I didn’t post my thoughts on this, but TL;DR:
Why build planet-size anything when the real action is in the small things? Small spaces, small units of time, everything gets smaller.
I kind of agree that large is inefficient and unnecessary. Look at the history of computers: from giant to tiny and tinier. From slow to fast and faster. Personally, I have a feeling really advanced life drops all that dumb physical stuff that slows you down as soon as they can, and moves into the infinite spaces between:
This is, of course, a variant on the Fermi paradox: We don’t see clues to widespread, large-scale engineering, and consequently we must conclude that we’re alone. But the possibly flawed assumption here is when we say that highly visible construction projects are an inevitable outcome of intelligence. It could be that it’s the engineering of the small, rather than the large, that is inevitable. This follows from the laws of inertia (smaller machines are faster, and require less energy to function) as well as the speed of light (small computers have faster internal communication). It may be – and this is, of course, speculation – that advanced societies are building small technology and have little incentive or need to rearrange the stars in their neighborhoods, for instance. They may prefer to build nanobots instead.
— Seth Shostak
The whole interview is a very good read.
He has a TED talk as well, apparently
Seth Shostak: ET is (probably) out there -- get ready | TED Talk
