I discussed the Kardashev scale and how it quickly one might expect a civilization to climb it (answer: not long compared to the age of the Galaxy).
One objection that is commonly heard is that civilizations powerful enough to generate large amounts of energy are also powerful enough to destroy themselves, and so we should assume or at least consider that they have finite lifetimes. I actually don't put much stock in this line of reasoning (I put it in the same bin as the logic Dr. Donald Kessler uses
(no, not that one
Basically, because we have come close to causing or nearly causing our own extinction using nuclear technology, it might seem to follow that the more advanced we (and therefore other civilizations) get, the easier it would be for us to destroy ourselves. But actually, we are near a maximum in the probability that we will go extinct, and the probability is about to go down again.
The primary way we could destroy ourselves is through massive fallout and climate change from a global thermonuclear war. A comet or asteroid could strike the Earth to similar effect. But once we have a self-sufficient colony on the Moon or Mars, this danger lessens considerably: a truly self-sufficient colony would serve as a "lifeboat" for the species until the Earth was habitable again. This makes the threat of extinction more remote. The more colonies we have, the more robust we become against extinction.
Of course, one could always imagine the deliberate eradication of the species from an attack against the colonies, or a Solar-System-wide event like some sort of Solar anomaly or planetary dynamical rearrangement. But once we head off to the nearby star systems, this becomes nearly impossible. An attack on a distant colony would take generations to execute, during which time the technology of the colony would advance considerably, and its defenses presumably improve (imagine Cortéz landing on the shores of the Yucatan today and attempting to conquer Mexico with 500 men and 15 horses. I suspect the local Veracruz police forces could handle him in a matter of hours what with their "automobiles" and "rifles").
From there, the number of scenarios that might destroy the species dwindles: you're basically left with some sort of supernova or other catastrophic event that wipes out a swath of the Galaxy. But, again, once our colonies are far enough apart even this isn't enough. Once Galactic shear spreads our colonies around the Galaxy, our species will become essentially immortal.
But couldn't we be wiped out by another species or set of robots left behind by a prior civilization? Sure, but then that species or defense system would itself constitute a K3 civilization, so the original argument that K3 civilizations should exist still holds.
But couldn't there be some sort of Galactic environmental movement or "one child" policy that slows down growth? Not really. Communication is limited by the speed of light (I'm willing to assume), so any cultural or administrative coordination amongst the stars would be strongly hindered by the fact that a single back-and-forth would take longer than a human lifetime. Even if large swaths of the Galaxy kept themselves in check, the growth rate would always be dominated by the parts of the civilization growing fastest, and the slow-growers would quickly be overrun.
But couldn't we discover that virtual reality is much more interesting than real reality and stop growing because everything we need is in computers? I guess, but in order to argue that this would prevent K3 civilizations from being everywhere you have to argue that this will inevitably happen to every part of every civilization that ever arises. Otherwise the part that doesn't do this overruns the VR-ers and you're back to a K3.
In fact I think it's misleading to talk about a K3 "civilization". Really, we're talking about a "super-civilization" composed of many disconnected civilizations of a common origin. It would be no more able to prevent itself from infesting the Galaxy than ants can prevent themselves from infesting a countryside.
So the timescale from space flight to a K3 civilization is pretty short -- of order the crossing time of the Galaxy, which is around one Galactic rotation, or around 200 million years -- and there is actually only a narrow window within which a K1 civilization can destroy itself. Once it's a K2, there's really no stopping it.
Growth and energy uses are parts of the definition of life. Any argument that it will eventually be checked by something other than the available amount of energy must argue that such a check is universal and inevitable, regardless of the form of life, intelligence, and technology behind it. I find that hard to understand.
But what about alternative physics? Surely I'm being too narrow minded by assuming things like the speed of light as a fundamental speed limit and that aliens have to get rid of their waste heat? Again, I guess so, but if I can't assume conservation of energy, the laws of thermodynamics, and relativity are correct then I can't assume anything and so I can't do any reasoning at all. Also, these really do seem to be bedrock principles of the Universe; it's hard to understand how they could be wrong in any way that would change my argument.
Finally, couldn't there be dark sector physics that provide a source (or even a sink) of energy, obviating the need for starlight (and radiators?). Yes, certainly, and that will be part of the parameterization I describe in the next installment.
[I should also correct the record: the "Keeping up with the Kardashevs" joke was from Matt Povich, not Steinn. Sorry Matt; I'm sure it was bugging you that you didn't get credit for that gem ;)]