We are self-healing, regenerating, low-power, versatile, autonomous, and most of us have a pretty decent array of sensors built-in, along with some communications equipment that's capable of interpreting the signals from our sensors and transmitting that information to other humans in a remarkable variety of ways. All of these are approximate and relative of course, if someone replies with e.g., "but actually we're not as low power as...", it will be easy to ignore.

Specialized machines can do things humans can't, of course. No single human could have survived as long in the Martian environment as any of the rovers have.

But nobody has yet designed a machine that can do all the things humans can do.

Take the single problem of mobility: many very smart engineers have worked together to develop a set of wheels that can usually move the rovers around their environment without getting stuck or damaged, or at least have a chance of getting unstuck. A human that hasn't climbed a set of stairs in a decade can still outpace the rovers, and do so over more varied terrain, and with less chance of getting stuck.

So, yes, from an engineering point of view, building new robots that can do things and shipping them to Mars to do those things presents a lot of very interesting technical challenges to solve. It's all endless puzzles and little unsung feats of science and engineering -- assuming there is a country left with both the will and the resources and the talent to pursue such things.

But from a human exploration perspective -- our instinctive drive, or compulsion, or whatever it is, that has spread our species across the entire planet -- no machine will ever quite satisfy the desire to have that experience with the sensors we were born with.

My enthusiasm for a human mission to Mars has waned quite a bit in the last few years, largely owing to its most vocal advocate. Still, all the same, I think we should acknowledge that robots are poor substitutes for geologists.