I work at a NASA research lab, and find it a rewarding way to spend my time... I've seen exoplanets through the eyes of space telescopes. I've invented AI algorithms and then flown them on smart satellites. My code has run on a rover traversing the surface of Mars. I agree that commercial enterprise has a role to play - but for all its imperfections, NASA is still a pretty remarkable institution at this particular moment in human history.

Point taken, but if science is our goal then our performance metric should be discoveries achieved per dollar spent.

The Mars Exploration Rover mission cost less than $1 billion total. In contemporary dollars the Apollo program cost $150-200 billion (and going to Mars would be WAY tougher than the Moon). Imagine - the price of a human mission we could fill the solar system with squadrons of rovers. The numbers are rough, but they suggest that we can get more science for our buck with robots.

The big reason for 3d virtual worlds in business is the same as for gaming: They have the potential to be a richer, more natural and engaging user experience. Imagine trying to sustain participants' attention in a 30-minute SMS business meeting.