + airports, too.

$20 is the unsubsidized 20-year-amortized cost of building the thing, there would be additional costs to run it, not the least of which being insurance.

That's the estimated cost to build it. I'd guess double it for operating expenses - maintenance, security personnel, etc.

Well it has to beat a standard commuter car (~$45 one way@34mpg) by a significant margin to be useful for consumers.

What about freight?

Time is money too. The drive from SF to LA is about 5.5 hours (per Google Maps). 5.5 hours of stress and frustration, compared to 30-40 minutes of hyperloop...

Just as an anecdote.. We had some friends from Orange County come visit us in SF this past weekend.

* OC -> SF -- Left at 3pm, took 9.5hrs to drive * SF -> OC -- Left at 9pm, took 7 hours to drive

There is random night construction as well as detours and traffic to worry about on I-5. Half an hour would be a no-brainer, even if it only went so far as downtown LA.

> The drive from SF to LA is about 5.5 hours (per Google Maps).

If I tried to do it in 5.5 hours in the real world, with traffic, I'm pretty sure I'd get arrested at some point.

Up to 8 hours when there's construction on I-5, jams into your destination city, or when you're taking kids who need bathroom breaks.

good point. Also my estimate is the marginal cost for a car vs the ammortized total cost for the hyperloop.

It's kind of cool that Elon Musk is competing with himself here -- Hyperloop vs. a 3rd generation Tesla (which should be even cheaper than $45 one-way per car, including depreciation and electricity.)

The real problem, even once you add self-driving cars, is I-5's limited capacity.

Isn't the real problem that it sucks to drive 7 hours?

Self driving cars seem plausible, especially for interstates, in 10 years. Technically, in 3-5.