I'd bet a large chunk of change that the HyperLoop is some variant of vacumn tube maglev. Not an idea Musk came up with by any stretch, but he's someone who might actually be able to make it happen which would be very awesome. His patent is going to hit a fair bit of prior art though.
Also think his dismissal of fuel cells is a bit disingenuous - Nissan apparently have them down to $50/kW (volume production estimate) which, given the low cost of energy storage ($/kWh), means they could be entirely viable at least as a range extension option.
But still, overall, admire the hell out of the guy. Legend.
He's already said it's not a vac tube (on twitter). I'm betting on some sort of launch loop (Lofstrom loop).
When he says there's a physics problem with fuel cells he's not being flippant. They're less efficient than batteries, have lower volumetric power density, and cars using off-the-shelf batteries already match the range of gasoline cars (so there's no energy density problem to solve). The only problem with batteries currently is cost, and fuel cells are no help there.
Pretty much the only attractive thing about fuel cells is that they can run on energy-dense liquid fuels like methanol, which can do greater power fluxes than last-mile grid. But battery swaps are even better. Ultimately, lithium and gasoline have about the same energy density, methanol half as much, and compressed hydrogen about 1/10 as much. And there are good reasons why hydride storage is unlikely to become practical in cars any time soon.
Also think his dismissal of fuel cells is a bit disingenuous - Nissan apparently have them down to $50/kW (volume production estimate) which, given the low cost of energy storage ($/kWh), means they could be entirely viable at least as a range extension option.
But still, overall, admire the hell out of the guy. Legend.