These are common misconceptions about what the future is likely to bring which have gotten enormous traction in the popular imagination.
In this post I will take on misconceptions about robotic unemployment, demographic space colonization, and cryonics. Each is a stand-alone topic, so skip ahead to the one that interests you most. Despite my bad track record when it comes to completing series, I’m going to save “strong” AI and progress/”the singularity” for another day. (Speaking of which: if anyone is terribly eager to see me finish one of the series of posts I started over the summer, feel free to bug me about it.)
The robot revolution has already come and gone. It was called the McCormick reaper. Tractors helped, too. For most of the history of civilization, 95% of our ancestors were engaged in agriculture. Only 5% of us are. So bringing “robotics” to the biggest single sector “destroyed” 90% of the jobs in a traditional agricultural economy… and what happened?
Well, people still work.
TECHNOFUTURIST: But This Time It’s Different™! When previous waves of mechanization automated plowing and sowing and reaping and threshing, and logging and fishing and trucking and canning and weaving, they were just replacing human labor-power with mechanical labor-power. That allows humans to move out of backbreaking manual occupations into more intellectually stimulating professions. But robots will be replacing human brain-power, which is an entirely different kind of dynamics and will leave us with nothing at all to do.
ME, A BLOGGER: Have you ever read Bartleby the Scrivener? If not, you probably have no idea what a “scrivener” is; it was a profession that required a fair amount of education, training and intelligence that has now been entirely replaced by mechanical solutions. And you probably think a “factor” is something you multiply and a “calculator” is something everyone has on their phones nowadays. That’s not wrong, but once upon a time calculators and factors were two of the most common jobs that required mathematical reasoning. In fact, if you read accounts of the Manhattan Project, you’ll discover that most of what the world’s top physicists were doing in those days could be done with a calculator today. — Once upon a time, the world’s most educated men were employed as scribes (rendered relatively low-prestige by alphabets, then made irrelevant by printing presses), as astrologers (replaced by almanacs), as artillerists (replaced by slide rules). Even the farmers, weavers, and potters whom we casually described as “labor-power” above were actually quite astute, skilled people who needed a whole universe of know-how to efficiently solve problems only modern industry can afford to solve with brute force (and which sometimes, it simply doesn’t solve at all). — There is absolutely nothing new about technological progress replacing brainpower with machinery. In every case, the new option leads to greater investment in capital goods, and substitutions of human labor and human capital out of some sectors and into others.
To clarify: I deny that robots will cause, as a social problem, unemployment in the macroeconomic sense (in the same way that labor unions, minimum wages, or strict employment contracts cause involuntary structural unemployment). But just as multi-millionaires can choose to live off their dividends and devote their days to taking long naps, a society with deeper capital investments can find ways to support more people in idleness.
Demographic space colonization
There are many ways to deal with overpopulation. You can have fewer kids. You can go to war. You can invest in capital goods that allow one slice of land to support a denser population. One thing that will probably never make sense is accelerating human bodies up to some considerable fraction of the speed of light and decelerating them again. Human beings may well colonize other solar systems, but only with some minimum feasible breeding population.
Historically, colonies with different underlying purposes have looked more or less identical to one another, which has led to conflation of what we may call demographic colonization with geographic colonization. Demographic colonization: a community has too many people and tries to find somewhere (anywhere!) to put them. Geographic colonization: a community wants to extend its control over some region and wants to find someone (anyone!) to go there and garrison it. In practice these look similar (build some boats, gather some supplies, send the colonists off), and typically colonizers try to divert people who want to be colonists towards areas they want to see colonized. But in each case, colonization is a substitute for very different alternatives. The alternative to geographic colonization is to conquer/enslave/hire indigenous workers and mercenaries, and to invest in projects that allow the home country to guard the colony with very few colonists (like forts or fleets). The alternative to demographic colonization is to take the resources that would have been used to establish a colony on virgin soil, and instead use them to improve improve production-per-acre in the homeland, supporting a larger population on the same territory.
The following calculation is 100% back-of-envelope (I’m not even taking relativity into account!), but should make the point. Say a human weighs 130# (low) and that a manned interstellar voyage has to go at 0.5c to have a reasonable chance of success (slow). If you add one additional human being to a manned interstellar voyage, you are going to need at least 6.7 x 10^17 joules to get that colonist humming along at 0.5c. Then you’ll need about that to get him to stop at the end of the journey.
Conveniently, ~1.3 X 10^18 joules is just a bit over 1 quad, the unit which the USA uses to measure total energy use. Total US energy use is ~95 quads [pdf], a bit under the bare minimum required to accelerate and decelerate 100 colonists on an interstellar journey. That’s not even enough emigration to solve over-crowding in the average ‘Kwa student dorm. (And that is just the bare minimum of energy required to accelerate/decelerate the bodies. Never mind the life support systems, the supplies for building the colony, the ship itself; never mind waste energy; never mind navigation; never mind escape velocity, never mind the difficulty of doing this all in space.)
There is no way we will ever be in a position to divert resources capable of generating 1 quad of energy to helping a single human being escape overpopulation. If we do have massive amounts of energy to devote to density-remediation, there will be much cheaper ways to invest that energy within our own solar system. Furthermore, possible futures in which density-remediation is necessary tend to be futures in which energy-use-per-capita declines as the centuries pass; and the poor techno-wretches who are bearing the brunt of hyperdensity will be the ones to whom society diverts the least energy.
TECHNOFUTURIST: But your parameters are silly. The ship could go much slower, for one thing, and… [blah, blah, blah]
ME, A BLOGGER: There are lots of underlying assumptions you could fiddle around with, but they all lead to the same conclusion: no interstellar demographic colonization. Are you going to put the colonists into cryonic sleep? Well, they don’t know what the annual chance of a malfunction is, so unless you speed up the ship enough to keep the risk of unplanned-thawing very low, they’re not going to be too keen to sign up. Are they going to stay awake for a (massively) intergenerational trip? Probably you’re not going to design the mothership with all the amenities of an Ohio-class nuclear submarine. For one thing, if you could give them that standard of living, why would you bother to also point them into deep space? But there is also the additional problem of accelerating all of those tennis courts and swimming pools. On the other hand, if you squeeze them into filthy, dystopian warrens, why would they want to leave their filthy, dystopian warrens on Earth? There is no way you could even start to imagine people voluntarily signing up for interstellar colonization under conditions that would make that kind of mission affordable for the population of a small town… but you would need to enlist millions (or more likely, billions) of colonists to make a dent in terrestrial overpopulation. — Now, you could pressgang the surplus population to get rid of them, but if you’re willing to force them to be miserable, you could also murder them or neuter them or (best of all) not worry about the unpleasant side-effects of population density in the first place.
You can keep playing around with the variables, but the best-case scenario is probably to prove the feasibility of large-scale geographic colonization of other systems. There would be enough volunteers for a minimum viable breeding population even with a 99% chance of never reaching the destination. (A 1% chance of opening your eyes under an alien sun would appeal to at least one volunteer per 1M in the general population, and you don’t need 8,000 colonists.) Demographic colonization of other planets/objects within our own solar system is also possible (although the complexity of calculations and comparisons that would be required to make a prediction is out of proportion to the inherent improbability of the event). But any adjustments you make to the interstellar demographic colonization scenario to solve one problem cause other insoluble contradiction to emerge. In short, interstellar migration is a solution in search of a problem.
Do you know anyone who died? What happened to him? Some people die suddenly. Stroke, heart attack, thrown out the windshield at 70mph, you name it. These people generally do not die in a cryonics facility. They face the embarrassment that their hearts are no longer providing their brains with blood: no oxygen in, no carbon dioxide out, the brain fails, the paramedic calls time-of-death. Denying your brain oxygen isn’t like tripping over a power cord. We know this because there is a form of pseudo-death that is pretty similar to yanking out the power cord, namely hypothermia; when someone “dies” of hypothermia, the cold slows down physiological processes enough that if the patient’s body temperature can be raised in a safe way, there will have been little damage to the vital organs (in particular to the brain), so he will revive. Little, but not none: the longer a hypothermia victim remains in freezing temperatures, the more permanent brain damage he exhibits after he is revived, up to the point that revival is impossible.
Same goes for resuscitating victims of cardiac arrest. There, the margin for error is unforgiving: after the first minute, successful resuscitation is unlikely. This is just to say that when you die of an acute condition, you only have a matter of minutes before you sustain enough brain damage that, even if your other organ systems would start working again, your brain would be reduced to a sort of potato, unable to support more than simple vegetable functions. In many cases the damage to your brain is debilitating and permanent before you can be pronounced clinically dead. (How would it feel to be Terry Schiavo… until the heat-death of the universe?)
Not everyone dies suddenly. Some people die slowly. Some of these diseases, like Alzheimer’s or Parkinson’s, are actually themselves the external signs of the breakdown of brain-machinery. These diseases will turn your brain to mush long before the rest of your body is weak enough to die. Others, like metastatic cancers that eventually reach the brain, manage to take significant chunks of it with them before killing you. Yet other chronic diseases stress the body to the point that all the organ systems start to wear out at the same time; someone with heart disease is nearing the end when they go from frequently confused to flat-out crazy. And if you don’t die of anything in particular, just of “old age”… well, no one is going to find you until the morning. Your brain will be useless. Sorry!
There may be a few diseases I’m not very familiar with where (a) the progress of the disease is so quick that you know to get ready for the cryonics team in advance, and (b) the symptoms of the disease are so specific that there is absolutely no damage to your brain until the very moment you stop breathing. But most people don’t die of these diseases. Most people die suddenly and unexpectedly of diseases that do irreparable damage to your brain before your wife calls the ambulance, or after a long lingering assault from a disease that does irreparable damage long before you’re dead. If you keep trying to stay alive another day, eventually your number will come up and you’ll die in a way that is, from the reanimation point of view, messy. Cryonics enthusiasts should know this, and if they really believed that cryonics gave you a shot (any shot!) at being raised glorious and incorruptible, they would all end their lives in a planned manner, to make sure that they get frozen while their brains are still hale.
TECHNOFUTURIST: You clearly haven’t thought about this at all. Here, come read this 100-page thread in the Less Wrong forums, this will give you a primer on the basics…
ME, A BLOGGER: No.
TF: I’m not certain that cryonics will work, but it’s only the second-worst thing that can happen to you (after rotting in the ground).
ME: The problem with this line of argument is, first, you guys never seem quite so interested in precisely calculating risk and reward when you blather about what a good deal cryonics is. If technofuturists never mentioned cryonics without clarifying that cryonic freezing is similar to burial/cremation/excarnation, but the chance of staying dead might not be quite 100%, then I wouldn’t call it “BS”. — The second problem is that your risk of dying rises day by day, so for any distribution of belief among enthusiasts about the odds of cryonics working, I would expect to see a distribution of ages at which they choose to off themselves to make sure that their brains get frozen in good condition. But you don’t see this (or if you do, y’all have managed to keep it extremely quiet). So until I hear about the FBI raiding cryonics clinics and charging the staff with assisted suicide, I’m going to go out on a limb and say that no one actually believes in cryonics. In fact, so little faith do cryonics clients have in cryonics that they actively avoid the information that would convince them they need to commit suicide for the cryonics to possibly work.
TF: I expect my brain to be mush when I die, but any future civilization that can revive my cryonically frozen brain will be able to repair brain-damage.
ME: This rationalization might work for certain very specific, limited kinds of brain damage (for example, in Parkinson’s it is specifically the parts of the brain that make/regulate dopamine that are destroyed, and it’s possible they are interchangeable from person to person), but come on… how many millions of neural connections are there in a human brain? How many different permutations of these are possible in a healthy, functioning brain? Even if your post-singulitarian redeemers are going to be able to take a frozen brain and raise it glorious and incorruptible, how could they possible know how to choose the precise permutations that would make it your brain rather than mine?
For some technofuturists, cryonics is a sublimation of the longing for that offer of immortality which Christ has extended to us all. For others, it is an ostentatious way to signal their rejection of God and any divinely ordained afterlife. Still others flinch away from the cold of the grave, and grope for any kind of distraction. Having found others motivated by similar concerns in the corners of the internet, they are convinced that they are bold truth-tellers; could so many people, many intelligent, all strongly convinced, all be misguided at the same time? In this they are like sexual deviants who use the internet to find strangers who will embrace their perversions.