Think of it this way.
Put our best EARTH robots down on a selected fossil rich piece of land, say N Dakota. Have them find dinosaur fossils. Still waiting? Or instead send a trained paleontoligist or 2. Which way gets you confirmation of past life within any reasonable timeframe?

Robots are great. We should fly as manyrobot missions as possible for the next decade or two. But I contend that despite all the AI (REAL SOON NOW) BS, trained human explorers are best. A VASIMR high ISP spacecraft reduces mission flight times to tolerable levels of a few months. This is very doable in the 2020s technologically, and achievable a decade or so later, financially.

I think odds are good that there's existant sub-surface life today on Mars. If there isn't, I'd start to wonder just how unique Earth is...