The Fermi Paradox

So, that post Robert mentioned about hash tables is still underway (3500 words and counting, it’s going to be a long one). But meanwhile, I wanted to respond to this post on the Fermi Paradox.

Basically, the Fermi Paradox assumes that civilizations grow exponentially- at which point, if there are any other alien civilizations of any antiquity, why aren’t they here? Since they aren’t here, then the only way to still be able to assume that other alien civilizations exist, then the exponential growth assumption has to be wrong. This generally assumes some sort of Malthusian catastrophe scenario. The more optimistic assume that a civilization might be able to learn to live within it’s limits.

I have a different theory.

Here’s an exercise. Got to The CIA World Fact Book, and download the data. Then do the following- create a scatter plot of all the countries. Make the vertical axis be GDP per capita, as an approximation of individual wealth. Make the horizontal axis be births per 1,000 minus deaths per 1,000, as an approximation of population growth excluding immigration (both directions). Now, take a look at that graph.

I’ve done this exercise a couple of times, and what I see is the following: there are a couple of outliers, which upon inspection turn out to be either small countries with large tax-haven banking establishments (Isle of Man, Luxembourg), or poor oil-exporting countries. In other words, they are countries with a small number of very rich people (bringing up the GDP per capita average), and larger numbers of either middle class (for the banking countries) or starving-poor (for the oil exporting countries) increasing the population growth rates.

But whether you include these outliers or not is irrelevant. At the bottom of the graph, those countries with very low GDP per capita, there is a wide spread of population growth rates- from countries with serious population explosions going on (births-deaths = +30 pop/1000/year) to countries with serious population crashes going on (births-death = -15 pop/1000/year, sort of a mini-malthusian catastrophe). But as the GDP per capita climbs, two things happen- first, the spread narrows considerably, and second is that it trends towards zero. It looks like once GDP per capita gets to somewhere between $40,000USD (if you ignore the outliers) and $120,000USD (if you include them), then the population growth rate would seem to want to fall to zero.

What is also interesting is that there does not seem to be much correlation between other social aspects of the countries. For example, Catholicism is very much a pro-population-growth religion, being opposed to both abortion and birth control (before accusing me of anti-Catholicism, please note that I was born and raised in a Catholic family. If the shoe fit, wear it. Sing along- Every sperm is sacred…). So you would expect heavily Catholic countries like Ireland and Italy would have higher growth rates for their GDP than non-Catholic countries would. But the data doesn’t support this conclusion- Ireland and Italy are very much in line with protestant countries like England or Germany, or even non-christian countries like Japan.

This makes sense if you think about it economically. In an impoverished farming lifestyle, children are an economic advantage. A child only a few years old can start doing simple chores, and helping with other chores, like gathering eggs, weeding gardens, etc. This directly and measurably contributes to the wealth of the family, and an older child quickly approaches the productivity of a full adult. Also, children are people’s retirement plans, having enough children so that some of them are likely to survive until you’re old is a necessity. So the more children you have, the wealthier you are.

In an industrial society, and even more so in an information/technological society, children are not a source of wealth, they are instead a drain on wealth. The “chores” the city kids do contribute effectively nothing to the wealth of the parents. Take out the garbage? Load the dishwasher with the dirty dishes? Mow the lawn? Even if the kid gets a part time job at a local fast food joint, generally the kid keeps the proceeds there. And the cost of having a child is much greater- just consider college costs alone. Economically, the parents would be better off to not have the kids at all, and just paying someone to mow the lawn and spending the ten minutes to take out the trash and load the dishwasher yourself. And as for retirement, these days we have Social Security and 401Ks and all sorts of means wherein an adult provides for their own retirement.

There are non-economic advantages to having children, granted. But those advantages don’t scale- the non-economic advantages provide no incentive, once you have 2-3 kids, to have any more kids. If you have two kids, you’ll get all the fun of the first day of schools, all the art work to hang on the fridge, the ability to scare the piss out of their prom dates, and so on. But these non-economic advantages don’t cause you to go out and have 12 kids, when 2 is enough. And having only 2 kids isn’t even population replacement.

On a geopolitical level, this remains true. Numbers don’t count (militarily) for what they once did- technology is such a force multiplier as to overcome even large numerical disparities. There is some disagreement over how many people Saddam Hussein had under arms in the first gulf war (the one in 91). Some estimates place it at 5 times the number of allied troops, some at only 2 times as many. But everyone agreed there were lots more Iraqis than Allies. Fat lot of good it did him. The technological advantage of the allies, while being only a few decades in terms of time, was more than large enough to compensate for the numerical disadvantage, to the point of turning the conflict into a one-sided slaughter.

And the rate of technological change is still increasing- the next forty years will bring more change, and more advantages, than the last forty. Any society that slows it’s technological change rate will rapidly find itself out gunned. This includes those societies that choose to spend their wealth on child rearing rather than technological development (both of which cost money), and the additional numerical superiority will not counterbalance the technological superiority of the other side.

On the other hand, the nice thing about this theory is that it doesn’t require any enforcement of any particular ideology. We don’t need to go to the poor third world and go “you’ve got to change your behavior”, which will likely encourage resistance. Instead we should help the third world become first world rich. They don’t need to change their culture or religion or anything except their wealth, and that they need to increase. Rather than being global moralists and cultural-imperialists, telling people what they can or can’t do, we get to be global altruists, who are simply helping everyone else be rich.

In a very real sense, I’m arguing that population will hit a resource limitation that will hinder further population growth. But that this resource isn’t going to be food, or energy, but will instead be the attention and effort of parents. And that the effects of hitting this resource limitation won’t be some huge catastrophe where everyone dies, oh the embarrassment, but rather a soft landing of (voluntarily) smaller family sizes and fewer children.

So the question then becomes, can the planet support 6-20 billion people each making $120K a year on average? I’d say yes, given three slight changes to our thinking.

The first is the assumption that energy is something you pump or dig out of the ground. There is plenty of solar and wind power available. Yes, the sun isn’t always shining and the wind isn’t always blowing, so what you do is generate extra power when the sun is shining or the wind is blowing, and store it (maybe in batteries, maybe as hydrogen, maybe just by pumping water up hill) to use when the sun isn’t shining and the wind isn’t blowing. So long as you average power generation is more than the average power consumption (plus what you lose to storing the power), you’re fine. Vehicles are just a problem of transporting the power- maybe we’ll use hydrogen, maybe we’ll use batteries, maybe we’ll use thermal depolymerization to make more oil.

The second assumption is that once we use something, we can never use it again. I’m talking on the atomic level here- not the macroscopic level. The car might be a junker and not worth repairing, but why throw the iron atoms away? Why throw any atoms away? Sure, they might not be in the chemical configuration we want them to be (damned iron atom bound to an oxygen atom, and now it’s rust), but that’s just a question of energy, and I’ve already established that energy isn’t a problem. There is only a finite number of atoms in easy reach (a very large number, granted, but still finite). But once we use an atom, why stop using it? In addition to having to constantly find new atoms to use, you’ve got to constantly find new places to store your used atoms. It makes more sense, economically and technically, to keep using the same atoms over and over. I’m talking 100% recycling here. Maybe it’ll be by thermal depolymerization, maybe by nanotechnology, maybe some other way, but it’s coming. And it’s safe to assume the long-time surviving alien civilizations already have it.

And the third assumption is we need to rethink is that we’re limited to the Earth. Even with our current technology we can get to most of the solar system, and once space elevators ramp up we’ll be able to get there cheaply. This just opened up several earths worth of raw materials, conveniently already broken up into bite sized chunks (called asteroids), and several millions of earths worth of free solar power for the cropping. Not to mention a near-earth or three, needing only a little fix me up (aka terraforming) to make a whole new planet for us to live on. And if the solar system isn’t enough, we can go interstellar.

This gives rise to sort of a self-fulling prophecy. If sufficient wealth creates a natural ZPG in a society, then once that’s achieved, wealth ceases to dissipate, and thus zero population growth is maintained. Indeed, as wealth probably continues to accumulate, there becomes even less of a temptation to positive population growth.

Imagine the following situation: 20 billion human beings spread across a thousand star systems. Now, there are about 1,200 star systems within a hundred light years of earth, so humanity hasn’t really spread all that far, especially if human civilization is a million years old. But this means the average star system will have only 20 million people in it- slightly larger than the New York metropolitan area is now. But the wealth (as measure by the energy and raw material budgets) of the entire star system is spread among these few people. This is a civilization wealthy beyond even the dreams of Bill Gates. If ever a civilization would achieve natural ZPG, this one is it.

And yet, for all that it is an interstellar civilization and sometimes empire, wealthy beyond imagining, it comprises less than 0.0001% of the entire galaxy. Ten thousand of these interstellar civilizations could be lost in the immensity of the Milky Way. This is a galaxy teaming with ancient, high tech civilizations, and yet almost all of it is empty. In terms of the Drake Equation, a galaxy with 10,000 civilizations is among the more optimistic assumptions. There is no paradox why the aliens aren’t here- the unlikely event would be if the aliens were anywhere near here.

I want to address one last objection: if the alien civilizations are out there, especially if they’re out there in large numbers, why can’t we hear their radio broadcasts? The problem here, I think, is that the number of bits you can transmit on a given frequency, over a given area, is fixed by a combination of physics and information theory. So radio waves are incredibly limited for long distance communication- there just isn’t enough broadcast bandwidth available to go around. So, even in our limited technology environment, we’re already shifting to using radio broadcasts for the last mile (or the last hundred yards, or even the last hundred feet). Wires of various sorts (copper or fiber) is what we use for long-distance communication.

When I make a call on my cell phone here in New York, to Robert’s cell when he’s in California, the call doesn’t go via radio waves the entire way. No, the radio communication is only between my cell and the nearest cell phone tower (across the street, it looks like), and between the nearest cell phone tower in California and Robert’s cell phone. The long haul, between New York and California, as done on wires. The advantage of wires is that if you need more bandwidth between New York and California, you can just lay more wires, you can’t make more radio frequencies.

But now consider this communication from the perspective of someone on, say, the moon. The communication going over the wires goes completely unheard. The radio communication can be heard, but even here there is a problem. Robert and I (or at least our cell phones) have no problem distinguishing between our local cell towers and other ones, the local towers are much, much closer. But from the vantage point of the moon, all the cell towers in America are more or less the same distance away. The result would be cacophony. From the moon, with some care, it might be possible to isolate individual cell towers and communications. However, since broadcast bandwidth is a finite, fixed resource, there is a strong incentive for any technological civilization to compress the information sent over radio waves- and since it’s public, to also encrypt it. Both of which make the signal look much more like random noise, unless you happen to know the decompression and/or decryption scheme. Even if you manage to isolate out a single transmitter, it is still transmitting something that looks an awful lot like noise. And if you back off to, say, Alpha Centauri, the problem becomes exponentially harder. It’s unlikely that we’d see an alien civilization’s radio broadcasts as anything other than noise.

For interstellar communication, directed beam communication would be preferred. For all-directions broadcast transmitters, the power required to reach a certain distance goes up as the square (IIRC) of the distance covered, while beam transmitters are more linear in their power requirements. And the problem with all-directions transmitters is that they blanket that frequency everywhere. If you’re transmitting to your friends 10 light years away at a given frequency, basically no one else in ten light years (including both you and your friends) can use that frequency for anything else. No, much better is to use a directly laser or maser beam for communication. But that means we’re unlikely to stumble upon such a beamed transmission, unless it’s directed explicitly at us.

So why haven’t they sent us a tight beam hello yet? Almost certainly because they haven’t heard us yet. AM/FM radio waves don’t leave the atmosphere, the ionosphere reflects them. It wasn’t until the introduction of TV, about 70 years ago, that radio broadcasts that the aliens could detect happened. But these waves have only traveled about 70 light years. This means that if the nearest alien civilization is only 100 light years away, which is right next door (and isn’t it crowded around here) in terms of galactic distances, our transmissions haven’t reached them yet. They won’t hear our broadcasts until 2037 or so, and even if they immediately send us a hello, it won’t get here until 2137.

And this assumes that photons are the best way to communicate interstellar distances- may I remind people that we’re so backwards we don’t even have a workable theory of quantum gravity. Declaring that there can be no faster communication than that is not unlike Lord Kelvin declaring heavier than air flight impossible. We simply don’t know enough. And if some form of faster than light communication is possible, it makes sense for that to be the “threshold of interest” for alien races. Rather than worrying about detecting their radio broadcasts with it’s centuries to millennia turn around times, wait until they develop subspace communicators, and then say hello to them in real time (or something much closer to it). Heck, they’ll probably discover subspace communicators before their TV broadcasts reach you anyways.

I have been making an awful lot soup from a lone oyster here. Unfortunately, that is rather unavoidable when having this discussion- at this point in time we only know about a single technological civilization and a single intelligent species: us. I have been blithely assuming that human economics, human science, human mathematics, human technology, human experience is in some vague sense “average” or “normal”, an assumption I have absolutely no support for. But the only thing that makes less sense to me than extrapolating from one to infinity, is extrapolating from zero to infinity. That is, ignoring the single data point in this discussion we do have in drawing our conclusions. But that seems to be a very common human failing.