[chrisboesing]

He talked about it a little bit in the video mentioned in the post. He thinks about "open sourcing"(his words) or patenting the idea and giving the patent to someone who has the financial resources to build it. I guess because of that he hasn't gone into much detail but here is a list of things about the Hyperloop I remembered:

-from downtown LA to downtown SF in 30 Minutes

-cheaper than airplane ticket

-solar panels on the top to make it self sustaining

-energy would be saved without batteries to run at night

-under no influence of the weather

-safe

(-Sarah Lacy made a reference to a TV show/movie with a tube as transportation system, and Elon said it was kind of like that [can't remember the title])

-Sarah Lacy: "Is it possible?" Elon Musk: "Yes!"

[stcredzero]

That could be a non-launch version of the Lofstrom Loop.

http://en.wikipedia.org/wiki/Launch_loop

Because it's not subject to the rocket equation, it would basically have the same effect on space access costs as a space elevator.

It is possible in principle for such a device to bring a vehicle up to hypersonic speeds, then regeneratively brake the same vehicle to a stop and recover a fraction of the energy.

[rwhitman]

I think this is much more likely to be what he's talking about than pneumatic tubes, as others have mentioned. Seems much more up his alley

[ableal]

Good reading there, but the part about the difficulties notes that it holds nuke-scale energies and states "Therefore for safety and astrodynamic reasons, launch loops are intended to be installed over an ocean near the equator, well away from habitation."

That would exclude the downtown-to-downtown feature.

[maxerickson]

If one vehicle is leaving when one is arriving, you have a nice place to put all that energy.

I guess if the transit is efficient enough, you store energy by getting a bunch of mass moving around the loop.

[dojomouse]

For what it's worth I think the energy related claims are pretty weak. Technically you can store energy in the kinetic energy of the mass moving at speed, and with diamagnetic ('quantum') levitation the losses are very low... however the costs, overall efficiencies, and dynamics make it barely interesting as a theoretical exercise, and totally worthless as a practical one.

Powering with PV on top of the tube, on the other hand, should be entirely realistic (at least on an average consumption vs average generation basis), and the incremental PV cost will be really low since you'd install it as part and parcel of the tube.