Who Should Exist?

The question of what people should exist seems complex and subtle, but basic economic theory suggests it may get a lot simpler in the future. Let me explain, via three possible questions.

1. First, consider the binary question, “should creature X exist or not?” Economically, creature X should exist if it wants to exist and it can pay for itself. That is, in a supply and demand world, if our only choice is whether X should exist, then an X that wants to exist should actually exist if its lifespan cost of resources used (including paying for any net externalities) is no more than the value it gives by working for others. To reject the existence of such a creature is to reject an efficiency gain, i.e., a way to potentially make everyone better off. (These costs and benefits are of course marked at market prices. A creature’s value might include donations from other creatures who valued its existence.)

2. Next, consider the question, “Which humans should be created today?” This question is complicated by the fact that each human likely to exist can be created by only one particular set of parents, and then only if the conception lottery goes a certain way.  Once a new human exists there remains the question of what resource endowment (positive or negative) the kid should get from its parents. Since this creation scenario is far from competitive, supply and demand doesn’t get us very far in analyzing it.

3. Finally, consider the question, “Which creatures should be created?” in a future where factories can make a wide range of creatures. This situation might arise with whole brain emulation, or advanced genetic engineering.  Imagine a supply-and-demand world where many similar competing profit-seeking factories can each make many possible creatures with great precision, endowing them with any preferred debts or rights, but aren’t overly limited by intellectual property rights. When creating creatures is such a competitive industry, supply and demand has strong implications.

All creatures would be created that could clearly pay for themselves (including intellectual property license fees minus existence donations). Since there are vastly more possible creatures than room for actual creatures, costs to exist would be prohibitively high. Most new creatures would have designs near the peak of factory profitability, and own little surplus relative to their cost. Residual control rights (e.g., “are they slaves?”) would rest in the hands of whomever could squeeze the most market value from them.  Yes a few would get lucky and become rich enough to have slack for leisure and existence donations; but theirs would be only a small fraction of total wealth. And in a supply and demand world, this distribution of existence, control, and wealth would be Pareto-optimal, economically efficient, and hence good for all the usual reasons.

Another exception to these creature patterns could be due to ancient legacies, of those who held large initial endowments before this competitive regime began. The designs of such ancient creatures, and of new creatures they favored with existence donations, might be unusually far from the peak of factory profitability. Other creatures might question the legitimacy of special creatures who would not exist if not for such legacy assets.  They might complain, “Why do such legacies get to be apparent exceptions to the general rule that creatures must pay their way to exist?” Of course if legacy assets were deeply entrenched in social institutions, yet represented only a tiny fraction of wealth, these might remain mere complaints.

Tin-pot dictators and supporting elites often keep their nations poor and inefficient out of (often valid) concerns that efficient economies might no longer tolerate their grabbing such large wealth fractions. Similarly, you might fear you would lose relative power in the above scenario of efficient future creatures. So you might prefer an inefficient legacy control scenario, where your generation coordinates to finely control of all future economies, to be tin-pot dictators of the future.  You might try to prevent the creation of these efficient creatures, in favor of creatures you decide should exist, serving you or not as you choose.

If asked what gives you the right to prevent the existence of creatures who could fully pay for themselves, you might respond that you need no right, if you have power and a will to use it.  Or perhaps you’ll say ethics assures you it is simply impossible to be unfair to creatures who don’t yet exist.  But wearing my efficient economist hat, I cannot support such naked selfish aggression, even if I thought it would work.  And knowing how hard is coordination, I have serious doubts re feasibility. If you can identify large negative externalities, I will help you to find ways to price them, to discourage the creation of creatures who cannot fully pay for themselves, and the theft of legacy assets.  But if not, I prefer to help creatures who can pay for their existence obtain that exquisite treasure.

Added 1Sep: Let “wants to exist” be “would want to exist if it existed.”

GD Star Rating
Tagged as: ,
Trackback URL:
  • Eadwacer

    A: All life is pain. We would be better off if we were never born at all.
    B: Yes, but how many people are that lucky?

    • If you think life is pain, there is always suicide.
      “The thought of suicide is a great source of comfort; with it a calm passage is to be made across many a bad night.” – – Nietzsche

      • Hedonic Treader

        Suicide doesn’t prevent existence. It only limits the total region of space-time of a suicidal person’s existence. It can never undo past pain and suffering, either.

        No non-existent entity can consent to their coming into being. And unless we invent super-duper biotech interventions, all life implies unpleasantness, and virtually all life implies non-negligible risk of severe suffering. The option of suicide does not negate this ethically relevant fact.

  • candy

    Good post. It’s a little upsetting to think that creatures of the future will be made as cheaply as they can be. I hope things aren’t too bad for them.

  • Economically, creature X should exist if it wants to exist and it can pay for itself.

    You’ve argued that under a competitive scenario, factories will only create creatures who can pay for themselves, but have not give an argument that such factories will only create creatures who want to exist.

    I think all else equal, such factories will actually want to create creatures who are indifferent about their own existence. The reason is that as technology advances, it will be more profitable for a factory to create/clone new generations of more productive creatures than to upgrade existing creatures. Creatures who are indifferent about their own existence will waste no effort defending against being “recycled”.

    • Nate

      In that case, one might instead choose to think of the factories as the “creatures” and their creations as functional pieces of that creature. For example, humans are creatures that create white blood cells for their own ends.

      It seems what this post is getting at is, if markets tend to become more competitive, what is the upper limit of competition in the biological market?

    • Tim Tyler

      Most creatures that we see want to exist. Those who don’t want to exist don’t defend themselves against being eaten by others – so usually they don’t exist for very long – and few observe them. Much the same seems likely to be true in the future.

      • John Maxwell IV

        OK, how about creating creatures that want nothing other than their employer’s profitability? So they’d self-preserve insofar as that helped their employer profit-maximize.


      • Tim Tyler

        John, that sounds *much* more sensible than the idea that future creatures are likely to be “indifferent about their own existence”. Most agents can’t accomplish any of their goals if they do not exist. Even Asimov’s laws specified self-preservation.

      • xxd

        Plenty creature exist today that don’t defend themselves against being eaten. They are domestic cattle who have outsourced their defense to humans. Humans eat them but they are protected from being eaten by other predators instead of humans.

        Outsourcing specialization is nearly always more efficient. In our economy, we have outsourced the military and the police to the government who (hopefully) are lower cost than everyone having to do it for themselves.

  • Wei, creatures would fit most comfortably with a norm that creatures should exist if they can pay for themselves, but not if they can’t, if such creatures actually wanted to exist conditional on being able to pay for themselves. Then when they lived past their usefulness they would no longer want to live. They would not want to live if they were a burden on others.

    • What do you mean by “being a burden on others”?

      Suppose I’m a factory-produced creature, and during my lifetime I managed to earn X dollars. Now I’m obsolete and can no longer compete with newer models in the workplace. With my X dollars I can live Y years in retirement, and ordinarily I wouldn’t be considered to be a burden on others.

      But why would my factory have designed me to want to keep living? If it instead programmed me to return to the factory for “recycling” then it would be X dollars richer.

      More generally, why would a profit-maximizing factory create creature with any terminal goals (including for existence, conditional or not) other than “maximize the income of my creator”?

      • Robin’s hypothetical was:

        “Which creatures should be created?” in a future where factories can make a wide range of creatures. This situation might arise with whole brain emulation, or advanced genetic engineering.

        So presumably, we don’t yet have the capability to design obedient AI slaves , and so e.g. cannot make them not have terminal goals, but have enough capability to adjust various preferences such as “desire to exist, conditional on being useful”. “Useful” or “not being a burden” meaning benefits outweigh costs (including externalities).

      • Brandon Reinhart

        Wei –

        Wouldn’t the marginal cost of an upgrade just have to be pretty minimal in order to justify the factory producing creatures that want to buy upgrades instead of producing creatures that just want to be recycled? Initially only a few rich creatures would be able to afford upgrades, but a market would exist and over time upgrades would become more accessible.

        Some ems might trade rights for upgrades, indenturing themselves in exchange for continued competitiveness. If ems were able to indenture themselves to a competing factory that offers upgrades, the originating factory might also be motivated to provide its own upgrade plan.

      • Sark, I don’t think Robin really intended to confine our attention to just the period of time after the invention of brain emulation and before the invention of more general AIs. At least, I hope not, otherwise the question becomes not terribly interesting, since presumably the period of time after the invention of general AIs is much longer and potentially can generate much more ethical value.

      • Wei, I agree values would tend to evolve toward not wanting to live longer if one was no marginally longer profitable in terms of living one more time unit.

      • Robin, suppose creature terminal values evolve into just “maximize the income of my creator”. Then wanting to live only when profitable falls out automatically as an instrumental value. (In a sense, they don’t really “want” to live any more than I “want” to wake up at 8 am and go to work.)

        What does your ethical philosophy say about a future where most creatures are like this, compared to one where most creatures care about their own existence as a matter of terminal values? Are you indifferent, or do you have a preference?

      • Wei, there are many possible ways to encode creature values such that they’d want to die if a burden, but live if they can pay for themselves. I don’t pretend to know which ones would win out. The main thing I feel confident of is letting creatures exist who can pay their way.

      • Robin, it seems that we disagree on either

        1. the distribution of evolutionary trajectories of creature values in a competitive scenario, or
        2. the ethical value of such creatures

        Probably we disagree on both but I’m having trouble telling which disagreement is the more serious one, especially given your tendency to dodge my attempts to probe your ethical philosophy.

        I guess we’re not going to resolve the disagreement here today. But can we at least agree that many who disagree with you on this issue do so out of genuine attempts to figure out the right answer to a difficult ethical question, and are not simply engaged in “naked selfish aggression”?

      • Wei, the ethical principle in which I have the most confidence is that Pareto improvements are good. This makes me very suspicious when ethical intuitions tells folks to prevent apparent Pareto improvements for unspecified reasons. Genuine attempts to follow such intuitions can also be naked selfish aggression.

      • Khoth

        Adding new creatures isn’t necessarily a pareto improvement – off the top of my head, the increase in competition is bad for existing people/creatures doing the same thing, and creating intelligent life for exploitation is likely to create a culture in which exploitation is more socially accepted, which is bad for people/creatures with low power.

      • Robin, many of those who disagree with you support the creation of a Singleton as the alternative to free market competition. In a Singleton scenario, an entirely different set of creatures would exist. It’s hard to see how either scenario can be considered a Pareto improvement over the other so we need some additional ethical principles to choose between them. This seems so obvious that I’m entirely confused by your statement.

    • It took me a while to make sense of my own prediction that in a competitive scenario, the creatures who will be created likely won’t care about their own existence. Intuitively, it seems wrong that efficient(*) markets would cause such waste (* putting aside the other problem that Carl Shulman pointed out).

      Eventually I recalled that a competitive market tends to satisfy the preferences of its participants in proportion to their initial endowments. Potential creatures who want to exist have zero initial endowments, so their preferences tend to be ignored in a competitive market.

      To put it another way, to ensure that creatures created are ones who want to be created, we need something like distributive efficiency, but a competitive market by itself can at best provide Pareto efficiency.

  • Andy McKenzie

    Very interesting. What are the implications here for the right to have an abortion? This is really what that whole debate comes down to.

    • These factories create what they intend “with great precision”; when they accidentally make something unwanted, they either sell abort, or continue and sell at a loss, whichever makes for the least loss.

  • Carl Shulman

    “Pareto-optimal, economically efficient, and hence good for all the usual reasons”

    If the legacies prefer not to create particular beings, such that their loss could not be compensated by transfers from the ‘efficient’ beings (thus their refraining from doing so given strong security), then creating the ‘efficient’ beings isn’t even Kaldor-Hicks efficient.

    “discourage…the theft of legacy assets”

    Combined with legacy claims to the resources of the interplanetary and interstellar frontiers, this is the legacy-dominance scenario. With secure property rights in resources (and a quick, easily duplicated ramp-up to near the limits of technology) ‘efficient’ designs are just those that serve the preferences of legacy resource-holders, or in turn serve those who do so, etc.

  • Carl, I don’t understand what basis you imagine folks having for not wanting to allow the creation of creatures who can pay for themselves. Our generation claiming to own the entire universe out there seems pretty extravagant to me; I had in mind more reasonable legacy asset claims.

  • Carl Shulman


    If creatures ‘pay for themselves’ through wasteful efforts to rush to seize extraterrestrial resources not yet protected by property rights, denying others the chance to obtain those same resources (and thus making them worse off), the reasons to object seem pretty standard. They would be similar to the reasons to object to the creation of vast numbers of fishermen who would “provide their own calories” by fishing oceanic fish stocks unprotected by property rights to extinction.

    • I think Carl has nailed the flaws in the reasoning. It will always be more profitable to “game” the system; to exploit flaws in the property ownership system and so produce non-sustainable transfers of property.

      The holders of the legacy resources have a tremendous incentive to never allow changes to the property ownership system that allow for the loss of their legacy control.

      The reason is because there are some components of value that cannot be measured via economics. For example the value of genetic offspring. To the parent, genetic children are of essentially infinite value. To the non-parent, someone else’s children are all of equivalent and finite value.

      The problem comes in trying to match inheritance of property and inheritance of genes.

      At one time that was dealt with by having only the oldest son inherit the property and titles. That was ok, but it provided a tremendous incentive for younger sons to kill the oldest, or for the oldest to kill the younger ones first. That selects for the meanest and most violent. Perhaps a “feature” when the first born has a “divine right” and can do no wrong, but when the siblings have to follow the same law as everyone else, not so much.

      The other approach is incest, to keep genes and property together. That was tried too. The Egyptian Pharaohs and European royalty both suffered from inbreeding. That doesn’t work in the long term either.

      Contraception does allow birth rate to be matched to personal economic growth. For the holders of legacy wealth, keeping their reproductive rate matched to their economic growth rate is feasible, especially when that wealth can be used to “game” the system by manipulating taxation and government spending.

      The problem then becomes that the “system” of property rights is a completely arbitrary societal convention. There is nothing universal or natural about property rights. When the holders of the legacy wealth extract too large a fraction of the wealth being created, then there is discontent in the society and there becomes the incentive for those who are discontented to change the system. Usually that requires the destruction of the holders of the legacy wealth if they do not allow for peaceful change, as occurred in France and in Czarist Russia.

      I think we are beginning to see that now. The birth rate is the lowest it has been in over a century.


      I think this is one of the first signs that the holders of the legacy wealth and those most expert at gaming the system have committed their own “tragedy of the commons”. The exploitation of the majority by the minority is producing real discontent, discontent measured by a falling birth rate and a decline in real standard of living.

    • Carl I agree it might be a shame if there is excess waste from insufficient property rights in distant “land.” I see that as a trade between racing waste and the cost of enforcement. I just don’t see the connection between that and letting creatures exist; letting more creatures exist doesn’t obviously make it harder to enforce distant land property rights.

  • Noumenon

    “Pay for themselves” seems like such a terrible answer to “who should exist.” It’s like asking “What actions should we strive to do?” and answering “Those that make money.” I would start over and say “those who pay for themselves will exist, now who should exist?”

    A lot of people who pay for themselves will produce negative externalities on me, and a lot won’t enjoy their lives as they are just eking out profit for somebody (hopefully me).

  • Interesting. An argument along the lines of The Abolition of Man by C. S. Lewis against the idea that one generation should have the power to determine the fate of all subsequent generations … but this time in favor of genetic (or other) engineering of humans.

  • OhioStater

    The key assumption is the economy is stable and it’s possible to determine who can pay for his existence. If peak oil happens, then a lot of people that could pay for themselves circa 2005 will have a hard time circa 2035.

    • Betting markets can convert uncertainty into certainty, for the purpose of knowing what actions give non-zero profits.

  • Finch

    @Andy McKenzie

    It’s a much stronger statement than that. Robin’s first question seems to imply that it is moral to bring into being any being that will pay for itself (presumably in the long term). So birth control, or even unenthusiastic reproduction, is less moral. Popular moaning about how terrible it is to live in modern civilization aside, I believe this is at least kind of related to deep human beliefs. Even slaves reproduce.

    I.e., following Robin’s logic, the Duggar family are heroes. They are maximizing their reproductive potential and producing beings that will pay their own way through life while throwing off externalities for the rest of us.

    Returning closer to the theme of the comments… As the fraction of these beings that are virtual (i.e., only exist as some sort of computation) becomes high, I don’t think they can be seen as resource limited in a meaningful sense. The rate of population growth would be limited by thermodynamics and the speed of light, but that’s about it.

    I suppose you could argue that you can only expand a sphere at the surface, assuming you have already achieved some maximum density (and assuming that there is a maximum density) so the fraction of the community that can grow is shrinking, and therefore the rate of population growth must eventually decline. But it will always be positive.

  • Cz

    Should there be a place in future societies for people wasting their time on blogging? 😉 Will there be only one homo sapiens species with coherent ethical system or will some groups branch into being something incompatible to replicate with? Will future bring us new fields of human activity after intellectual property (following information/knowledge economy) rewarding of which will have to be included in the law and enforced by the police as legacy people won’t have intuition to exchange their labour for such things. Is it more ethical for you Americans to subsidize reproduction of Chinese or Eastern Europeans by buying their labour? Who prints the currency in the future world? Won’t physical violence disrupt future markets? Would it be possible to discuss the future global societies like this with your market assuptions with all the different cultures occupying this planet?

    Anyway, you ask very good questions. I enjoyed this post a lot.

  • Khoth

    A large population of effectively-slaves living on the barest minimum doesn’t sound like a good future. Do you see an up-side to this that I don’t?

    If you don’t like it either, do you have ideas on how it could/should be prevented? A strengthening of laws against slavery to cover simulations sounds like a good start, but is there more that could be done?

    • Konvkistador

      Most of the agents you refer to as effective-slaves living on the barest minimum are quite likley to be content with their lives?

      I feel compelled to point out that there is no reason whatsoever that the more economical units need be self-aware at all much less share our utility functions or our ideas of what is “living in prosperity”.

    • Khoth, the whole point of doing careful analysis is to do better than just making a knee-jerk reaction to your vague mental image of the situation.

      Konv, yes creatures who want to live need not be self-aware. Why is that a reason to prevent their existence, if they can pay their way?

      • Jess Riedel

        Konv, yes creatures who want to live need not be self-aware. Why is that a reason to prevent their existence, if they can pay their way?

        I figure it would be because maybe our morality doesn’t intrinsically value creatures who aren’t self-aware, since “wanting to live” then probably just boils down to “takes actions which continues it’s existence” (which a plant is capable of).

        Also, I think all of this discussion is confused by ambiguity about property rights for creatures that don’t exists. By having new creatures “pay their way”, do you mean pay back the already-existing creatures for the resources the new creatures consume?

      • Jess, yes, once created, the new creature would pay back others for the cost of its creation.

      • Jess Riedel

        Sorry, I’m still confused.

        You think it extravagant for us currently existing creatures to lay claim to the entire universe. I agree. So when you say “pay their way”, you must mean that new creatures can (1) pay for their own creation and (2) pay any continuing support burdens on old creatures. But I assume you do not mean (3) pay for the value of the comic commons they consume during their life, because old creatures have no claim to that. Is this correct?

        (Also, does “legacy assets” just refer to things owned by previously existing creatures?)

        I guess a better way to put it: I don’t understand what equilibrium situation (specifically, what property rights exist) we are considering when asking whether new creatures should be created.

  • mjgeddes

    Me only. Everyone else is just too damn annoying.

  • Philo

    “Which humans should be created today?” In other words, in the next n hours, for some n between 0 and 24; for definiteness, let’s make it 24. ‘Created’ had better mean conceived; if it meant born, the question would concern the timing of birth of already-existing fetuses (not very interesting). But even if it means conceived, conception is mostly not achievable instantaneously. Some acts of sexual intercourse performed yesterday will result in conceptions today; almost all of these, I think, we will just have to accept: as of now they are inevitable. The question, then, is: “Which *additional* humans should be created today?”

    By the way, the *practical* question for today is: “What sequences of actions should each person do over the next 24 hours?” None of our actions will achieve conception with metaphysical certain, nor will any avoid conception with metaphysical certainty. Our possible actions just have greater or lesser probabilities of resulting in conception (today; toward the end of the 24-hour period this probability approaches zero).

    In answer to *your* question: Those additional human beings should be conceived who would in fact be conceived if everyone did the sequence of actions he ought to do. (And what sequences are those, for each person? Well, that would be a long story!)

    “Economically, creature X should exist if it wants to exist and it can pay for itself.” But economics is unhelpful here. There might be a choice between creature X’s existing and creature Y’s existing, and creature Y’s existing might be better. Then creature X had better not exist even if it would want to and would (not ‘could’ or ‘can’) pay for itself.

  • I think you have passed the limits of useful economic analysis. Or else I need to reread the post a few times. Probably both.

    The point at which we have factories to produce all sorts of sentient creatures is so far removed from the present reality that I would need to develop a picture in my mind of how we got there and what we learned along the way.

    Meanwhile, ya gotta love the liberal philosophy that got the common folk considered as something well beyond capital investment. Perhaps we will have some way to extend the ideas of individual liberty to factory created sentient beings long before overcrowding of the planet with them becomes an issue.

  • You write as if work and leisure will be mutually exclusive. I see that as a temporary artifact resulting from the nature of productive activities having changed faster than evolution can change what we feel like doing. Once most desires are the result of design, there’s no obvious reason not to have leisure-like desires coincide with productivity-inducing desires.

  • It seems like competition would drive existence costs down to the bare metal, and the most market valueable entities will be those with the most novel/efficient brain designs.

    But I don’t see how the factories can make a profit off this unless they can retain exclusive rights to the brain designs. And even in that case other factories could probably replicate the functionality without violating IP, if with were profitable to do that development.

    Ok, in an EMS copying situation, I would contract with a publisher to sell enhanced custom copier of me to customer specs. The publisher would demand exclusivity and agreement to a per-copy debt. I require limitation to only approved modifications, and some promised wage from the customer or other criteria to limit how long I might need to work to pay the debt.

    After I am copied, I have incentive to work for the customer, pay the debt, and maintain exclusivity to the publisher. (letting others copy me would hurt my wages and likely many other copies of me.)

    And my publisher has incentive to develop new efficient safe tweaks to my brain, which means I probably don’t want to self publish.

  • ChrisW

    Surely the ability of a creature to “pay for itself” would be compromised if it had the tendency to devise and persue its own goals? If our creation is to pay us back it needs to work for us, not for itself. The most economically viable creatures would be like the pliant robots of popular science fiction, existing to serve their owners. There seems no economic reason to create truly free creatures.

    Of course there is (in rich countries at least) no economic reason to have children; they are an expense. But our tendency to economic self-sacrifice here is a very special case and unlikely to be extendable to factory-created lifeforms. Our biological instincts to reproduce usually outweigh economic considerations and lead to us producing offspring who will never financially reward us for the effort of raising them. If artificial creatures are produced for gain why would they ever be free?

  • William H. Stoddard

    I’m sorry, but I prefer to help creatures who can pay for their existence obtain that exquisite treasure completely fails to parse for me. Obtaining things is an action; but action is possible only to things that already exist. (And conversely, if something never performs any action, there can’t be any evidential basis for saying that it exists.)

    And the point of value is to guide actions. I don’t see how it can be meaningful to say that anything has value for a notional thing that is incapable of acting.

    I didn’t obtain my own existence; it was obtained for me by my genitors, acting on values of their own, wisely or unwisely.

    Or do you want to say that when they engendered me, rather than engendering the millions of other possible sons or daughters who might have been begotten a few minutes, hours, or days earlier or later, my coming into being robbed all those potential beings of the exquisite treasure of actuality, so that I am in debt to all of them for the very fact of my existence? Really, that’s starting to sound like original sin. . . .

  • Pingback: The utility of existence, and the limitations of economics «  Modeled Behavior()

  • Evan

    Robin, reading your post, and similar ones (the “Dreamtime” series), I am struck by the impression that you seem to regard most of the horror people feel at the possibility of an impoverished, heavily populated world as due to a fear of a loss of status due to poverty (especially in this post when you refer to a desire to be “tinpot dictators”).

    I can’t speak for everyone else, but that is not the case at all for me, to me the reason this future sounds horrible is that sounds boring as all hell. What is the point of creating new creatures to enjoy existing if they spend 24 hours a day, seven days a week doing nothing but boring, tedious work. What’s the point of creating them if they have no time to appreciate their existence because they are too busy working to pay it off?

    I care very little about my relative status, I submit as evidence for that statement the fact that I have forgone many opportunities to improve my status because they would waste precious free time I could be using to have fun and enjoy myself. Why should I pay for cryonics (and let me say that you and Eliezer have convinced me that is a good idea, for the most part) if I’m going to awaken in the future where I’ll never have time to spend time with my family, socialize, watch movies, read books, or anything else fun because everyone is working nonstop to pay for their existence. Your future sounds like a horrible day at the office that never ends.

    Robin, why are you paying for cryonics if you think this will happen to you? I know you’re lucky enough to have a rewarding, stimulating job today, but do you really think you’ll be that lucky again when they thaw/emulate you in the future? Are you paying for cryonics on the off chance that you’re wrong, or am I visualizing this future incorrectly.

    P.S. I am still signing up for cryonics because it is possible that I am incorrectly visualizing this projected future of yours and it isn’t as mind-crushingly dull as it sounds.

  • Is it more ethical for you Americans to subsidize reproduction of Chinese or Eastern Europeans by buying their labour? Who prints the currency in the future world? Won’t physical violence disrupt future markets?

  • Stuart Armstrong

    Or perhaps you’ll say ethics assures you it is simply impossible to be unfair to creatures who don’t yet exist.

    Well, that’s precisely the point (though phrased somewhat one-sidedly). If we disagree on that (which we do) then every other argument we can make to each other is irrelevant.

    But wearing my efficient economist hat, I cannot support such naked selfish aggression, even if I thought it would work.

    Reading your pledge, it all depends on the meaning of “clients” – whether they exist or don’t exist. Back to the previous point, again.

  • Stuart Armstrong

    As for why I don’t agree to take into consideration the “preferences” of non-existent entities, it’s simply because the question “would you prefer not to have existed?” is meaningless. I can kill you, and I can make you not have existed from the point of view of the outside world, but I can’t make you not have existed – because you did. Even if I erase your existence, it still happened.

  • Aaron Armitage

    1) “Should” looks like an ethical question. Why reduce ethics to economics?

    2) Your two answers to “should creature X exist or not?” contradict each other. A idler of inherited wealth, or for that matter a person living off government welfare, can pass the first version — they want to exist and can pay — but not the second. Are we to read the second version as an endorsement of revolution against the wealthy? And if not, why?

    3) In the comments you make it clear that you use “creature” in an inclusive sense, including non- or sub-sentients. It seems clear you’re including animals. Therefore very few humans, and certainly no American or meat-eater, can say he (not “it”) “works for others” nearly enough to compensate for the harm we do. And furthermore we just don’t care. If “creature” includes plants, all animals fail, by definition. Should we first cleanse the natural world of at least predators and perhaps all of our fellow animals, then kill ourselves? If not, then obviously we have some freedom to ignore your principle; why not then opt out of the whole thing in favor of something more palatable?

    4) Your second version of the principle states than we should continue existing if our work for others exceeds our costs (rather like machinery). But the principle is stated in universal form, which would imply that those others in turn justify their existence by working for others. Unless you allow for someone who gets to exist for his own sake, there must be circularity in who works for whom. Either way is an objection. The circularity speaks for itself. If we admit of someone who exists for his own sake, why not, say, everyone with at least a human-level consciousness?

  • Jonnan

    “Economically, creature X should exist if it wants to exist and it can pay for itself. That is, in a supply and demand world, if our only choice is whether X should exist, then an X that wants to exist should actually exist if its lifespan cost of resources used (including paying for any net externalities) is no more than the value it gives by working for others. ”

    Maybe I’ve missed a comment or twelve, but this seems to me to be an impossible standard to have, based in the simple restriction of Entropy.

    Life can only exist in an energy gradient (In our case, powered by the Sun.). The very nature of life is that it takes useful energy and survives, like any other machine, by converting some of that to heat energy – only the sheer abundance of useful energy coming from our sun makes the creation of so many sub-strata of energy possible, from plant’s to animals, microbes, et al. But the nature of the game is that it is a losing game.

    Any life form that passed this test would be a perpetual motion machine.


  • xxd


    Assumption: most profit derives from labor.

    If profit derived from capital exceeds profit derived from labor but labor is *required* to guide the capital then you could have the situation whereby entertainment sims are very likley in direct contrast to your hobbsian view of EMS having to pay for their very existence with no leisure.

    But only if the mediocrity principle is true also. If it’s not then you are most assuredly right in your competitive winner-takes-all scenario.