Rockets are a lousy way of getting stuff into space, because most of what they lift is their own fuel. We're still using rockets because we don't have a space elevator yet, but an innovative interim solution could be a maglev vacuum tube that launches cargo up the side of a mountain and straight into low Earth orbit for cheap.
NASA has been looking at railgun and maglev launching systems for a while now, but the solution that a company called StarTram is offering doesn't need any rockets or turbojets or scramjets or anything like that. Instead, StarTram proposes to construct an 80 mile long tunnel, the end of which follows the curve of a mountain up to the peak. The inside of the tunnel would be held at a vacuum, and magnetically levitated cargo spacecraft would be fired down the tunnel at about 20,000 mph. They'd exit at the top of the mountain through a plasma window and continue up into space with 35 tons of payload on board, requiring only one tiny little correction with on-board rockets to circularize their orbit.
It would certainly cost a whole bunch of money to set up a system like this: estimates put it as something on the order of $20 billion over ten years, and that's if we really push to make it happen. But once it's up and running and launching 35 tons of cargo ten times per day, the cost of getting something into low Earth orbit would be just $20 per pound, which is hundreds (or thousands) of times cheaper than it costs now.
This generation of StarTram wouldn't be able to launch humans (since the cargo capsules would be subjected to an eye-pancaking 30gs of acceleration), but if we want to think bigger (much bigger), it would be possible to scale up to something human rated. To make up for a tolerable level of acceleration (2-3gs), the vacuum tunnel would have to be much, much longer (1,000 miles or so), and it would have to end 14 miles above sea level to be above enough of the atmosphere such that you wouldn't die of deceleration when transitioning out of the vacuum tube. To keep the tube up there, the idea would be to charge ground tethering cables up with 280 mega-amps of electricity, and the tube itself with 14 mega-amps flowing in the opposite direction, creating four tons per meter of repulsive force and causing the tube to strain upwards against the tethering cables. The estimated cost of something like this? $67 billion.
StarTram is certainly not cheap, but it has the advantage of being more or less ready to go in terms of the technology required: there's nothing ultra-futuristic about it, and no major advances that would have to be made. Given the funding, we really could build something like this in a decade, and it would enable so much stuff to be sent into orbit so inexpensively that things like orbital solar power stations might at last be cost-effective, sending as much clean energy down to Earth as we could possibly want.