Support the IEET




The IEET is a 501(c)3 non-profit, tax-exempt organization registered in the State of Connecticut in the United States. Please give as you are able, and help support our work for a brighter future.



Search the IEET
Subscribe and Contribute to:


Technoprogressive? BioConservative? Huh?
Quick overview of biopolitical points of view




whats new at ieet

Enhancing Virtues: Caring (part 1)

Death: Why the Brain Matters

Obamacare’s “Secret Trick” Could Cure CEO Pay Excesses

Here’s a Terrible Idea: Robot Cars With Adjustable Ethics Settings

Thoughts on Bostrom’s ‘Superintelligence’

Skepticism, the Singularity, Future Technology & Favorite Frauds


ieet books

Superintelligence: Paths, Dangers, Strategies
Author
by Nick Bostrom


comments

Rick Searle on 'Death Threats, Freedom, Transhumanism, and the Future' (Aug 30, 2014)

CygnusX1 on 'Death Threats, Freedom, Transhumanism, and the Future' (Aug 30, 2014)

Rick Searle on 'Death Threats, Freedom, Transhumanism, and the Future' (Aug 30, 2014)

CygnusX1 on 'Death Threats, Freedom, Transhumanism, and the Future' (Aug 30, 2014)

Rick Searle on 'Death Threats, Freedom, Transhumanism, and the Future' (Aug 30, 2014)

CygnusX1 on 'Death Threats, Freedom, Transhumanism, and the Future' (Aug 30, 2014)

instamatic on 'Death Threats, Freedom, Transhumanism, and the Future' (Aug 29, 2014)







Subscribe to IEET News Lists

Daily News Feed

Longevity Dividend List

Catastrophic Risks List

Biopolitics of Popular Culture List

Technoprogressive List

Trans-Spirit List



JET

Transhumanism and Marxism: Philosophical Connections

Sex Work, Technological Unemployment and the Basic Income Guarantee

Technological Unemployment but Still a Lot of Work…

Hottest Articles of the Last Month


Enhancing Virtues: Self-Control and Mindfulness
Aug 19, 2014
(7758) Hits
(0) Comments

Is using nano silver to treat Ebola misguided?
Aug 16, 2014
(6416) Hits
(0) Comments

“Lucy”: A Movie Review
Aug 18, 2014
(5638) Hits
(0) Comments

High Tech Jainism
Aug 10, 2014
(5253) Hits
(5) Comments



IEET > Life > Directors > Mark Walker

Print Email permalink (2) Comments (7186) Hits •  subscribe Share on facebook Stumble This submit to reddit submit to digg


Universal Superlongevity: Is It Inevitable And Is It Good?


Mark Walker
Mark Walker
in Death And Anti-Death, Volume 3: Fifty Years After Einstein, One Hundred Fifty Years After Kierkegaard Charles Tandy, Ph.D., Editor. Ingram.

Posted: Feb 17, 2006

[This is an early version of the essay]

Abstract

To the extent that it is possible to predict the composition of the future population of the world, I argue that the most likely scenario is that (almost) everyone will choose to adopt technology to live hundreds of years, perhaps indefinitely. Further, ethically speaking, this seems like the best option for our world, for such a world is one where there are higher levels of happiness and achievement.

1. Population Predictions

Rapid advances in the understanding of human aging and possible ways to combat it, e.g., discoveries in genetics (Kenyon, 1996), stem cell research (Shostak, 2002), the cessation of aging at the cellular level (de Grey, 2005), and nanomedicine (Freitas, 1999), point toward the possibility that sometime this century we will have to confront the decision of whether we should radically extend the healthy lifespan of humans. Just imagine being 1000 years old with the body of a twenty-five year old (or so) to get some idea of the potential promise of developments in biotechnology. Given that such an extended lifespan (‘superlongevity’) is technically possible, I propose to examine two related questions: Will our descendents opt to use this technology or not? And if they do so, is this a good thing, morally speaking? The answer to both questions, it seems to me, is “yes”. Let us take these in turn.

Given the enormous contemporary interest in looking young among certain segments of our society, it may seem reasonable to think that everyone will line up to access this technology. But perhaps this judgment is skewed precisely by the contemporary hype over looking young e.g., make-overs, botox injections and cosmetic surgery.[1] Why not think that this is a fad, and that it is more reasonable to suppose, as with so many other choices in life, that some will say “yea” and some will say “nay”?  Alternatively, perhaps it might be thought that the nearly universal adoption of technology (at least in the wealthier nations) ought to be a good indication of how many will opt to use superlongevity technology. The qualification of “nearly universal” is necessary since there are of course exceptions: the Amish are one conspicuous example of the rejection of much modern technology. But this comparison may be misleading. Not all technologies are widely adopted, and ‘superlongevity technologies’ refers to specific technologies designed for specific purposes. It is true that certain technologies, such as the telephone, television, and radio, are used by over 90% of the population in the rich nations; yet, most people who live in first world nations could afford a jetski, but its use is confined to a small fraction of the population. Will superlongevity technology be the jetski of the future? Furthermore, there is at least one crucial difference here that ought to give us pause. The type of technology under consideration involves altering human biology in a manner that many might consider “unnatural”. The idea of humans living for hundreds or thousands of years is unprecedented in our history. Conceivably, many will reject this as fundamentally different from other types of technologies, and so find no contradiction in using dishwashers and cell phones, etc., while forgoing the use of radical life extension technology. Perhaps some will do so for religious reasons, others might reject it for reasons similar to the opposition to genetically modified foods, that is, that we ought not to alter the biology of any species including humans.

It may be thought that systematic empirical research should put us on a firmer footing here, since drawing analogies from our current practices does not seem to provide a univocal answer. Such research might survey people today to find out whether they would adopt superlongevity technology if it were available, and use this as an indicator about how future populations might decide the issue. However, I do not know of any such studies.[2] Furthermore, we run into the problem of surveys asking people whether they are likely to adopt some emerging technology are notoriously unreliable. For example, asking about microwave ovens in 1970 would have indicated that there would be little interest or demand for this device [Ref]. Obviously, then, such research would be a bad predictor of the fact that microwave ovens are overwhelming embraced today in the rich nations.

To add further difficulty here is the (seeming) problem that long-term predictions might be confounded by small changes in individual preferences. This point is nicely illustrated by a report “World Population to 2300” (2002) released recently by the population division of the UN’s department of social and economic affairs. The report predicts 8.97 billion humans on earth in the year 2300. However, the report is quite clear how sensitive this forecast is due to a number of factors including fertility rates. The report notes, for example, that fertility rates at .3 children above replacement would lead to a quadrupling of the prediction to 35 billion in 2300; and with fertility rates at .2 children below replacement leads to a prediction of just over 2 billion inhabitants in 2300 (a significant decline from our present population of over 6 billion (pg1)). So, small changes in preference for the number of children, therefore, can make a huge difference in the long-term.

So, it may seem at best incautious and at worse pointless to speculate how the future population of the world will divide between those that seek superlongevity and those who choose a natural life span—we will refer to them as ‘mortals’—given we do not know what people’s preferences will be in the future. However, I want to argue that, if given the choice, those adopting superlongevity technology will eventually make up (almost) the entire population. This conclusion holds even if there is a fairly small preference rate[3] among future individuals for superlongevity to the mortal life.

2. A Simple Model

Admittedly, the claim that superlongevitists will eventually compose almost the entire population may seem counter-intuitive. For example, if the rate that individuals opt for radical life extension is 50% then shouldn’t we predict that 50% of the population will choose a “mortal life” and 50% that of a superlongevitists’, and if the preference rate is merely 10% then they will comprise 10% of the population, and so on?  The answer is that this is certainly what we should expect during the first generation of mortals. However, since superlongevitists do not die, their numbers will continue to accumulate and so they will tend to become an increasingly large segment of the population.

To see this in more detail let us think about a simple model based on the following assumptions.

1.    Availability: Superlongevity technology is available to all by 2050.

2.    Population Stability: The world’s population is stable at 10 billion from 2050 for the foreseeable future.

3.    Preference Rate: For every individual, there is some non-negligible probability that he or she will choose to adopt superlongevity technology.

4.    No Defection: There is no defection among the superlongevitists (none of them die).

It may be helpful to work through an application of these assumptions before discussing their plausibility. For our first application, let us assume a constant 50% preference rate for the adoption of superlongevity technology, as Fig. 1 illustrates. That is, we assume that every individual has a chance to decide for him or herself whether to adopt superlongevity technology, and half opt for it, and the other half choose to live a mortal life.

Figure 1

The mortal population we will assume experiences a complete turnover in its number every 70 years. Given the 50% preference rate, five billion of those alive in 2050 adopt superlongevity technology and five billion choose the mortal life. So by the year 2120 the original 5 billion mortals will be replaced. Since we are assuming that the population is stabilized at 10 billion, this means that by 2120 5 billion new persons are born. Since the preference rate is constant at 50%, 2.5 billion of the 5 billion born between 2050 and 2120 choose superlongevity technology. So in the year 2120 there are 7.5 billion superlongevitists and 2.5 billion mortals. From 2120 to 2190 2.5 billion mortals die. Their replacements again choose the mortal life at a 50% rate, so by 2190 the mortal population has been reduced to 1.25 and the superlongevitists’ population has swelled to 8.75 billion. As Fig. 1 indicates, looking 330 years into the future (only slightly longer than the scope of the UN population study) the mortal population is but a small fraction of the total population (about 3%).

        This result is not simply an artifact of the 50% rate; since even with a relatively low 10% preference rate, inextricably superlongevitists become a larger and larger proportion of the population. As Fig. 2 indicates, within just over 300 years they comprise over 40% of the population.

Figure 2

Fig. 3 shows the 10% preference rate for the next millennium. The population of mortals has slipped to 25% of the total population within one thousand years.

Figure 3

So, we end up with the somewhat strange position where, even though there is an overwhelming preference among individuals born in the next thousand years for the mortal life (90%); the superlongevitists constitute the majority of the population within 500 years, and within a millennium make up about three quarters of the population.

3. Assumptions

        Let us now turn to the assumptions of the model, asking how crucial and plausible they are.

3.1 Availability.

        Our model assumes that superlongevity technology will be available in 2050. The start date here is not particularly important for the logic of the argument; for if superlongevity technology does not become available until 2100, or it is available as early as 2025, this does not affect the conclusion that eventually, mortals will tend to disappear from the population. However, the question of availability seems inextricably linked with the question of affordability. Notice that our initial model makes no distinction between preference and realized-preference: basically the difference between what people want and what people can get. If superlongevity technology is extremely expensive then there may be a large gap between the desire for the technology and the means to obtain it, that is, between preference and realized-preference. The applications of the model above are clearly based on realized-preference. For example, imagine the technology becomes available in 2050, and 50% of the world’s population would use it if they could, but only 20% of those who desire it can afford it. Here, realized-preference would be 10% of the world’s population.

Are there good reasons to suppose that there will be a large gap between preference and realized-preference? The UN model predicts that poverty will still be a major social issue in 2050, which suggests that there may be some gap between preference and realized-preference here. How large this gap is will depend on (among other things) how much of the world remains impoverished when superlongevity technology is developed, and the cost of superlongevity technology. Admittedly, we do not know the answers to these questions, but it is worth commenting on a line of thought that suggests that superlongevity technology will inevitably be expensive. This line of argument, which I have encountered in conversation a number of times, basically runs as follows: “Look how expensive medical treatments are presently, and they really don’t do a whole lot in terms of extending our lives. Superlongevity technology will have to do so much more, so it is likely to be very expensive.” In my experience this line of thought is so prevalent that I propose that we dub this the “iron lung fallacy”. To see why, imagine someone in 1940 reasoning thus: “Look how expensive it is to treat polio victims. Polio victims eat up an enormous share of our health budget, the cost of running an iron lung[4] to keep a single patient alive is staggering. To actually cure polio once and for all is to ask so much more of technology, so we should expect that a cure for polio will be prohibitively expensive for the foreseeable future.” Of course, we know that it costs a few cents per polio vaccination, and that it is much cheaper to treat everyone for the cause of polio in this way than attempting to help patients deal with their symptoms over a lifetime. Similarly, it may be the case that it is much cheaper to treat the causes of aging than it is to treat the symptoms of aging. As I said, we don’t know at this point how much superlongevity will cost, but it is wrong to assume that it will necessarily be expensive. It may turn out to be much cheaper per year to keep a superlongevitist alive than to treat mortals today, precisely because at present we must contend with all the problems associated with deterioration. Moreover, looking simply at health costs is to overlook the fact that one of the costs of choosing a traditional mortal life is that enormous financial resources must be reinvested in the young. So, the cost of any superlongevity treatments would have to be weighed in part against the expense of raising and educating children to replace the dying adult population.[5]
3.2 Stable Population

As mentioned earlier, the UN study estimates that the world’s population will climb to about 9 billion in the next fifty years or so, and then will hover around there at least until the year 2300 (the end of the projection period). For ease of illustration, our model assumed a population of 10 billion. Obviously, the size of the initial population does not affect the general line of argument. If the world’s population is 10 million or 10 trillion, and there is a 50% preference rate, the same decline (as a percentage) of the mortal population would occur as illustrated in Fig. 1, thus, whatever the initial population, mortals will constitute only about 3% of the population within 300 years.

One possible means to ensure a continued presence of mortals is if the population continues to grow, and there is some preference for the mortal life. Consider, for example, a 50% preference rate and doubling of the population every century:

Figure 4

Superlongevitists, in this scenario, grow from 50% of the initial population to level off at about 2/3 of the population. A continually expanding population, then, is one means to ensure the continuation of a mortal population. As a long-term hope for the continuation of a mortal population, the obvious difficulty is that our solar system will support only so many persons (Bostrom, 2003). To support a single person requires some minimum quantity of matter and energy, and so at a doubling rate the resources of our solar system will quickly reach its carrying capacity. Looking speculatively to the future, we might imagine traveling to uninhabited solar systems as a means to continue the population expansion. However, even if some leave for other stars there is a finite amount of resources around this sun to underwrite such adventures, and for every such voyage there will be that much less matter and energy for supporting those that remain. So, even if intergalactic travel is possible, the finite nature of the resources of our solar system leads to the prediction that the number of inhabitants born in our solar system is finite. So, our model predicts that mortals will all but disappear from the local population around the sun.[6]

3.3 Preference for Superlongevity

That there might be some preference for superlongevity seems undeniable. The long standing aspiration to find the “fountain of youth” is some testimony to the desire of many to live a much longer time than is presently permitted by our biology. The argument above suggests that the prediction that superlongevitists will be the largest segment of the population in short order is not as sensitive to preference rate as one might expect; still, we must assume that the preference rate here is not negligible, e.g., if the preference rate were only one in a billion then it would take a ridiculously long time for the superlongevitists to become a significant portion of the population.[7] So, it seems that we cannot avoid entirely the question:  What might the preference rate be?

We noted that peoples’ stated preferences about whether they would adopt some new technology are notoriously unreliable. It might be thought that this means that we should predict that the actual use of superlongevity technology will be higher than what survey results at present might indicate. Here, however, we have to be wary of the problem of the ‘saliency of success’. We should not simply look at a popular technology such as microwave ovens. For this is to beg the question of how widely adopted the technology is likely to be. The relevant class to extrapolate from would be all technical innovations. Here it is less clear that there is an upward trend for greater adoption, for there are lots of technologies that are not widely adopted or underutilized. As mentioned above, jetskis are not widely adopted; and think of all the electric juicers that remain for years in kitchen cupboards, unused. As well, there are tens of thousands of technical innovations that are never commercially manufactured.

        So trying to predict the actual preference rate may prove to be quite difficult. My informal surveys of students, friends and colleagues suggest about a 30% preference rate for superlongevity. Of course such survey results are hardly systematic and, as we have noted, not necessarily a reliable indicator of how people will act. I think there are at least four reasons to think that the actual preference rate will be much higher than this. The first of these is what we might think of as the “backing in” route to superlongevity. The idea here is that superlongevity technology is not likely to be easily labeled “to be used for superlongevity only”. Stem cell technology, for example might be used as an integral part of a superlongevity strategy, but it could also be used to help with a number of ailments that mortals suffer from. Imagine Hiram says that he is not interested in superlongevity, yet when his heart starts to fail at age 70 he avails himself of stem cell technology, which completely repairs and rejuvenates his heart.  When he reaches 90 he has problems with emphysema, and has his lungs rejuvenated with stem cells. At 110 he is diagnosed with Alzheimer’s disease, and again stem cell technology is used to repair the problem. When he is 130 he again finds there are problems with his heart, and again it is renewed with stem cell technology. Continuing in this way there is no principled reason why Hiram should not live indefinitely if every organ and biological system is rejuvenated in this manner. In other words, there is a tension between people not wanting to suffer from some debilitating disease, and not wanting to opt for superlongevity. While we sometimes still occasionally speak about ‘dying of old age’, ultimately we always die from the failure of some organ or biological system. Like Hiram, people may back in to superlongevity simply by not wanting to die from any particular organ or system failure.

Second, superlongevity technology may make converts amongst those who are morally or psychologically opposed to suicide. If we think of suicide as ‘voluntarily ending one’s own life’ then any death that results from refusing to use superlongevity technology looks like an instance of suicide. As a parallel, imagine someone falls and suffers a compound fracture. Instead of having the arm set he refuses all medical attention knowing full well that this could lead to his death in a matter of days; weeks on the outside.  So long as we think of this decision as an act of suicide, it seems that by parallel reasoning we ought to think of the refusal of radical life extension treatment as a form of suicide. Let me emphasize, that I am not claiming here that this is in fact a form of suicide, or that suicide is immoral; I am merely pointing out that some may see it this way, and so provide them with some reason to join the ranks of the superlongevitists. Furthermore, even if one does not think about the moral issues here, psychologically, many find it difficult to commit suicide or otherwise renounce life and so it may be difficult psychologically to refuse radical life extension treatment just as one might find it difficult to refuse treatment for a broken arm.

Third, and what may turn out to be of greatest influence, there is the issue of human bonding. So long as one’s loved ones are around there will be at least some pull not to choose the mortal life. Grandparents may not want to die before seeing their grandchildren grow up and have children of their own. But once they have great grandchildren they may want to see their great grandchildren grow up, and so on. Likewise, imagine an entire family unit decides to forgo radical life extension technology, but then there is a single defection: a woman who is both wife and mother in the family unit decides that she would like to try superlongevity. Presumably we would not be too surprised to find that she attempts to persuade her husband to join her in superlongevity, and asks the same of her children. Of course this is not to say that such persuasion will always be successful, but it does seem a likely influence on peoples’ preferences. Indeed, imagine in our example that the wife is not successful in persuading her husband to adopt superlongevity, and he eventually succumbs to the ravages of old age. A consequence of this is that his voice is now silent, while the mother’s voice (because she has availed herself of superlongevity technology) continues to be heard, and so, still has the opportunity to influence her children to become superlongevitists—the “saliency of survivors” phenomenon. Furthermore, there is the snowball effect: as the superlongevitists population grows it may become harder and harder to resist the influence of human bonding. In the extreme, imagine all your friends and family have opted for superlongevity. It is easy to imagine that there would be enormous social pressure to adopt it for yourself.

Finally, we may see an increasing rate of preference due to “adaptation”. Children born after superlongevity technology is available may be more comfortable with the idea of superlongevity than those of the first generation. Historically, people are more likely to adopt a technology that they are familiar with from childhood. The same may be the case with superlongevity technology.

So while we don’t know what the actual preference rate will be, we can see that there are good reasons to suppose that it may be quite high, and potentially increase over time as more people use it and individuals are born into a world where it is already in use.

3.4 No Defection

Our model assumes that there will be “no defection” among the superlongevitists, that is, that none of them die. However, it seems almost certain that the “no defection” assumption is false. After all, superlongevity technology is not a guarantee against death; at best it promises to reduce or eliminate death due to disease and age related causes. Of course there are many other ways to die. Death due to accidents and violence are still possibilities. As is the intentional taking of one’s own life. The question then is how much these other causes of death would affect the superlongevitists’ population.  Let us assume that the chances of dying from some non-disease factor (where we assume that aging is a disease) in any year is 0.2%, which is extremely conservative [ref]. Even with this conservative assumption, and a conservative preference rate of 10% it still turns out that superlongevitists will compose 30% of the population within 300 years, as Fig. 5 indicates.[8]

Figure 5

3.5 Predictions

Before hazarding a guess about the composition of our future population we need to discuss one more question: who will raise the children of the future? The importance of this is apparent when we think that on average children are more likely to adopt the customs and behaviors of their parents. So, we should predict that children raised by mortal parents are more likely themselves to choose to be mortal than children raised by superlongevitists. So suppose that it is decided as a matter of some policy decision that mortals are allowed to have children while superlongevitists are not. To access superlongevity technology one must sign a “non-proliferation pact” and undergo a vasectomy or a tubal ligation (Walker, 2003).  (I am not suggesting that such a policy is a good thing, I propose it here merely as part of a thought experiment). This may seem like the best hope for the continuation of a mortal population, for in these circumstances it seems that the mortal population might have a relatively low defection rate. But how low? One of the most successful groups in terms of retention is the Amish who are able to retain 85 to 90% of their members; so an overwhelming majority of Amish children stay with the Amish community as adults. Even if mortals could be as successful as the Amish in their retention rate, still this would mean that the 10% defection rate would accumulate as in Fig. 5. So, even under these circumstances it seems that superlongevitists will tend to become an increasingly large portion of the population. However, the possibility of a large population of mortal parents engendering this sort of retention rate seems unlikely. The Amish, for example, have somewhat unique circumstances in the rather insular nature of their communities. This insular nature seems important since the defection rate is approximately 50% higher (it jumps from 10 to 15%) for those families that live closer to towns. The thinking is that children (and adults) living closer to towns with non-Amish populations have more contact with alternate possibilities. Since it seems unlikely that the world’s population could be organized along the rural model of the Amish, we should predict that in all likelihood the defection rate among children raised by mortals will be much higher.

        Thinking about these different factors in conjunction leads to the following predictions. I believe that initial preference for superlongevity technology might be somewhere at least in the 30 to 50% range. As noted above, I suspect that there will be some initial opposition in the first generation to this idea from some of those who are unaccustomed to it. For the reasons listed above, for the first generation that grows up in a world where there are superlongevitists I suspect that the preference rate will be even higher, perhaps 80 percent or more. If this is the case, this would mean that within the first two generations superlongevitists would be the overwhelming majority of the population. Again these are very risky predictions since there are so many unknown variables here.

The safer prediction is that superlongevitists will inexorably compose a larger percentage of the population over time. This prediction holds even if mortals, like the Amish, have a very high retention rate.


4.0 Ethical Implications

The discussion thus far has been predicated on the assumption that individuals will be free to choose radical life extension technology when they can afford it. However, it may be wondered how it is possible to know that this will come to pass. How do we know that the worlds’ major religions will not react in a strong negative manner to this technology? How do we know that individual countries will not pass laws to prohibit individuals from making this choice? The short answer here is that we don’t. If some countries pass laws that prohibit (at least some of) their citizens from adopting this technology then clearly the analysis will not apply in such cases. However, the analysis would still apply to those countries that do not pass such laws. Obviously, if a world government is able to enforce laws that prohibit all citizens of the earth from living beyond 122 years (the length of the current record holder) then the analysis does not apply. Nor does the analysis show that an undetected comet won’t come crashing into earth tomorrow destroying all human life. Any projections to the future are saddled with similar limitations: the UN projections about the population of the World to 2300 have to make similar assumptions, as do environmental models of global warming. So long as we are cognizant of these limitations, such models and predictions may help us think about policy decisions we should make.

How might the analysis be applied to policy decisions? One thought is that some sort of “preemptive affirmative action policy” ought to be instituted in order to ensure the continued survival of a significant mortal population. For example, suppose it is discovered that there is a 50 % preference rate for superlongevity, a policy might be introduced to ensure that the superlongevitists never exceed 50% of the population. Our previous analysis shows a serious potential complication here. Since the preference does not abate, the demand for superlongevity will outstrip the supply of spaces available. Looking at Fig. 1 again, we see that in the second generation 2.5 billion more people will desire superlongevity over and above the spaces available.[9] How would the 5 billion spaces be justly distributed among the 7.5 billion who desire them? If it is “first come first served” then the 5 billion original superlongevitists would hold all the spaces available, meaning that no one in the second generation could enjoy superlongevity. Perhaps some lottery could be instituted, but this would obviously entail refusing superlongevity to 1/4 of the population each generation. Certainly not an attractive option and indeed, one might wonder how politically stable such a policy would be, given that those that lost the lottery might see themselves as being condemned to die in the name of affirmative action.

Even to contemplate such a radical policy seems to presuppose that there is some great moral good to be preserved by instituting it. What might this good be? Is there something intrinsically good about the mortal life that is worth preserving? To see one possible argument, let us suppose that there is only a 10% preference rate for superlongevity. As we have said, even with this low preference rate, eventually superlongevitists will constitute a majority of the population and eventually compose (almost) the entire population. But doesn’t this show us that there is something inherently wrong with a policy that allows people to choose whether to become superlongevitists or not? For even though this form of life is seen as unsatisfactory by most who face the choice, nevertheless it eventually becomes the dominant preference in the population. So, it may be thought that something has gone wrong here, morally speaking.

But what exactly has gone wrong? This is harder to pin down. Consider an analogy; suppose 90% of some population prefers beer to wine. Through some tragic accident a virus is accidentally introduced into the beer and wipes out the entire beer drinking population. Surely it would be wrong then to force 90% of the wine drinkers to start drinking beer. After all, we might be inclined to say that wine drinking is right for them even though we suspect that if new persons are born, 90% of them will choose beer over wine. Here it seems wrong to override the preferences of actual persons in favor of “possible persons”—hypothetical future persons. Similarly, it does not seem right to think that superlongevitists should have their preferences overridden simply because we think that 90% of possible (future) persons would choose differently.

Of course this does not provide us with any positive argument for thinking that superlongevity is a morally good option; however, I believe that such an argument can be made. The argument, roughly, is that superlongevity best accords with our understanding of the ‘good life’.

Let us think first then about the ‘good life’. In the history of philosophy, there are two schools of thought on how we should understand ‘the good’, and the ‘good life’ (Sumner, Hurka).[10] One answer, sometimes referred to as ‘welfarism’, says that what makes a life good is the quality of the experienced life. There are a number of recommendations about exactly what counts as ‘the quality of the experienced life’ including ‘pleasure’, ‘happiness’ and ‘preference satisfaction’. The other major contender is ‘perfectionism’, which says that what makes a life good is the extent that we realize excellence in some mental or physical endeavor. To see the difference here it might be useful to think of a concrete example. Imagine that Abigail is a very promising mathematician just completing her doctoral work. She is trying to decide whether to take up an offer from a prestigious university or to take over her uncle’s tavern. She knows that she has ample talent for mathematics, and she may even make some great discoveries. Yet, she finds the academic life does not make her happy. On the other hand, she knows that she has been extremely happy working in her uncle’s tavern during breaks from school. Abigail does not think that she will be any more than adequate at running a tavern—it is not as if she thinks she possesses a special aptitude for business in the way that she does for mathematics. But this does not dissuade her, because the attraction to running the tavern is simply the sense of well-being it brings her. What should Abigail do? Perfectionists like Plato, Aristotle, Aquinas, Kant, and Marx would enjoin Abigail to pursue her special talent in mathematics and take up the university post. Welfarists like Bentham and Mills would recommend the life of a tavern owner.[11]

I do not propose to solve this longstanding controversy between welfarists and perfectionists, but rather, argue that on either conception, superlongevity tends to promote the good. Let us take welfarism first and assume that happiness is our measure of welfare—that is, the happier people are the better off they are. There are two arguments that suggest that superlongevity will promote happiness. First, there is extensive empirical research that indicates that the elderly amongst our population is the happiest segment, which clearly contradicts ageist stereotypes that see the elderly as “cranky” or unhappy (Overall, 2003).[12] Of course, we do not know for sure whether this trend will continue, e.g., whether the fact that 60 year olds tend to be happier than 40 year olds means that 160 year olds will be happier than 60 (or 40) year olds. Still, the most straightforward extrapolation from our current knowledge is that the elderly should on average be happier than the young. Second, the fact that the elderly may be happier is supported by consideration of a possible self-selection process: we know that suicide and unhappiness are correlated. If, as seems likely, the unhappiest segments of the population commit suicide (or refuse to obtain superlongevity treatments) then they will tend to be eliminated from the population. In a slogan: survival of the happiest. So, superlongevity will tend to make the population happier on average as compared with an exclusively mortal population. So, given that we understand goodness in terms of peoples’ level of happiness, it seems a future where individuals have the option to become superlongevitists is morally better. 

        Perfectionism also favors superlongevity. The basic argument here is that to achieve excellence takes time, and so the more time individuals have the more likely they are to achieve a higher level of excellence in a broader range of activities. Here it might be best to first contrast our world with the fictional world of “Logan’s Run” where individuals are executed on reaching age 30.[13]  It seems obvious that dying at such a young age will mean that much in the way of achievement would go unrealized. Think, for example, about mathematics and physics. These two disciplines are known as ones where the relatively young can achieve great breakthroughs. However, the qualification here of ‘relatively young’ is important. Andrew Wiles was in his early forties when he proved Fermat’s last theorem; and Einstein was 36 when he published his general theory of relativity. If either Wiles or Einstein had lived in the Logan’s Run universe, their personal achievements would probably have gone unrealized; as would have Lance Armstrong’s record-breaking multiple winning of the Tour de France run been cut short if he were dead at thirty. Indeed, we do not need to invoke such outstanding levels of achievement to make the point: think how few people would complete a PhD in the Logan’s Run universe, given that the average age of completion is close to thirty. And the idea of striving for a high level of achievement in a number of areas, say playing the violin in a chamber group and earning a black belt in karate in addition to (say) advanced degrees in medicine and biochemistry would simply be unheard of in the Logan’s Run universe.  As perfectionists we must lament their short lives because it does not allow nearly as much in the way of achievement in comparison with our world, for when we look at their lives in aggregate it would pale in comparison to ours. It is readily apparent that they simply do not have the time to achieve what we can with our greater lifespan. Similarly, imagine what level of achievement would be possible if we could live hundreds of years (or more). Einstein is reputed to have said, “It’s not that I’m so smart, it’s just that I stay with problems longer.” We can put to one side the questions of to what extent it is true that he is “not so smart”, and think about staying with problems longer. Einstein spent much of his life attempting to formulate a “theory of everything”. He was unsuccessful in this endeavor, but one can only wonder if he had had another hundred or two hundred years to work on the problem whether he would have succeeded. And of course the point about achievement does not apply simply to geniuses like Einstein. It is simply ageist stereotypes of the elderly that suggest that as a group they are bored, or inactive, or no longer seeking achievement (Overall, 2003). Imagine what could be done with the help of the full vigor and health of a youthful body even when we are hundreds of years old. Any of us with more time at our disposal might achieve more in a variety of different areas. One could start their fourth PhD program at age 140, and climb Mount Everest at age 180, learn to fly a plane at age 220, become a master chef at age 260, etc.

        It is important here to caution against a potential misunderstanding. The claim is not that a world of superlongevitists would be one where the population is hyper-motivated to achieve excellence, any more than it is to say that we are hyper-motivated to achieve in comparison to the denizens of Logan’s Run. In fact, the opposite might be true. Perhaps knowing that one will die at thirty means that one will be more motivated to achieve more per unit of time compared with us. Even if this is the case it still would not show that the Logan’s Run world has higher levels of achievement than ours. At best this would show that by age thirty they may have achieved more than we have by age thirty, but since we have so many more years to live we can outpace their level of achievement. Similarly, given the comparatively brief length of our lives, perhaps we are more motivated to achieve per unit of time than superlongevitists. (I suspect that they will be motivated in much the same way we are—by looming deadlines!) However, perfectionism is not worried about such matters, but with the actual level of achievement. On average we should expect the level of achievement of our world to fall short of that of the superlongevitists given that they will have so much more time to pursue achievement in intellectual and physical activities. Superlongevitists, after all, would have time to earn a dozen PhDs, learn to play every symphonic musical instrument, and become proficient at every summer and winter Olympic event. Such possibilities are not available to those who live but “three score and ten”. So, clearly perfectionists should enjoin us to adopt superlongevity technology as one way to further our level of excellence and achievement. Therefore, whether we think of ‘the good’ in terms of welfare or perfection, superlongevity will tend to promote the good.

        So, we have reached the pleasant conclusion regarding our future: what seems almost inevitable is also good, morally speaking. Of course, to say something is inevitable is not to say that we have no influence over its timing. We should bear in mind too that to the extent to which we do not attempt at the present to help usher in this future as quickly as possible reflects badly on us.


References

de Grey, A. 2005. SENS (Strategies for Engineered Negligible Senescence) http://www.sens.org/index.html

Freitas, R. 1999. Nanomedicine. Landes Bioscience

Hurka, T. 1993. Perfectionism. New York: Oxford University Press.

Kenyon, C. 1996. “Ponce d’elegans: gentetic quest for the fountain of youth”. Cell 23 84(4):501-4

Overall, C. 2003. Aging, death, and human longevity: a philosophical inquiry. California: University of California Press.

Shostak, S. 2002. Becoming Immortal. New York: Albany Press.

Sumner, W. 1991.  “Two Theories of the Good”. Social Philosophy and Policy 9, 2 (Summer), 1-14

United Nations. 2002. http://www.un.org/esa/population/publications/longrange2/WorldPop2300final.pdf

Walker, M. 2003. Killing the Physically Immortal: the Ethics of Prohibiting Access to Life Extension Technology.  http://www.permanentend.org/Immortality.html

——————2005. Apologism, Prolongevistism and Utilitarianism. www.permanentend.org/SuperlongevityandUtilitarianism.html

Notes

[1] Of course there is some question as to how much is “contemporary hype” and whether there are other factors at work here, e.g., perhaps evolutionary advantage in looking young. I take no position on this question.

[2] There is a poll on this subject on the Better Humans website: However, the results indicate that only 12% would not utilize the opportunity to use technology to radically extended their lifespan. The poll can not (nor does is claim to be) considered to be a reliable indicator of how the general population would respond to the poll question.

[3] In what follows I shall mean by ‘preference rate’ the probability that an individual will choose or refuse superlongevity technology. ‘Preference rate’ then does not refer to a property of a population, but of individuals.

[4] Iron lungs were large devices (about the size of a small car) designed to respirate polio victims.

[5] There is too the cost of having “non-productive” members of society—children and those infirmed by age.

[6] A prediction for the universe itself would have to take into account the question of whether the universe is infinite or finite.

[7] Indeed, they may never. See the point about defection among superlongevitists below.

[8] The numbers here are derived by assuming that 14% of the superlongevitists population will die every 70 years due to non-disease reasons such as homicide, suicide and accidents.

[9] I am ignoring here defection from the superlongevitists’ ranks.

[10] This needs to be qualified: we are speaking about a single notion of ‘the good’. Some theorists adopt items from both theories. We won’t worry about this qualification here.

[11] We are considering here the good only in relation to Abigail’s life. If we had to consider others then the recommendation might be different. For example, if Abigail was likely to make mathematical discoveries that might help make many others happy then some welfarists would recommend this life for her, even though she personally would not be as happy.

[12] For this evidence and an expanded version of welfarism and superlongevity see Walker, 2005.

[13] In the original book the age of death is 21.


Mark Walker Ph.D. serves on the IEET Board of Directors, and is Associate Professor of Philosophy at New Mexico State University, where he occupies the Richard L. Hedden Endowed Chair.
Print Email permalink (2) Comments (7187) Hits •  subscribe Share on facebook Stumble This submit to reddit submit to digg


COMMENTS


Hi Mark,

Thanks for the thought-stimulating article!  Some comments:

In the population model presented, it seems as though the replacement rate would be applied to EVERYONE, not just the mortal.  There is no assumption given that long-lived individuals won’t breed at the replacement rate.  So if reproduction continued only at the replacement rate, there would be an increase in population as long-lived individuals had replacement
offspring but did not die.

There is at least some reason to speculate that very long-lived individuals may actually reproduce above the replacement rate.  Assuming that such individuals retained the capabilities of the relatively “young”, I would not be surprised to see second or third relationships yielding children as a way of further strengthening a relationship, or even out of a desire to see how a different pairing results in differences in children.  After all, if one might live for a thousand years, the 20 or so years that it takes to raise a child may seem pretty trivial, to be undertaken whenever it seemed interesting.

If one then chooses to hold the population constant (that is, no growth; which seems like a relatively arbitrary decision given no other information) then we get in to a situation where whatever factors are conspiring to keep the population steady force increased pressure on members of the population, potentially leading to revolts on one side or another (or the creation or exacerbation of factions not strictly aligned on the long-/short-lived spectrum) resulting in people being selected to die in some way (lottery, war, etc) to relieve the pressure.  Worth keeping in mind is that pressure is likely to be applied across all individuals living at the time, not just the mortals.

One thing one might consider to help relieve this pressure would be to offer individuals large cash payments to maintain a “short-lived” life.  This, on the face of it, solves very little in that as long as the long-lived continue to breed they will continue to put pressure on whatever is acting to stabilize the population (resources, for example).  Since it seems likely that individuals would not be considered old enough to actively decide to live a short life until after the age where a short-lived individual would be able to conceive and care for children, it is therefore unlikely that rules that a free society are likely to adopt would be abe to successfully curb the expanding population.

Speaking of resources, things get complicated when one talks about resources too far into the future assuming that individuals are undertaking bio-modification of some sort or another.  Given the present rate of technological progress, and the fact that life extension body modifications are likely to be developed concurrently, or at least only slightly ahead
of modifications for less dire reasons (such as improving the capabilities/efficiencies of the body system).  It is therefore likely that whatever modifications humans undertake to themselves is likely to significantly alter the resources required to sustain an individual body.  This may come from such abilities as to reconfigure ones body to suit a situation, which may allow an individual to adopt a much less resource heavy physical construction until need for a more resource-heavy body is required.  This could alternately be accomplished by separating one’s “consiousness” entirely from any physical presence,
again, except when needed.  Inded, if consciousness is able to be separated from a physical body and Moore’s law keeps moving along for several more generations, then an “individual” may take up very few resources at all (and any further efficiencies discovered would likely be easily propigated to all other “individuals” existing at the time).

I

Interestingly, if superlongevity does come about, it seems very likely that views on suicide will need to change at least somewhat.  If it is theoretically possible for someone to live forever, we are likely to have to respect the will of some percentage of this long-lived population to be done with their living experience and either opt out of further life-supporting care or, equally likely, actively terminate their mortal existance.  I wouldn’t be surprised to find, hundreds of years hence, age of consent laws that treat suicide as akin to drinking: you must be old enough (say, 50 or 75) to be considered to have enough experience to make an informed choice.

It would be interesting if the backlash over the ability to live for a very long time was an increase in individuals not opting for life-saving procedures of any kind.  The dissonance that is likely to arise between people who have chosen to become very long lived versus those who see it as antithetical to their religious/moral beliefs seems likely to cause at least some fracturing along these lines; it will be interesting to see what causes people to take sides.  Even if the whole process of becoming long-lived unfolds over a long period of time so that individuals don’t have to confront the idea of suddenly having a super-old population, I wouldn’t be surprised to find a group of “young” people who become outraged by the idea (perhaps with a return to fundamentalism such as we are seeing now among some groups of people here and abroad) and gain traction on the idea as part of some larger dogma attractive to those who have not yet had to face the decision of living just a little bit longer.





Mark,

You have put forward a very intriguing article that gives super-longevity some focused projections into the future.  This issue has been on my mind a great deal as I firmly believe in the benefit of enhanced humans via technology.  I do, however, believe that indefinite life spans are a dubious advance when seen in comparison to the system of which we are a part.  Whereas AI and improved quality of life can help us contribute to the larger system, near-immortality does not.  Here’s why:

1. We are part of a larger evolution and evolution is of greater necessity than individual happiness.

2. Evolution is predicated on life-death cycles

3. Without death evolution is slowed

Again, I wish to stress that I believe in the continued evolution of humans and our surroundings.  The progress of the latter is inevitable, while, due to free will, that of ourselves is not so secured.  We are connected to our environment and our environment changes.  We are participants of something larger than each individual’s contentment.  Each individual is an important contributor to the whole, but without a connection to the whole we lose relevancy.  Since nature must continue to evolve by resolving the endless variations presented to her on her journey, so must we.

We sever this critical connection when we detach ourselves from the life cycle.  The life cycle is intrinsically a cycle of change that weaves its way between stability and instability, absorption and expression.  Each of these stages of life mirrors the development and critical contribution that every entity in the universe exhibits within its respective home system.  The notion of an individual is a convenient yet scale dependant human construct.  The individual is by design a temporary entity manifested out of inherited diversity to temporarily express possible solutions to currently relevant hindrances to evolution.  Technology is just such an example of a solution.  A deathless society is not.

Without death evolution slows down since the individual, no matter how liberal, has a vested interest in preserving his/her surroundings to suit the patterns that that individual has habituated.  The brain itself is dynamic, but over time its pathways cohere upon themselves and become entrenched.  This is not the work of a single gene or hormone that can be turned on or off.  It is a fundamental property of conscious development.  Consciousness is predicated on behaviors being built upon behaviors and these behaviors are contextual to a given time and place.  In other words, they are assumptions that are ingrained in a person that have particular temporal and spatial relevancy.  In order to maintain evolutionary vitality the individual must absorb that time and space, challenge it, resolve it, manifest it and then leave that manifestation to be resolved by future “individuals” that have future relevancy.  It is a dynamic process and to try and hold in stasis those pathways and habits is to hold the world hostage.

I respectfully disagree not just with you, but with a good majority of the transhumanist community of which I proudly consider myself a part.  I have been waiting for the right moment to engage this debate and I felt that your very clear, concise and thought provoking article was a good place.  I believe that greater quality of life and an enhanced ability to perceive and process our world is a worthy goal.  I have as of yet, however, not heard compelling reasons why a deathless society is good for our overall evolution as a species and a planet.  Happiness is a comforting notion, but I believe that sacrifice for the greater good is just as important and deserves at least equal consideration.  And as far as perfection is concerned, well, that’s relative to the time and place such a purported perfection is exerted.  And ultimately, no matter how long we live, we can not stop the world around us.  Would we really want to?





YOUR COMMENT (IEET's comment policy)

Login or Register to post a comment.

Next entry: From Heaven to Doomsday: Seven Future Scenarios

Previous entry: The Illusiveness Of Immortality

HOME | ABOUT | FELLOWS | STAFF | EVENTS | SUPPORT  | CONTACT US
SECURING THE FUTURE | LONGER HEALTHIER LIFE | RIGHTS OF THE PERSON | ENVISIONING THE FUTURE
CYBORG BUDDHA PROJECT | AFRICAN FUTURES PROJECT | JOURNAL OF EVOLUTION AND TECHNOLOGY

RSSIEET Blog | email list | newsletter |
The IEET is a 501(c)3 non-profit, tax-exempt organization registered in the State of Connecticut in the United States.

Contact: Executive Director, Dr. James J. Hughes,
Williams 119, Trinity College, 300 Summit St., Hartford CT 06106 USA 
Email: director @ ieet.org     phone: 860-297-2376