In a comment to the previous post on the Fermi Paradox, David Wallace wrote this terrific exposition of it. I agree with almost everything he says, so I will save my own quibbles for a comment. David writes:
Just to clarify the force of the Fermi paradox (which is intended to rest on quantitative factors, not just a general "where are they?"), here are the main basically-uncontroversial premises that get it going:
- The Galaxy contains an awful lot of stars, and plenty of them have planets. There are 100+ billion stars in the Galaxy, and in our neighbourhood, around 5% of them are G-type stars like the Sun in non-binary systems. Our best theories of planet formation make planets look generic, and the current era of exoplanetology backs that up. I think it would be very hard now to argue for fewer than several hundred million Earth-type planets, and that's almost certainly a significant underestimate.
- The Galaxy is much older than it is large. The Galaxy is something like 100,000 light years across. But it's nearly ten billion years old, which is to say that light could have travelled across the Galactic disk and back some 50,000 times.
- Slow interstellar travel (c.0-1%-1% of light speed) is technologically possible. Pioneer 10 is already a starship, travelling at (1/30,000) of the speed of light, and that was by accident. There are a quite large number of designs and plans for spacecraft capable of reaching between 0.1% and 10% of light speed. That we are unlikely to build any of them any time soon - the mission time is too long and the cost is currently rather painful - isn't my point; the point is that this kind of speed is borderline possible with *current technology*, let alone the extrapolated capabilities of much-more-advanced civilisations. (Contrast relativistic starflight, which doesn't violate any known physical principle but isn't remotely possible with any feasible technology we can think of now; contrast faster-than-light starflight, which actually looks *physically* impossible.)
- Self-replicating intelligent machinery is technologically possible. We have an existence proof for this: humans. (I guess I can't absolutely rule out some strange principle of emergent engineering which makes it physically impossible to make a slow-moving space probe capable of carrying machines (humans or not) that can reproduce the probe even though self-replicating intelligence and probes are separately possible, but it seems ad hoc.) Given 2-4, it is technologically possible for a civilisation to send physical vehicles to every solar system in the Galaxy in a timescale *much* less than the age of the Galaxy. (Err on the pessimistic side: assume 0.1% light-speed propulsion and a 1000-year turnaround time; you still get an expansion wave moving at 0.05% of lightspeed, enough to span the galaxy in 200 million years, or 2% of its current age. Why might a civilisation do so? Several plausible reasons: (i) colonisation. As Allen Olley points out, this isn't a plausible way to move your existing population (I don't really agree with Eric's resource-based argument for colonisation) but it is a plausible way to spread your civilisation, or your species, across the Galaxy. (ii) exploration. Robot probes is a really great way to explore space, as we've found ourselves. (See the Wiley paper cited above.) (iii) malevolence or prudence. People very often react by noting that each of (i)-(iii) is speculative, and maybe intelligent civilisations don't tend to do that kind of thing. Fair enough. But the crucial issue in the Fermi paradox is that *it only has to happen once*. And unless life *basically never* evolves, or intelligence *basically never* evolves, or civilisations *basically always* die off before they get to the stage of exploring the galaxy, there ought to have been so many civilisations that it seems implausible that not even one would do this. (Note that I'm pretty sure *we* would do it, if we got to the point at which it was reasonably inexpensive.) An illustration: suppose life only turns up on 1% of earth-type planets, and that intelligence only evolves in 1% of biospheres, and that 99% of civilisations wipe themselves out through war or resource exhaustion before reaching the stage when it's feasible (indeed relatively cheap) to spread across the Galaxy. Then, on a pretty conservative estimate of 1 billion habitable planets, there should have been 1000 suitable civilisations so far in Galactic history. Yet *not even one* decided to try the explore-the-galaxy strategy? Things get several orders of magnitude worse when you consider that the strategy works perfectly well over intergalactic distances too, and that there are hundreds or thousands of galaxies in realistic exploration range. Could they be here but not be noticed? Well, it's quite hard to hide a decelerating starship, but maybe none happen to have turned up recently. There could be all manner of things hiding in the Solar System. But again: does *every single* civilisation behave that way? It seems unmotivated to suppose so. (Unless there is a single hegemonic species or coalition that requires it coercively.) I think the force of the Paradox is this: yes, for any given civilisation we can imagine reasons why we could be completely oblivious of it despite it being millions of years older than us and quite capable of having spread across the Galaxy. But unless technological civilisations are fantastically unlikely to arise on a per-planet basis, there should have been such an enormous number of them by now that it becomes really implausible that *none of them* have done so.