[rflrob]

The problem is, humans are so much more effective than robots at planning and independent decision making, which is critical when there's a 8-40 minute communications round trip time. The two wildly successful MER probes Spirit and Opportunity traveled a combined ~40km in around 4,000 sols. A human areologist could cover that much in a week.

[mechanical_fish]

And just how many Spirit probes could one send to Mars for the cost of sending a single human?

Based on a bit of Googling, the answer seems to be on the order of "ten", as a minimum. Numbers of $20 billion or so were being thrown around for a manned mission, and the currently-in-transit Mars Science Lab is up past $1.5 billion after a 30% cost overrun:

http://www.aviationweek.com/aw/generic/story.jsp?id=news/Bal...

Given that the $20 billion is a pie-in-the-sky planning number and has yet to experience overruns of its own...

Meanwhile, what's the rush? Mars is not going anywhere.

And meanwhile too, is your human really going to cover 40km in a week after a ten month trip in free fall? Not at the present state of the art:

http://www.space.com/8978-trip-mars-turn-astronauts-weakling...

Or are we shipping an SUV to Mars along with the human and all of her groceries? The Spirit-probe-to-human ratio continues to grow.

[ericd]

It's not something that can just be parallelized away with identical probes. It's the difference between a multifunction calculator and a PC. The PC may weigh 100x as much, but it has an infinitely greater breadth of capability. That's why human spaceflight is going to be important for the foreseeable future. Unless you want to design a separate probe for every conceivable task (some of which we undoubtedly don't know yet).