Morality Should Be Adaptive

Yesterday I said:

Morality should exist; … there should exist creatures who know what is moral, and who act on that.

Many commenters disagreed, yet today I will go further:

Morality should be adaptive; it should help groups survive.

Humans evolved moral feelings as an adaptive response to difficult coordination problems in forager communal living. Culture tweaked those feelings to better fit farming life. Related feelings in other animals evolved for related reasons. So morality evolved to help us survive, and it has been intricately but not infinitely matched to that purpose. If, after a sudden unexpected change in our environment, we apply that morality in such a way as to make ourselves go extinct, that seems a rather dysfunctional broken application of such morality!

Our moral feelings are crude and imprecise – they can have error. Given how complex is our world and crude our minds, and given how weird is our modern world relative to our evolved expectations, we should expect a lot of error. We should not blindly follow our moral intuitions, but should instead correct them as best we can whenever we can estimate a non-zero net error. And if your intuitions suggest that people like you should go extinct, well seems like  a pretty damn big clue of error. Of a BIG error. Correct!

Added 4p: The evolutionary context of our moral intuitions gives a rich detailed framework for defining and estimating moral error. If you reject that framework, the question is what other framework will you substitute? How do you otherwise define and estimate the error in your specific moral intuitions?

Added 21Apr: Richard Chappell comments here.

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

    Morality evolved, therefore morality should embrace evolution. See the fallacy?

    • Adam M

      If we have incomplete information of our own morals, which we do, then knowing that our morals are evolved adaptations is a pretty handy way of guessing at what they might be.

      • Dremora

        In all honesty, I see no way in which this knowledge could help me change my terminal values to more “correct” ones.

        Morality is about emotions, ideally in reflective equilibrium. It’s a matter of taste, if you will, and I find it hard to see how you could be wrong about your own taste.

        If I look at the world and see more uncompensated suffering than I would voluntarily accept or coercively force onto any non-consenting party, then the only way in which my preference for that world to not exist in this state could be wrong would be an epistemic failure in my world model. In other words, I could be mistaken about the actual distribution of pleasure and suffering, but not about my preference that the trade-off stop existing, given that it in fact does exist.

  • Richard Quigley

    I would go even further … ” it should must help groups survive.

  • marris

    I think there’s a bit of equivocation going on. I think many people in the previous post were talking past each other.

    Many people think of morality as something “out there,” either God given or conceptually true statements. They don’t think of it as “rules designed to prevent coordination failures.”

    Instrumental practices should certainly adapt to the environment. After all, the oobjective is given. A practice that fails to adapt to achieve goals is one which is less likely to achieve those goals as the environment changes. So rules that are instrumental for coordination should adapt to improve coordination.

    However, many people don’t think of morality as purely instrumental. They think of it as specifying the good. And the good as the goal. To them, “adapting morality” means changing.the goal away from the (previously specified) good. Which must be “bad” by (the old) definition.

    Anyway, it’s easier to convince that an adaptation is good if you can show that you’re only changing the instrumental parts.

    • Roger

      Building on maris,

      Life requires problem solving. Over time the problems change, as do the solutions. Indeed, the solutions often cause new problems as life expands and flourishes.

      Morality refers to a special class of coordinated, often cooperative problem solving among humans. Once again, over time the coordination problems change and the solutions need to adapt in response to these changes.

      As is true for any solution, morals need to be instrumental — that is they need to actually solve problems. If they create more problems than they solve or don’t work at all, they are objectively dysfunctional. Interestingly, moral solutions work better if people tend to treat them as intrinsically good. I suspect this tendency is just an evolved (culturally and biologically) solution.

      Going back to Robin’s claims…I’m not sure I would word it the way he does, but I will say that since, for living beings solutions are by definition better than problems (they replace a better state for a worse — that is what a solution is), that for a species which can solve more problems collectively than alone, a means to coordinating this problem solving is better than the absence.

      In summary, life is about problem solving. Life creates value, absent life the term has no meaning. Failure to discover value and solve problems leads over time to non life (no problems, no solutions, no value). A social species should solve problems of coordination and cooperation, and these solutions should adapt over time. Robin is right.

  • Poelmo

    I don’t think morality has to adapt. What I do think is that a lot of people are confuse morality with religious/societal notions of “sinning” (which should move with the times), hence people calling pre-marital sex immoral (while this has nothing to do with morality because pre-marital sex is not inherently evil or good).

    Some core tenets of actual morality are universal because they are rooted in logic, like “women are just as smart as men, so not allowing women to have a career means inflicting unnecessary suffering and is therefore wrong” or “nobody wants to be killed and there is no logical reason to give some people more rights than others (or at least, we can’t test for “worthiness”), so the same laws should apply to everyone, which means we should outlaw murder”. Changing these tenets requires changing logic and that’s impossible.

    • Ari

      There’s nothing about morality that is rooted in logic. Even when you can use predicate logic to analyze moral statements it cannot give definite answers on syntheticity of the statements. All abstraction and observations leak. Just revise epistemology. And don’t forget Bayesian interference either, not only for abstracting but also in decision theory.

      There’re many ways to inflict as much as harm as murder as long as you assume transitive preferences. And there’s hardly anything objective (let alone rooted in logic) to tell how much they should cost. Plus things like life and murder are not discrete, which can get as complex as there is entropy, and math to understand it. Assuming even concept of punishment is already framing custom premises. Moral actors don’t need to be divisible. The list just goes on.

    • As a descriptive matter, those beliefs are far from universal. It’s quite a reach to call some competing claims “actual morality” and “universal”.

    • Poelmo writes:

      Some core tenets of actual morality are universal because they are rooted in logic, like “women are just as smart as men, so not allowing women to have a career means inflicting unnecessary suffering and is therefore wrong”

      Nothing in the above statement is a logical necessity nor a valid logical deduction. Judgments of facts and/or values come in at every point of the chain.

      I’d like to be able to believe in an absolute morality, but I’ve never seen a compelling argument for one. All I can say is what *my* morality is.

    • Poelmo

      Yes, taking the standard freshman philosophy textbook position is very safe and very easy. But really, please try a little harder and give some constructive criticism. The only assumptions I had to make in my examples were that humans don’t like being killed and don’t like other people getting preferential treatment for no good reason. Yes, these stances don’t follow from the laws of physics but virtually every human being that has ever lived and will ever live agrees with them (and the ones that don’t are usually psychotic), so do the great apes and most likely intelligent aliens as well (these stances are pretty much an inevitable result of the evolutionary path social sentient beings had to survive).

      • bhw

        Yes, these stances don’t follow from the laws of physics but virtually every human being that has ever lived and will ever live agrees with them (and the ones that don’t are usually psychotic), so do the great apes and most likely intelligent aliens as well (these stances are pretty much an inevitable result of the evolutionary path social sentient beings had to survive).

        A ridiculously arrogant statement. This is to assume that modern western secular human liberal morality is somehow the FINAL, INEVITABLE, end-of-history product that can never be changed that will likely apply to ALL intelligent beings under all possible environments. And you dismiss anyone who could possibly disagree with you as “mentally abnormal” (“psychotic” as you called it).

        You would agree that if a society operates under a different moral set as yours and they manage to function just fine, then you can not claim that your set is superior to theirs, or somehow they will eventually convert to yours, right? Countless of society have survived for much longer than the modern western moral set have ever existed (maybe ~40-50 years).

        The seemingly invincibility of modern social movements seem to have convince people that somehow their belief is an universal inevitable tenets. Just because you have not observed anything that surpassed or will surpass your preferred moral set, does not mean nothing ever will. You realize your statements are just as stupid and short-sighted as some Christians in medieval era declaring that the Christian moral norms and their ideal lifestyle are the end-of-history and inevitable. They also think that anyone who disagreed with them as evil/unexplainable/possess by the devil/people who have not seen the light, which I think is a good parallel to your “psychotic”-label. I think we are in agreement how wrong they have proven to be, by our current world.

        Also, to address your actual moral preferences that are somehow infallible.

        The only assumptions I had to make in my examples were that humans don’t like being killed and don’t like other people getting preferential treatment for no good reason.

        These assumptions do not naturally lead to your tenets. Humans don’t like to be killed, but they might not have such a dislike to kill others. For example, in a moral set that might seem even feasible to us (let’s constrain the possible moral set only to things that seems feasible to us, and not possibly appeal to anything that will ever exists), the leading morality might be that everyone should be responsible for their own defense. If you got murdered, it is your own fault. Perhaps this world might have a stronger evolutionary advantage than our own, such as much faster genetic evolution. Their society might be even more competitive than our own.

        For your second assumption, “don’t like other people getting preferential treatment for no good reason”. The key point lies in who defines good reason. If you get enough people to agree with you that certain traits are good reason of discrimination, then it becomes a good reason, so I doubt that this qualifies as being universal. For example, people who have bad body odor (to be even more objective, let’s say what WE considered to smells bad:) faces a severe disadvantage in society. You could say they can get gels to cover their smell up, but then again there is always a second possible direction you can argue. Why is it that they should be cover themselves up, instead of getting people to accept their smell? They could decide to that their right to their own traits trumps the preferences of society. (Maybe they could even have BO pride parades :).) The point is that the selection criterions on people that we have as a society are never universal and somehow logically objective.

      • Poelmo


        “Our” society is far from the ideal: we still discriminate against LGBT people, atheists and the poor (who are only poor because of the sick economic system we have).

        You cannot discuss my two assumptions separately for the murder thing, you need both of them to outlaw murder. You talk about the definition of “good reason”, which is fine, but you forget that any logical mind can figure out that religious resons are not good reasons, calling on the authority of “nobility” (or wealth) is another example of what is obviously not a good reason (religion and nobility/wealth are the source of more than 90% of all discrimination).

        Can you logically argue that people should put up with bad odors? Yes you can. Can you argue that being women should be disqualified for work? No, you can’t. So go ahead and argue for more freedoms than we currently have, that’s fine and sometimes even the right thing to do, but arguing against rights we already have is usually demonstrably illogical.

        To reiterate the validity of my “assumptions”: ask any person anywhere on Earth whether they want to be killed or not and whether or not they believe others should get preferential treatment and they’ll answer “no” and “well, sometimes, but only when it’s in accordance with my complicated belief-system (usually a form of religion, that leaders need to justify giving some people preferential treatment against the inherent will of all other members of society)”. I can guarantee you that no one will say “sure, give some people preferential treatment, I don’t care”, instead they’ll always cite some complicated belief-system to base their exceptions on.

  • richard silliker

    How do you otherwise define and estimate the error in your specific moral intuitions?

    And the answer is ……..AMBIVALENCE. Defined as concurrent feelings of like and dislike.

  • Mark M

    Humans evolved moral feelings as an adaptive response to difficult coordination problems in forager communal living. Culture tweaked those feelings to better fit farming life.

    Can you prove that?

    I’ve heard it said that “We have as many morals as we can afford.” I don’t know who said it, but I agree, which means I agree that morality is adaptive.

    The whole should be part is kind of nonsensical. Morality is an abstract concept that doesn’t exist outside the context of those who can experience it. Morality is, almost by definition, what we experience it to be.

  • Dave

    In prosperous times such as now, where people are well off but not equal, morality dictates that the weak are entitled to the amount of good that will make them equal to the strong on a per –capita basis. (True Democracy)

    If that is not happening to perfection someone, apparently for their own entertainment and at no cost engages in self-righteous finger pointing. (I will not name these culprits, but they post comments frequently on the internet.) It seems that they want to achieve nobility by being vicarious champions of justice, as they wallow in their own prosperity. However from an evolutionary point of view, the weak need to be weeded out.

    In the past, in the real ancient world outside of Robin’s imagined farmer/forager dichotomy, cultures battled for superiority on horseback and in chariots in a quest for superiority and nobility.

    When times got tough, the key to survival was to prevent each person from putting himself first but instead to maintain group survival. Here the non-self-serving characteristics of morality improved survival. In soft times it allows the weak the moral advantage. Not that this means life was or is a bowl of cherries. I like it lke it is now, so I tire of listening to the do nothing complainers.

  • I learned through trial and error that if someone much smarter than you appears to make a mistake that is covered in the relevant introductory undergraduate course, it’s foolhardy to try to correct them on it, because it’s almost certain that you have made an error interpreting them….

    ….but this just seems overwhelmingly like a dispute which becomes as-resolved-as-possible during an introduction to moral philosophy course. Yes, the overwhelming evidence appears to be that moral intuition were finely shaped by selective pressures. But the standard discussion of relative versus absolute morality makes it pretty clear that anything worth being called “Morality” (rather than just “what I like” or “what my culture likes”)–that is, anything wielding a true categorical imperative–has just got to be treated like an absolute morality, regardless of how unlikely such a thing seems given our understanding of humans and the universe.

  • Gulliver

    Robin Hanson, your reasoning that morality should exist is circular. What should or should not be is necessarily based on a value judgment. It cannot exist a priori a moral framework. What you’re proposing is a metamoral value, which only works if there exists an objective morality. For there to exist an objective morality, there must be an agent to choose it whose existential authority overrides all others. Some would call this entity God, a non-falsifiable proposition.

    However, I suspect you are trying the other garden path moral objectivists often go down. It may indeed be true that a moral system that seeks to hasten the extinction of all conscious beings is evolutionarily unfit for survival, especially if its agents fully achieve its goal (though given the unlikelihood of that outcome, the meme itself may occupy a stable niche and may even be beneficial to the species and its ecosystem by serving as a benign pressure working against overpopulation). But survival fitness is not a value judgment because evolution evidences no teleological objective; the universe as a whole has no goals, it simply is. So immolation is not objectively amoral, because nothing is objectively amoral (unless you believe in a High Power). Values are chosen by conscious agents, even if all they choose is what someone else tells them. Moral reasoning extrapolates from values. Ergo, morality is subjective.

    A moral system that seeks the destruction of its agents and all other conscious and/or emotive life forms is entirely self-consistent. It’s also awesomely megalomaniacal, but that doesn’t make it objectively wrong, evil, immoral or amoral. Don’t conflate objective existence with your own subjective values and goals.

    To be taken seriously, any antinatalist ought to answer the following three questions:

    1) Would you prefer yourself never to have been born?

    2) If so, why do you choose to remain alive?

    3) If not, why do you assert the right to decide whether or not to be born over all other conscious or emotive life forms?

    The biggest hurdle to antinatalism isn’t immorality. It’s the fact that values, suffering, and spiritual enlightenment are all subjective experiences, and most people are highly resistant to letting others deprive them of those experiences.

    @ Dremora

    Morality is about emotions

    True only insofar as morality depends on values the selection of which is inextricably influenced by emotions. Moral reasoning itself requires on the logical extrapolation from one’s selected values. Errors in that extrapolation, i.e. inconsistencies with the chosen values, reflect flawed reasoning. Taste, on the other hand, such as whether I prefer Japanese cuisine to Mediterranean, need have nothing to do with reason at all.

    • Dremora

      Good questions. I’m not an antinatalist (more of a utilitarian and a conditional natalist, i.e. people should only have children if they can meet their basic needs and the conditions are quite safe, especially for infants and very young children who cannot commit suicide yet), but I’ll answer the questions anyway.

      1) Currently yes; I would not voluntarily repeat my life as it was. If the future turns out to be very good, that may well change.

      2) Because the expected utility of committing suicide now, given the suicide methods that I have at my disposal and the current circumstances of my personal life, is lower than the expected utility of not committing suicide now. I am politically in favor of better suicide options for everyone, including minors.

      • Gulliver

        I agree that a couple should only reproduce if they can meet their offspring’s basic needs. I am also in favor of better suicide options for adults only, and an end to the punitive laws surrounding the estates of suicides. I believe suicide should be legally sanctioned for individuals who can demonstrate that they have discharged all responsibilities which they have voluntarily assumed (such as raising a child).

        That said, I would do everything in my power short of legal or physical coercion to discourage suicide. But, because I do not regard the lives of others as justly belonging to me, either personally or through the power of the state – and, conversely, do not recognize any nonconsensual claim on my life – I do not support depriving citizens of self-ownership (where ownership is defined in its most rudimentary sense as the sovereignty over one’s body). For this same reason I support neither capital punishment nor laws that violate the sovereignty of any woman’s body (such as the abominable forced invasive sonograms my state is trying to mandate).

        In my moral system, it should be up to the citizen whether or not their life is worth living.

  • I think the problem that Robin is trying to dance his way around without explicitly stating it, is why are the children of some elites actively deciding to not have children themselves.

    This is causing cognitive dissonance among the elites because they feel that because they are elites they have elite genes, and have produced elite children with elite genes and so those elite children should also produce even more elite children with elite genes so that future generations of elites can continue to exploit future generations of non-elites.

    This is the existential problem that elites are facing. Why are their children more concerned with overpopulation than with continuing the exploitation of non-elites?

    This is one of the drivers for cryonics, indefinite life extension and uploading humans into ems.

    I think that Robin is trying to frame this existential problem that elites are facing as a problem of morality, that it is moral for elites and the children of elites to have children and continue the cycle of exploitation of non-elites by elites. He is framing it as a group problem, but the group that is having the problem is not the group of all humans, it is the group that Robin identifies most strongly with, the group of wealthy elites.

    We know how to fix the problem for the group of all humans. Increase taxes on the wealthy elites and provide better jobs, education, housing, food, medical care, etc. for the non-elites. This solution is completely unacceptable because it goes against everything that the elites have been doing their entire lives. They can’t even conceive of it as a possibility.

    • Mitchell Porter

      “This is the existential problem that elites are facing. Why are their children more concerned with overpopulation than with continuing the exploitation of non-elites?
      This is one of the drivers for cryonics, indefinite life extension and uploading humans into ems.”

      The super-rich, with a few oddball exceptions, are not showing any interest in those topics. They are certainly avid consumers of medical advances, but not of futurist vaporware. The most avid supporters of transhumanism belong to a sort of high-IQ techno-proletariat of code-monkeys who like science fiction.

  • Ari

    “Morality should be adaptive; it should help groups survive.”

    I definitely agree. I actually posted elsewhere the same thing yesterday. Evolutionary psychology shaped whatever (deontological) rules we have got, very likely there’re no god-given rules.

    People who assume morality should be global and constant, are making assumption that those rules don’t need to take details of ever-shaping institutions and environments into calculation. That is like an engineer thinking “well this will work matter how fast it will go.” before relativity was discovered. Even physics hundred or whatever number of years ago was still quite different from modern theory of it (quantum physics). Its simply a delusional assumption that morality wouldn’t be more prone to errors than anything else. In fact it would be more prone thanks to biases, and status games. In fact, physics is rather simple compared to myriad of problems faced by humans and their institutions. At least as long as you work with highly accurate macro abstractions. Moral statements work on so high level of abstraction that they leak in enormous ways.

    The forces of entropy are mighty. Almost all places where there is entropy, there is lots of complex systems and change all the time. Rules need to adapt. Whether its machines, institutions, plants, animals, they need to adapt — or they die.

    Too bad a lot of people get caught up defending some tradition because of signalling reasons. Everybody loses.

  • John Maxwell IV

    Morality is what I want, for myself and others. Telling me how to be moral amounts to telling me what I want. At best, you could point out inconsistencies, in which case I could think and try to decide what I really want.

    Let’s say there was a species of ants that grew intelligent. The ants had social rules. Over time, the ants evolved to think that their social rules were just as real as trees and rocks, because ants who believed that had higher fitness. Ant philosophers argued about their rules in the same way they argued about questions of fact like whether their world was round or flat.

    I view human discourse on morality as being similar to the discourse of the ants. I’m happy to see humans argue about what laws there should be, but it seems silly to try to derive some basis for supposedly “objective” morals.

    Sure, you can try to convince me what I want, in the same way a salesman tries to convince me what I want. Or you can suggest to me that I should try to want something for some sort of game theoretic cooperative reason, in which case I will take your argument into consideration. But don’t try to pull a fast one and suggest that what I want should be determined based on some sort of abstract objective justification.

  • David

    I don’t think there’s any doubt that evolution selected for creatures with group-supporting psychological attitudes. These attitudes, which have helped us survive, feel from the inside like moral sentiments. But the question is whether precisely these moral sentiments make up the content of true morality. Note that evolution will produce the same pro-social sentiments no matter what the moral truth really is, or even if there are no moral facts at all. Now we might be lucky and it might turn out by happy accident that moral truth and the sentiments implanted in us by evolution perfectly match. But why should we think that as a matter of fact they do?

  • mjgeddes

    Lets be clear. Minds are programs that form representations of things – or signals instantiated in brains. As Hanson pointed out previously, brains are signal processors. And signal processors are explained by algorithmic information theory

    The results of algorithmic information theory are both objective and universal…. universal terminal values could be implicit in algorithmic information theory. So commentators are too quick to give up on the notion of universal morality.

    I gave a clear metric in the other thread for moral value/moral error. I suggested a complexity metric based on minimization of cognitive complexity during knowledge representation of goals This metric would be defined by algorithmic information theory, so it would be completely objective and universal in scope.

    The key point I’m suggesting here is that morals are defined on representations of actions/goals, not actions/goals themselves.

    • V

      I can program a computer to output “2+2=5”, this doesn’t make it correct.

      • mjgeddes

        The key point is that we can indeed dispense with the ‘evolutionary’ framework and replace it with ‘algorithmic information theory’, which is based on objective, universal math. We can ditch that CEV nonsense also.

        As the brain is a signal processor, ‘moral feelings’ are simply another form of ‘signal’, and like all signals there is ‘noise’ (‘moral error’). The ‘noise’ can be quantified using the complexity metric I suggested. This leads to a super-duper revolutionary new theory of information, and constitutes the ultimate ‘moral calculus’.

      • V

        Algorithmic information theory makes no claims about morality. I’m not aware of any complexity metric, in information-theoretic sense, that can be used to approximate moral intuitions.

      • No, this is just silly. The brain is not an algorithmic processor. It does not operate using algorithms.

        Algorithms manipulate data without losses and with perfect fidelity. Brains don’t do that except in very special circumstances.

        Brains can emulate an algorithmic processor, that is what you are doing when you learn to implement algorithms, such as the counting algorithm or the arithmetic algorithm. The brain can’t run algorithms “native”, the computational structure of the brain does not have the native ability to run algorithms.

        Human brains evolved to have a hierarchy of values because organisms that acted as if they had those values survived and reproduced and have descendants that are alive today. Organisms that did not act as if they instantiated these values don’t have descendants.

        To the extent that a human acts counter to this hierarchy, that action is “immoral”.

        1. Survival & reproduction of species
        If species extinct, then individual organism has no mates, no conspecifics for offspring mates, no surviving offspring. No extant species is descended from an organism that successfully favored individual survival over species survival.

        2. Survival & reproduction of organism
        All extant organisms are only descended from ancestors that survived until they reproduced. Behaviors favoring descendants included here.

        3. Survival of organ
        Example beating heart with necrotic spots following myocardial infarct. Organ will fail trying to maintain organism viability, liver, kidney failure. Death from exhaustion running from bear.

        4. Survival of tissue compartment
        Pruning of neuroanatomy via excitotoxicity following stroke to prevent future seizures. Tissue compartments will fail to preserve organ function.

        5. Survival of cell
        Mitochondrial failure due to overload, high potential, superoxide.

        6. Survival of organelle
        Mitochondrial pore transition release cytochrome c when superoxide level falls too low.

        7. Survival of DNA
        Cell viability is more important than maintaining DNA error free.

      • Kith

        Algorithms manipulate data without losses and with perfect fidelity. Brains don’t do that except in very special circumstances.

        That’s not true. An algorithm is simply a step-by-step method for the manipulation of objects (usually, but not always containing data). An algorithm can be very lossy, like JPEG compression or the algorithm for converting some of the things in my kitchen into a pie; or it can be lossless, like ZIP compression or conversion of a base-two number to a base-sixteen number.
        As for your specification of morality, I think that is a very close view of a specific kind of morality (human biological morality, to be precise). Try to think of the most general specification you can, one that would cover the most volume of morality-space. I think a specification like that would look more like a utility function that outputs “shouldness” over the space of all possible futures. A function like that must be adaptable in order to work at all.
        As to evolution, it is responsible for what we are now, and will eventually update our morality. What we should be deciding is, is it more “should-y” to allow that (slow, imprecise) process to proceed apace or to self-modify based on our current understandings?

      • The “lossyness” of algorithms is built into them. Yes, an algorithm can be designed to erase data, but if that is the function of the algorithm, it is not the kind of losses that I was thinking about. A Turing Machine doesn’t have losses. The losses in the brain are due to the substrate the brain uses (neural networks with unknown properties), not the algorithms the brain uses.

        The brain is not algorithmic. There are no algorithms that are being operated to manipulate data to achieve results.

        Making a pie is not using an algorithm to manipulate data. Pie ingredients are not data.

        Implementing algorithms is substrate independent. Any Turing Equivalent can implement any and every algorithm that any other Turing Equivalent can implement. If human brains are algorithmic, then they should all be equivalent. They should all be equally capable of implementing the human intelligence algorithm and all humans should be equivalently intelligent. All humans are not equivalently intelligent. Human intelligence is not reducible to an algorithm operating in the human brain.

    • Dave

      Math is value free.I am no math genius,but I thought it was just internally consistent ,not able to explain externalities,just in some cases model them.

      Morality on the other hand always leads to logical paradox which people then unsuccesfully try to paper over using ever more complex digressions, caveats,and tricky word play,because they think morality should be logically consistent so that it can be universal. I believe that there is morality and immorality. Math won’t be able to nail it down. It is some sort of religious like thing.

      • mjgeddes

        I didn’t say ‘math’ could solve the problems of morality, I said ‘algorithmic information theory’. Whilst math alone is indeed value-free, algorithmic information theory is math + additional empirical ingredients (properties of computation).

        The failure to solve moral problems is simply due to the rampant stupidly of information theorists to date, who have failed to grasp even the most basic implications of their own ideas.

        I’ve already hinted that when the problems of ethics are re-formulated in terms of knowledge representation of goals rather states of the world, all of them fall to the power of algorithmic information theory very easily, with ‘beauty’ turning out to be the primary universal terminal value. Algorithmic information theory is the only ‘religion’ you need 😉

      • Dave

        We, the unwashed lay persons say don’t understand your Latin,so we are in need of a pontiff. Any volunteers?

  • Evan

    Morality should be adaptive; it should help groups survive.

    I agree that morality should help people survive. But “adaptive”in evolutionary biology means more than “survival.” It means “converts as much of the environment as possible into copies of oneself.” I can’t agree with that. I think that morality should fulfill human values by helping us lead eudaemonic existences.

    Obviously survival is an important part of that. And reproduction is too to some extent, since increasing total well-being is an important value. But there are so many other values besides reproduction. It would be better to use a large portion of the universe’s resources for non-reproductive purposes so that people can live fulfilled and happy lives. Any morality that claims we should convert as much matter as physically possible into people seems just as insane to me as one that says people should go extinct.

    • Roger


      I agree and disagree, depending upon the scope and time frame.

      Life and value are woven into each other. 3.8 billion years ago there was no life and no value. Now there is a lot of both.

      I do see the expansion, flourishing and persistence of life across the universe as good. As life expands, as solutions improve, quality expands with it. They are a packaged deal. Life can expand too fast though and create more problems than it solves.

      I think the universe would be immensely better a billion years from now if life, concsiousness and value is ubiquitous.

  • V

    Claiming that ‘morality should …’ is begging the question, since in order to say ‘should’ you already need a morality.

    If your morality involves maximising evolutionary fitness, then the logical conclusion seems to be that you should devote as much effort and resources as possible to cultivation of bacteria or HeLa cells.

  • Dremora is right.

  • Adaptive morality: Nietzscheanism, beyondism.

    • Gulliver

      Adaptive morality: Nietzscheanism, beyondism.

      This is a common mistake made by, among others, the person who coined the term moral objectivism, Ayn Rand. Nietzsche was an existentialist, not an objectivist. The thesis of Beyond Good and Evil was that a slave has no moral obligation to its master because it cannot make its own moral decisions with regard to its master since its choices are made for it. In other works Nietzsche argued that a maladaptive morality cannot survive, but he never said that made it immoral or amoral.

      However, in biological evolution as in any dynamic evolutionary system, a trait may be neither adaptive nor maladaptive. Many genes are propagated by their chromosomal association with adaptive genes, or the simple fact that genes can code for multiple proteins and may therefore have side-effects that have nothing to do with the adaptive phenotypic trait to which they contribute. Parasites can survive in a metastable niche so long as they do not mutate and run amok, and may even develop symbiosis at a later stage of their or their host’s development. Leaving aside Hanson’s stab at an objective metamoral value, this is the fundamental flaw with his hypothesis that a moral system must be adaptive to survive. Remember, evolution is a balanced equilibrium statistically driven by punctuated disequilibrium; survival is incidental, even though it doesn’t always seem so to sapient organisms for whom the survival instinct holds significant importance.

    • You appear to need to check out the relevant “Nietzscheanism” page.

      It credits the “Nietzscheans” in Andromeda as the source of the name.

      • V

        Trying to derive philosophical insight from pop sci-fi (a cheap rip-off of Star Trek, which is itself not exactly known for its depth) is usually a bad idea.

        The Nietzscheans in the show are the stereotypical straw-man social darwinists, amoral and individualistic to the point of sociopathy, backstabbing each other and their allies over and over. Only sloppy writing prevents them from going extinct.

      • If Andromeda’s Nietzscheans went extinct, they wouldn’t have been implementing their own philosophy very well.

      • V

        It depends. Individual fitness doesn’t always coincide with group fitness.

      • In the show, the Nietzscheans had spread throughout known space, reproducing rapidly until they constituted 8 percent of the overall Human population.

        In practice, group and individual fitnesses are often correlated. A group of very fit individuals is itself often fit. If your group dies, then normally you and all your relatives die.

        However, as you say, the Nietzscheans in the show are a bit of a “straw man”.

      • V

        There can be ‘tragedy of the commons’ where every uses a strategy that at all times maximises its individual fitness even if it eventually leads to extinction, where a cooperative strategy could have avoided extinction.

  • Wouldn’t it make more sense to ignore anti-natalists entirely? After all, if you ignore them, they really will go away … in the very long run.

  • CStout

    Moral instinct. If evolution is defined by adaption can we justly say that through our existence in this world, we have adapted and perhaps generated a moral instinct? Sociology, biology, astronomy, etc. can not determine this separately but give reverence to the entity of our being. Neurological processes equip us with reaction as do sociological processes. Life is complex and the process of morality even more so. If we can only examine one layer at a time when can we put it all together? Adaptive morality exist amongst different perceptions but instinctive morality exist in all human beings, it has to. If it does not, than we as a race of beings are screwed. Say goodbye to type one civ….

  • Gulliver

    @ Tim Tyler

    It credits the “Nietzscheans” in Andromeda as the source of the name.

    You may want to mention that as the term Nietzscheanism will reference the actual philosopher for most people, as opposed to a fictional future subspecies whose backstory harkens to the same uncareful reading of Nietzsche’s philosophical works as Ayn Rand’s (much more offhanded) references to Nietzsche. If memory serves, the regular Nietzschean cast member can even be found reading a copy of Rand’s The Fountainhead as though it were a liturgical text. And even he comes to appreciate the long-term benefits of sustained eusocial cooperation.

    Also, your video’s argument that Dennet’s informal surveys failed to draw out anyone openly stumping for his or her genome’s maximal propagation because of self-deluded signaling misses at least two important points.

    One, memetic processors as sophisticated as humans are perfectly capable of being selfish with regards to motivations other than simple reproduction, and those selfish motives are quite capable of overriding the biological imperative to reproduce. Memes don’t need genetic relations to survive. If you ever watch one of Dennet’s full lectures you’ll see that he argues for the manifest ability of memetic imperatives to override instinctual genetic ones.

    And two, if it was a façade, education and “acculturation” wouldn’t result in actual statistical drops in the cultured populations’ birthrates. Since your argument boils down to memetic imperatives acting contrary to genetic imperatives are a form of self-delusion, you may want to note that humans didn’t dominate the ecosphere by mere reproduction, and numbers of progeny are no sinecure for memetic survival. Nor are they a sinecure for genetic survival. Overpopulation is a seriously flawed survival strategy.

    • FWIW, I tend to use the term “memetic hijacking” to refer to meme interests overriding gene interests.

      Memes don’t cause fertility decreases through self-delusional facades used for signalling purposes. Rather they more directly cause fertility decreases by diverting resources away from gene-reproduction and towards meme reproduction. The classic example is the chaste priest – who cares much more for his memetic descendants, than his genetic descendants.

  • John

    “Morality should be adaptive; it should help groups survive.”

    If we assume that EMs are not human and that the creation of EMs will lead to the eventual total dispalcement of humanity (both propositions seem fairly reasonable), what does that tell us about what our stance towards EMs should be?

    Robin, if it turns out that the EMs economy requires biological humans to go extinct because of their inefficiency, would you say that we, as a species, are justified to prevent the creation of EMs?

  • Ari

    “Intelligence is the ability to adapt to change.”
    – Stephen Hawking

  • Pingback: Should Morals be Adaptive? « azmytheconomics()