There's some recent speculation on the web about the puzzling problem of why we can't seem to spot any alien civilizations.  The latest buzz surrounds the idea of spotting Dyson spheres in our galaxy or elsewhere.  Bruce Dorminey sparked the discussion with a piece on physicsworld.com, and others such as Paul Gilster over at Centauri Dreams have picked up on it.

I've written about the Fermi paradox before; in fact, my novel Permanence offered a new solution to the problem.  Astrophysicist Milan Cirkovic wrote a nice analysis of the ideas in that book for JBIS, and he and I have corresponded ever since. 

As part of our recent discussions, I wrote him a little note about the logic behind such monstrous engineering projects as the "Kardashev-II civilization," where a species decides to capture all the energy radiated by its sun, generally by building a giant Dyson sphere around it.  I think the idea's a perfect example of homocentrism, or more exactly the kind of techno-centrism that assumes that future civilizations will orient themselves around the same central issue as 20th century humanity (in this case energy use).  Here's my off-the-cuff comments to Milan about energy efficiency as it relates to the visibility of spacefaring civilizations:

Notes to Milan

I’ve been doing  a lot of consulting/writing about “green” technologies lately, and one idea that comes up a lot is the concept of ecosystem services.  An ecosystem service is something you get for free from nature, whose value can be directly calculated by estimating what it would cost for us to provide the service ourselves.  For instance, water treatment:  recently a greenbelt area was declared around Toronto, basically a crescent-shaped region where real estate and industrial development is banned.  A key reason for doing this was the discovery that these forested lands filter and treat the entire aquifer for the Toronto region.  If they were developed, much of the fresh water in the region would dry up.  We’d then have to import/produce fresh water ourselves, and the cost of doing that can be directly calculated, and compared to the financial benefits of developing the land.  It turns out that the land, left alone, provides a set of essential services more cheaply than we can provide them technologically.

Now in the realm of information processing, it turns out to be cheaper for many organisms to offload calculations into the natural world; cockroaches use a clever mechanism that’s directly tied in to air movement and shadow angle to directly cause leg movement (they scurry away when something swings at them).  This mechanism essentially bypasses the nervous system because that’s too slow.  A partial program is in general any algorithm where key steps in the algorithm are offloaded in this manner:  the classic example is (for Americans) how do you catch a pop-fly in baseball?  AI researchers used to think that it required a sophisticated internal model and some nasty differential equations solved by the nervous system; in fact, runners catch a ball by running backward while keeping the ball at a fixed angle with respect to the horizon.  This combination of factors substitutes successfully for the calculation.

 Combining these two ideas, of ecosystem services and partial programs, we can propose an economic argument for the invisibility of advanced civilizations.  A settlement that uses solely ecosystem services is called a ‘zero footprint’ settlement (another word for sustainable).  Zero-footprint means environmentally neutral; it also means invisible to the mechanisms we usually use to detect the presence of technological activity (because our means for doing so generally involve detecting the waste products of systems running against or in parallel to natural processes).  In addition, a civilization that offloads as much of its data processing as possible into natural processes in the physical world, through partial programs, is more energy-efficient than one that builds "computronium" to do its thinking, and probably calculates faster (because the energy required by an algorithmic process and the speed with which it's executed are related).  The more such processes are substituted by integration with the natural world, the harder it will be for us to see the operations of that civilization from interstellar distances.  In fact, I would argue that a civilization that integrates efficiently with its environment on these two levels will be invisible by definition. 

A corollary to this is that colonizing other planets means moving into environments that provide few or no ecosystem services.  This implies that a spacefaring civilization is visible only in those places that do not provide such services; between its worlds, in other words.  Such a civilization’s visibility is then tied to its ability to directly adapt itself to alien environments including the environment of outer space itself.