"Robot spacecraft bring back far more information per dollar spent."

There is no reason why both of those statements can't be simultaneously true.

The trick is to find the optimum combination of the two. I believe that
robot spacecraft are the most logical and cost-effective way of making manned space exploration (and eventually, exploitation and and even colonization) inevitable. Sooner or later, robots are going to find a really good reason why we need to send people, and when they do, we had better have the technology ready to do so. We probably won't have the time to develop it from scratch.

The two concepts should not be self-exclusive, and the debate about which is "better" is damaging to both. My own opinion is we should complement a strong science-based robot program with continued orbital stations as a technology testbed and as a precursor to planetary exploration when we are ready, and when we actually have a good reason to go there. I believe manned missions to the planets are a bit premature, and to the moon redundant, and both too expensive for any payoff we can currently anticipate.

I also think our most effective long term project would be a space station in an orbit designed primarily as an assembly point and refueling port for staging elaborate robot missions into the asteroid belt and the outer solar system. When we find the justification, this station could easily be converted to a real spaceport for manned planetary missions. The problem with all our space enterprises so far is that each has been conceived as a demonstration or proof-of-concept. After a few years they have been abandoned as obsolete. What we need is something we can build on and expand for future use.

History is filled with examples of old technologies that are still providing useful service because they were designed with long-term utility in mind. Anyone who's ever been through the Panama Canal knows that good engineering can have a VERY long and useful life span, IF it is part of a long-term strategy. Soviet-era spacecraft come to mind too, they may be old and clunky, but they are reliable and are still providing useful service. We need to design our major space projects so they fit into an overall, long-term plan for space exploration. Instead, we put together flash-in-the-pan applications about which just a few years later we ask ourselves why we even bothered.