NASA's moonshot  program  is in tatters, and the shuttle's due to fly for the last time soon ... but it doesn't mean there's no exciting space news. For example: Did you know NASA could send an android  to the moon inside just three years?
This crazy scheme even has a funky, mysterious title to go with its radical science and engineering: Project M. It's not funded at Agency level yet, let alone at the governmental level (from where dedicated funding would have to come) but it seems to be a highly mature project, with some engineering precedents already in place, including robotics and shuttle-derived rocketry.
The moonshot would work like this: NASA would fire aloft a modified existing rocket, making the most of the fact that the rocket wouldn't have to be human-rated, and no pesky (and heavy) supplies like oxygen, water, and food need be hurled moonwards to accompany the robot. This launch would place a small autonomously flying capsule to the moon, which would be propelled using green fuel (liquid methane and oxygen) and which would make an automatic landing. When all was safe, the capsule would pop open to reveal the robot, a humanoid-shaped walker with manipulator arms that are more or less analogous to human arms. The 'bot would be self-sufficient to some extent, but it will also be steered by Earth-based astronauts. The purpose is to test and refine the basic engineering issues that any future long-term lunar or Martian missions would face, in terms of construction. But there'd also be room for perfecting lunar mission management processes, performing opportunistic science with the benefit of more adept manipulators than typical rovers possess, and to do student-sourced experimentation.
You may even recognize the upper half of Project M's android: It's a development of the amazing Robonaut device that'll be being put through its paces aboard the International Space Station later this year. Its legs, like a more sophisticated version of Asimo, are already being developed by NASA independently, as the video below demonstrates.
For all sorts of reasons, this fast-track moon mission makes sense. Scientifically, there's a definite plus to be gained from lunar experiments. The engineering lessons that would be learned would be valuable for future manned  missions. There're even financial benefits, since unlike humans, Moondroid could actually be "parked" on the moon indefinitely. And there'd be a definite good-feeling PR spike (and associated effects like more kids interested in learning science) earned by doing something so very bold. So exciting. So damned sci-fi. In fact, thinking about it... isn't there some sort of space opera precedent in shooting a capsule to a dusty alien terrain, and watching a golden android trotting through the dusty sand?