By shocking your readers by advocating contrarian stances on war, slavery, democracy, death, you have galvanized us to think about what are the variations on our culture that could be beneficial (since people generally like their cultures).
The contrarian position on death is that it's bad.
There is no single contrarian position on democracy; the possibility space is barely explored IMHO. I think we should further explore the space within democracy-meritocracy hybrids before paying countries to trade slaves.
It sounds as though you're saying that cultural homogeneity has its own costs and constraints. This is an interesting point, since it seems to be increasing at a historically unique rate across the world - at least in elite circles.
> And that should cut the rate at which natural selection of DNA can improve that population or adapt it to changing conditions.
This doesn't follow at all. The actual speed and efficiency of selection will depend on many factors like effective population size or population structure, maternal/paternal age at reproduction, background mutagenesis (what altitude do you live at?), mortality sources and their connection to inclusive fitness, modern medicine saving so many lives (especially Mendelian or de novo mutations). This should be obvious because you can't just jack up mutation rates to 100%; so there must be a point at which mutation rates or genetic variance are 'too high'. And there is no reason to believe that current mutation rates + conditions hit any kind of optimum because all of those variables have changed massively over the past millennia, and still are even within generations, far too fast to remotely reach any genome-wide equilibrium. Personally, eyeballing stuff like increasing age at reproduction, disconnection of fertility from social dynamics, a large and more interconnected global population etc and comparing to efficient selection processes, I strongly suspect that natural selection would benefit from *lower* mutation rates or total genetic variance, not higher.
Even if we have too much DNA variance now, strong genetic engineering might plausibly lead to too little. But my main point was about culture variance; DNA is a motivating analogy.
My points also apply to culture, as these are general facts about evolution/selection. (Most of these carry over quite well analogously, and some literally: modern medicine *also* permits more pathological subcultures or practices to survive harmful side-effects. Some are stronger: culture, being so much smaller in effective population size in a group selection setting, would be even further from equilibrium etc).
"This doesn't follow at all. The actual speed and efficiency of selection will depend on many factors like effective population size or population structure, maternal/paternal age at reproduction, background mutagenesis (what altitude do you live at?), mortality sources and their connection to inclusive fitness, modern medicine saving so many lives (especially Mendelian or de novo mutations)."
Obviously speed and efficiency of selection will depend on many factors. That doesn't affect Robin's general claim that the power of selection is limited by the genetic variation available.
Increasing variation could be detrimental is if there is so much variation in the gene pool that all the best variants are already present, or if there are many combinations of variants which are valuable only when they all occur together, so that having many variants splits up these combinations more rapidly. I've seen the second happen in evolving CA which must decide on which way to break a symmetry. I've never seen the first happen.
"eyeballing stuff like increasing age at reproduction, disconnection of fertility from social dynamics, a large and more interconnected global population etc and comparing to efficient selection processes, I strongly suspect that natural selection would benefit from *lower* mutation rates or total genetic variance, not higher."
Optimal mutation rates depend on many factors, and these factors may push in either direction, so I would hesitate to make any claim about higher or lower rates being better in general. GA devs tend to fiddle a lot with the mutation rate.
Genetic variance, OTOH, is usually lower than you'd like it to be, because of a kind of feedback: if fitness is increasing rapidly, that tends to make variance decrease rapidly. GAs which are functioning well naturally starve themselves of genetic variety. So it would be unusual for a system that has been evolving for a long time to have too much genetic variance. This is why premature convergence is the bugaboo of GA, and GA devs put a lot of effort into increasing variance, but not into decreasing it.
Pay people to have slaves?? I think I’d rather have a maladaptive culture. What are the odds that we are too morally overconfident that slavery is bad?
I think that, if slavery is ever part of a "good" culture, it would be in a pre-industrial near-subsistence society, not in a high-tech one. Or in a society which admitted other animals as members, or transhumans, or AIs, so that there was a stratum of society that needed supervision.
I'm confident that subsidizing slavery doesn't have enough marginal utility in any near-future world to use it as an example here. It's just provocation. Robin, you're increasingly going for shock value in these posts, and IMHO, given the lack of proponents for your ideas, it just taints all of them. You're pushing them further outside the overton window. Being technically correct is not always helpful. If your ideas ever catch on, somebody is going to read all your posts and make a short list of the most-offensive ones, and you and those ideas will be finished for a generation. We're living in a one-strike-you're-out world, and this post has two in it.
The point of this post isn't some incidental issue I'm highlighting to be provocative. This is a central point! Yes, it may be hard to swallow, but ignoring it won't help.
The modern norm against conquest and similar agression came about along with nuclear weapons. This norm was a strengthening of one that arose after the first world war, after experience with chemical weapons and the first long range bombers making real the prospect of gas attacks killing millions of civilians in major cities. Today there is also potential for biological weaponds to add to that. As technology advances, destructive possibilities increase.
If a big war with an extreme culture can result in extinction, it becomes far more risky to encourage "deviant" cultures. It may not be adaptive in the end unless limited.
From an adaptivity point of view, I would guess that you would think that t's only ones related to existential issues for all of humanity. I might be slightyl more expansive on those than you, but not much. I don't know the full scope of them. Probably they are all technology related, but Itechnology limits are tricky. The same technology that can take us to Mars and help humanity survive can put a payload anywhere on earth within two hours.
Also if you accept any of them as special enough to warant some universal bounds, it can be difficult to draw the lines, and it opens the door quite a lot for cultures to interfere with each other. Look at Israel and Iran over the last few years. One or both cultures can accusse the other and away we go. Difficult to restrain and sort out, and you end up with embargoes and all sorts of things going on. With nuclear weapons or similar available everyone ends up "interested" in it.
Conflict is just one example, but if your focus is adaptivness, I don't see how humanity becoming extinct inadvertantly is any worse than from a conflict.
Maybe you are thinking that a conflict has a winner, and so it can't lead to extinction, but it is not difficult to imagine a cult (perhaps on a national scale) that suicides (maybe when it can't win, or maybe just anyway) and takes down the whole of humanity with it. A few people survived Jonestown, but with more effective technology (perhaps biological), who knows.
If you want to have a complete hands-off world where many different cultures are given free reign, you'll need to convince a large number of people that the doomsday type scenarios are very unlikley. Talk of people like Putin being heros for conquring lteritory isn't likely to get you there. It also runs counter to cultres leaving each other alone to develop in their own ways. Surely a norm against conquest would be a fundamental part of separate cultures going their own ways.
If a technology arises that at least temporarily defeats trump weapons like nukes, a pacifist global monoculture may be extensively overthrown by conquest by those cultures which did not follow that
If you subsidized a fighting culture within your own it might be protective against a no trump weapons development, if you tried to have a global no-conquest norm it might be conquered by a single defector with no shared values
It might be protective often, but either way you might bring about your own destruction.
A world with a few good rules that maximises diversity is going to beat one with no rules and also one with too many poor rules. I think no-conquest is one of the few good ones.
"And more for deeper more consequential cultural differences, such as having contrarian stances on war, slavery, democracy, death, and gender equality."
In your examples, you are literally saying we should pay nations to start wars and take slaves. Let's not go that far.
Then take me as speaking pragmatically, politically, not theoretically. Now is not the time to argue over edge cases. Now is the time to show that your idea has value, by restricting ourselves to cases which might convince people the idea has value.
When people argue whether human cloning should be legal, someone always says, "What if someone clones himself to make organ donors?" This is a silly objection, because we already have a consensus not to do that with clones. Likewise, it would be silly to shut down your case for cultural variation because some people might want slaves. But it would NOT be silly for someone to shut down your case for cultural variation because some people want slaves when you yourself have proposed it. What you did was like proposing that laws against cloning humans should be lifted SO THAT we may make fully-conscious human clones to use as organ donors, when you really expect most of the benefits to come from cloning exceptional people.
We would have to make some compromise to get the idea used at all, and "no slavery" is an easy compromise, and one we should expect *must* be made in any case. Why kill your idea in its infancy to hold onto one theoretical case which we already know will not be accepted?
The monoculture is probably like a chromatography spectrum where a dominant peak can mask weaker signals. As the dominant civ declines, we will see more clearly the underlying differences (or those subcultures will feel less need to bind themselves to the dominant culture). The process of decline is inherently a process of diversification. The benefit is that the subcultures that separate and emerge in the wake of a declining civ were also along for the ride with the dominant civ, so will likely carry on some aspects of the original civ. Then we get a selection process where would-be new civs share some aspect of the old civ but vary in which aspects are conserved plus a variety of innovations - so the result is likely to be better than the original. (That’s basically the Toynbee model.) This makes subsidizing random pirate cultures unnecessary.
I agree that human extinction isn't likely here, but future scenarios vary in how much of our current civ they inherit. I want to save more, not less, of the values we now share.
If a deviant culture is adaptive, shouldn't it flourish without help from the maladaptive mainstream culture?
As you say yourself, our current norms won't last whatever we do. The process you suggest might happen best if done naturally - people putting forward competing values that they argue for on their own merits or attempt to push through resistance.
"Added 26Nov: Whatever it is that you want to promote in the future, including all of the good things you think follow from our monoculture norms and values, you must tie those good things to an adaptive package. Else they will go away. And yes many big compromises may be required to tie them so."
Also, you must consider them AS a package. You don't get to call "low-cost house prices" good if it must come together with "high-cost rent", as in the US, which has a history of house subsidies that penalize poor people for not buying houses. You don't get to call "extremely safe drugs" good if it also requires "extremely expensive drugs" and "very few new drugs".
Evolution is usually modeled using as parameters mutation, migration (connectivity and intermarriage between subgroups, or a spatial probability distribution that governs "matings"), and selection. Variance is a function of all three: it increases with mutation, and decreases as selection or migration increase. What we have now in the evolution of human societies is a sudden decrease in selection and a sudden increase in migration.
In my experience, and in the consensus opinion, the biggest problem in genetic algorithms is premature convergence (loss of nearly all variation and a halt to further evolution). Increases in migration overpower both decreases in selection and increase of mutations in simulations, causing premature convergence. You can't solve this by increasing mutation or reducing selection. Both slow down the fitness increase exponentially. What's worse, increasing mutation requires increasing selection to avoid devolution. Decreasing selection is a terrible way of fighting premature convergence, because the variation it adds is random, unselected noise.
There are some better tricks, though unless your fitness function increases in complexity over time, they can only delay the end of evolution and/or decrease the odds of getting stuck in a local minimum. We can class them in 3 approaches: ones which change the evolution algorithm, ones which adjust the parameters, and the kind of direct intervention to increase variation that Robin is proposing (called "fitness sharing" in GA).
Changing the evolution algorithm imposes highly artificial modifications on evolution (eg simulated annealing), which is not easy when you're dealing with real-world evolution rather than a simulation. Adjusting the parameters means reducing migration (strongly disincentivising inter-cultural exchange and marriage: nationalism), increasing mutation (random cultural drift), and increasing selection (competition). Schopenhauer grasped the "competition" aspect dimly; combining Schopenhauer and Darwin led Nietzsche to attend to both migration and selection in /The Genealogy of Morals/, which led directly to Nazism. (I am having none of the standard "the Nazis misunderstood Nietzsche" bullshit; I consider it a dishonest position.)
So Robin is probably correct: our problem probably is, or will be, loss of variation. Our only alternatives seem to be bizarre centralized micromanagement of life to change the algorithm of life, Nazism, intervening directly with variation with subsidies (or punishments) as Robin proposes, or the end of cultural evolution.
(Note that Robin is talking about competition and selection between cultures. The parameters of evolution have changed on both the individual level (eg, members of every society on Earth find it easier to travel far now) and by group (eg some countries have mostly immigration while others have mostly emigration). I may have overlooked some subtlety in applying my experience fighting premature convergence in individual selection to fighting it in group selection.)
The Amish and Satmar maintain cultural boundaries without Nazism. "Reduced migration" can mean community formation under regimes that tolerate it, rather than top-down nationalist enforcement.
I agree if we were collectively trying to survive we'd be subsidizing or at least not actively suppressing skin-in-the-game cultural experimentation. There are some legit questions about how to reconcile that with the need to maintain and police some beneficial norms, but I see little point in discussing the details absent a context where there's some desire to *implement* such solutions if they become available.
Overall it seems like our current values as expressed are better modeled as emerging as an evolved memeplex, selected in part for their immune efficacy (propensity to suppress alternatives), than as a solution to a problem people had.
If we want to do better, we'd be better off trying to identify and appeal to subsets of the population whose culture is nearer the "solutions to a problem" attractor, who might therefore benefit from being better informed, and might be persuaded to reciprocate by doing things that you like, than to try to reason about survival value with something that isn't currently trying to think about how to survive.
This is a very bad idea all around. First we don't know enough about human DNA to do this safely and 2 e have too big an ego. As a Buddhist person I have become perceptive to the effects human ego has on society ans as a psychologist the language to express this. We are living in what one Sci-Fi opiinionator calls "egotastic fun time" Our political leadership is based entirely on the ego of one demented con man. But here are the fallacies I see in this. First, we only know what half the DNA in our genome does. We also know of the origins of less than half of that. is is H. Sapiens? H. Neandertlis? How about H. Erectus or Hidelbergensis? What about the inactive alleles? I don't think the European descended folk will want to be black again. What about gender, gender preference, and identity? Many untrained people think it's all in the X and Y chromosomes. It's not. Researchers have found over 46 alleles including those controlling the adrenal glands. Altering any one fo these to produce a pure heterosexual male or female may cause unforeseen complications in other parts of the organism. Also if science ever discovered a "gay gene" religious objections to abortion will quickly give way to religious bigotry. This is why I do not ascribe to the theory that aliens came here in our prehistoric ear, grabbed a bunch of apes, shaved them and augmented them to make a slave race. If that was the case, they did a piss poor job of it. Humanity is a colony animal. It is composed of a multitude of cells that were once once celled bacilli called Arcana. When they combined into a multi-celled being the cells specialized. To say we are a walking universe is a bit poetic but not so far from the truth. Each of those cells has DNA, not some DNA but the entire genome in it. CRISPER is a wonderful invention but it cannot change every cell in the body all at once. We are nowhere near that level of technology. Right now we could create a few Khan Noonian Singhs. But that's it.
But shouldn’t we say that some norms already proofed to be maladaptive (dictatorship etc), while others proadaptive (open inquiry) because the latter allow for more variance and adaptation, within a culture, others do not? I’m not getting making variability a value in itself, unless it drives towards something. Then, within, say, democracy, the question is how to make it less homogenous perhaps? And why exactly the culture as such should be a principal and crude evolutionary unit?
It seems to me that, in projecting the collapse of our civilization, you are projecting the trend over the last few centuries towards cultural uniformity, supposing it to continue indefinitely. But other factors are looming that may overwhelm the one on which you concentrate; for example, our growing mastery of biotechnology, and our growing ability to produce rational machines. The Singularity is near—perhaps several singularities in rapid succession—and the future will be unimaginably different from the past.
(Also, I think you underestimate the ability of mankind to respond rationally to the most obvious threats. It is far from perfect, but it is not totally nil.)
Your proposal seems to depend on a view of Sam Harris’s “Moral Landscape” that is not static but dynamically reshaped by exogenous forces such as technological advance and cosmological events. If the moral landscape shifts by itself over time, then subsidizing high-variance cultural experiments might indeed hedge against locking humanity into a local but ephemeral peak. Your argument is effectively an exploration strategy for a non-stationary Harris landscape. We can see futures—AGI, your own emulated people—where Parfit’s Repugnant Conclusion and the possibility of large-scale slave-based or recreationally sadistic cultures can no longer be dismissed as mere dystopian fantasy.
A stress test of your idea is to consider the most extreme pariah-like objective conceivable: universal promortalism—the view that conscious experience is intrinsically bad and the morally correct endpoint is to prevent any conscious beings from existing, lest they suffer, so obliterating the universe to eliminate (and preclude the re-emergence of) consciousness is the moral imperative. What happens when such a group applies to the “pariah fund” for AI compute to explore manually collapsing the quantum state of vacuum to end the universe? Or maybe they just want a biolab to sterilize the Earth using self-destructing mirror life. Sharks, are you ready to hear the next pitch?
Back that insanity off to your own point of indecision about how much horror you're willing to subsidize in exchange for possibly averting human extinction. I venture that Harris’s landscape is more fixed than your argument would seem to require, but I’d be eager to hear the two of you discuss it.
Many here focus on the most extreme possible bad cultures they can imagine. But I didn't make an absolute claim, which wouldn't allow for any exceptions. I'm focused in the middle of an important distribution, not on every imaginable tail.
By shocking your readers by advocating contrarian stances on war, slavery, democracy, death, you have galvanized us to think about what are the variations on our culture that could be beneficial (since people generally like their cultures).
The contrarian position on death is that it's bad.
There is no single contrarian position on democracy; the possibility space is barely explored IMHO. I think we should further explore the space within democracy-meritocracy hybrids before paying countries to trade slaves.
It sounds as though you're saying that cultural homogeneity has its own costs and constraints. This is an interesting point, since it seems to be increasing at a historically unique rate across the world - at least in elite circles.
It's a well-known law of evolution.
> And that should cut the rate at which natural selection of DNA can improve that population or adapt it to changing conditions.
This doesn't follow at all. The actual speed and efficiency of selection will depend on many factors like effective population size or population structure, maternal/paternal age at reproduction, background mutagenesis (what altitude do you live at?), mortality sources and their connection to inclusive fitness, modern medicine saving so many lives (especially Mendelian or de novo mutations). This should be obvious because you can't just jack up mutation rates to 100%; so there must be a point at which mutation rates or genetic variance are 'too high'. And there is no reason to believe that current mutation rates + conditions hit any kind of optimum because all of those variables have changed massively over the past millennia, and still are even within generations, far too fast to remotely reach any genome-wide equilibrium. Personally, eyeballing stuff like increasing age at reproduction, disconnection of fertility from social dynamics, a large and more interconnected global population etc and comparing to efficient selection processes, I strongly suspect that natural selection would benefit from *lower* mutation rates or total genetic variance, not higher.
Even if we have too much DNA variance now, strong genetic engineering might plausibly lead to too little. But my main point was about culture variance; DNA is a motivating analogy.
My points also apply to culture, as these are general facts about evolution/selection. (Most of these carry over quite well analogously, and some literally: modern medicine *also* permits more pathological subcultures or practices to survive harmful side-effects. Some are stronger: culture, being so much smaller in effective population size in a group selection setting, would be even further from equilibrium etc).
"This doesn't follow at all. The actual speed and efficiency of selection will depend on many factors like effective population size or population structure, maternal/paternal age at reproduction, background mutagenesis (what altitude do you live at?), mortality sources and their connection to inclusive fitness, modern medicine saving so many lives (especially Mendelian or de novo mutations)."
Obviously speed and efficiency of selection will depend on many factors. That doesn't affect Robin's general claim that the power of selection is limited by the genetic variation available.
Increasing variation could be detrimental is if there is so much variation in the gene pool that all the best variants are already present, or if there are many combinations of variants which are valuable only when they all occur together, so that having many variants splits up these combinations more rapidly. I've seen the second happen in evolving CA which must decide on which way to break a symmetry. I've never seen the first happen.
"eyeballing stuff like increasing age at reproduction, disconnection of fertility from social dynamics, a large and more interconnected global population etc and comparing to efficient selection processes, I strongly suspect that natural selection would benefit from *lower* mutation rates or total genetic variance, not higher."
Optimal mutation rates depend on many factors, and these factors may push in either direction, so I would hesitate to make any claim about higher or lower rates being better in general. GA devs tend to fiddle a lot with the mutation rate.
Genetic variance, OTOH, is usually lower than you'd like it to be, because of a kind of feedback: if fitness is increasing rapidly, that tends to make variance decrease rapidly. GAs which are functioning well naturally starve themselves of genetic variety. So it would be unusual for a system that has been evolving for a long time to have too much genetic variance. This is why premature convergence is the bugaboo of GA, and GA devs put a lot of effort into increasing variance, but not into decreasing it.
> Optimal mutation rates depend on many factors, and these factors may push in either direction...
Indeed.
> So it would be unusual for a system that has been evolving for a long time to have too much genetic variance.
And humanity is a highly unusual system which has not been evolving for very long.
So then, we agree.
argh, I'm moving my long comment elsewhere, as I missed that you were responding only to Robin's first 2 paragraphs.
Pay people to have slaves?? I think I’d rather have a maladaptive culture. What are the odds that we are too morally overconfident that slavery is bad?
And for how many of the hundreds of dimensions of global monoculture are you that confident it is best?
I think that, if slavery is ever part of a "good" culture, it would be in a pre-industrial near-subsistence society, not in a high-tech one. Or in a society which admitted other animals as members, or transhumans, or AIs, so that there was a stratum of society that needed supervision.
I'm confident that subsidizing slavery doesn't have enough marginal utility in any near-future world to use it as an example here. It's just provocation. Robin, you're increasingly going for shock value in these posts, and IMHO, given the lack of proponents for your ideas, it just taints all of them. You're pushing them further outside the overton window. Being technically correct is not always helpful. If your ideas ever catch on, somebody is going to read all your posts and make a short list of the most-offensive ones, and you and those ideas will be finished for a generation. We're living in a one-strike-you're-out world, and this post has two in it.
The point of this post isn't some incidental issue I'm highlighting to be provocative. This is a central point! Yes, it may be hard to swallow, but ignoring it won't help.
The point is important. But you could omit the most-provocative example.
Deviant stances on war obviously impact other cultures!
Of course
The modern norm against conquest and similar agression came about along with nuclear weapons. This norm was a strengthening of one that arose after the first world war, after experience with chemical weapons and the first long range bombers making real the prospect of gas attacks killing millions of civilians in major cities. Today there is also potential for biological weaponds to add to that. As technology advances, destructive possibilities increase.
If a big war with an extreme culture can result in extinction, it becomes far more risky to encourage "deviant" cultures. It may not be adaptive in the end unless limited.
The modern norm against aggressive war is only one of many norms of our monoculture. How many of them do you think are "special" in this way?
From an adaptivity point of view, I would guess that you would think that t's only ones related to existential issues for all of humanity. I might be slightyl more expansive on those than you, but not much. I don't know the full scope of them. Probably they are all technology related, but Itechnology limits are tricky. The same technology that can take us to Mars and help humanity survive can put a payload anywhere on earth within two hours.
Also if you accept any of them as special enough to warant some universal bounds, it can be difficult to draw the lines, and it opens the door quite a lot for cultures to interfere with each other. Look at Israel and Iran over the last few years. One or both cultures can accusse the other and away we go. Difficult to restrain and sort out, and you end up with embargoes and all sorts of things going on. With nuclear weapons or similar available everyone ends up "interested" in it.
The possibility of conflict is far from the worse possibility here.
Conflict is just one example, but if your focus is adaptivness, I don't see how humanity becoming extinct inadvertantly is any worse than from a conflict.
Maybe you are thinking that a conflict has a winner, and so it can't lead to extinction, but it is not difficult to imagine a cult (perhaps on a national scale) that suicides (maybe when it can't win, or maybe just anyway) and takes down the whole of humanity with it. A few people survived Jonestown, but with more effective technology (perhaps biological), who knows.
If you want to have a complete hands-off world where many different cultures are given free reign, you'll need to convince a large number of people that the doomsday type scenarios are very unlikley. Talk of people like Putin being heros for conquring lteritory isn't likely to get you there. It also runs counter to cultres leaving each other alone to develop in their own ways. Surely a norm against conquest would be a fundamental part of separate cultures going their own ways.
If a technology arises that at least temporarily defeats trump weapons like nukes, a pacifist global monoculture may be extensively overthrown by conquest by those cultures which did not follow that
And then you've got a different monoculture.
A global norm against conquest is surely a necessity to maximise diversity. Robin needs to admit some global rules to achive his aims.
If you subsidized a fighting culture within your own it might be protective against a no trump weapons development, if you tried to have a global no-conquest norm it might be conquered by a single defector with no shared values
It might be protective often, but either way you might bring about your own destruction.
A world with a few good rules that maximises diversity is going to beat one with no rules and also one with too many poor rules. I think no-conquest is one of the few good ones.
"And more for deeper more consequential cultural differences, such as having contrarian stances on war, slavery, democracy, death, and gender equality."
In your examples, you are literally saying we should pay nations to start wars and take slaves. Let's not go that far.
How many more alternate culture features will you cross off your list as unacceptable?
Then take me as speaking pragmatically, politically, not theoretically. Now is not the time to argue over edge cases. Now is the time to show that your idea has value, by restricting ourselves to cases which might convince people the idea has value.
When people argue whether human cloning should be legal, someone always says, "What if someone clones himself to make organ donors?" This is a silly objection, because we already have a consensus not to do that with clones. Likewise, it would be silly to shut down your case for cultural variation because some people might want slaves. But it would NOT be silly for someone to shut down your case for cultural variation because some people want slaves when you yourself have proposed it. What you did was like proposing that laws against cloning humans should be lifted SO THAT we may make fully-conscious human clones to use as organ donors, when you really expect most of the benefits to come from cloning exceptional people.
We would have to make some compromise to get the idea used at all, and "no slavery" is an easy compromise, and one we should expect *must* be made in any case. Why kill your idea in its infancy to hold onto one theoretical case which we already know will not be accepted?
The monoculture is probably like a chromatography spectrum where a dominant peak can mask weaker signals. As the dominant civ declines, we will see more clearly the underlying differences (or those subcultures will feel less need to bind themselves to the dominant culture). The process of decline is inherently a process of diversification. The benefit is that the subcultures that separate and emerge in the wake of a declining civ were also along for the ride with the dominant civ, so will likely carry on some aspects of the original civ. Then we get a selection process where would-be new civs share some aspect of the old civ but vary in which aspects are conserved plus a variety of innovations - so the result is likely to be better than the original. (That’s basically the Toynbee model.) This makes subsidizing random pirate cultures unnecessary.
I agree that human extinction isn't likely here, but future scenarios vary in how much of our current civ they inherit. I want to save more, not less, of the values we now share.
In your scenario, would you preferentially subsidize the new cultures that share your values, to stack the deck in that direction?
If I were king maybe, but I worry about our monoculture trying to coordinate to manage a regime like that, we might undermine it
If a deviant culture is adaptive, shouldn't it flourish without help from the maladaptive mainstream culture?
As you say yourself, our current norms won't last whatever we do. The process you suggest might happen best if done naturally - people putting forward competing values that they argue for on their own merits or attempt to push through resistance.
A single adaptive culture is much worse at adapting to future changes than a variety of cultures, even if each member is less adaptive.
"Added 26Nov: Whatever it is that you want to promote in the future, including all of the good things you think follow from our monoculture norms and values, you must tie those good things to an adaptive package. Else they will go away. And yes many big compromises may be required to tie them so."
Also, you must consider them AS a package. You don't get to call "low-cost house prices" good if it must come together with "high-cost rent", as in the US, which has a history of house subsidies that penalize poor people for not buying houses. You don't get to call "extremely safe drugs" good if it also requires "extremely expensive drugs" and "very few new drugs".
Evolution is usually modeled using as parameters mutation, migration (connectivity and intermarriage between subgroups, or a spatial probability distribution that governs "matings"), and selection. Variance is a function of all three: it increases with mutation, and decreases as selection or migration increase. What we have now in the evolution of human societies is a sudden decrease in selection and a sudden increase in migration.
In my experience, and in the consensus opinion, the biggest problem in genetic algorithms is premature convergence (loss of nearly all variation and a halt to further evolution). Increases in migration overpower both decreases in selection and increase of mutations in simulations, causing premature convergence. You can't solve this by increasing mutation or reducing selection. Both slow down the fitness increase exponentially. What's worse, increasing mutation requires increasing selection to avoid devolution. Decreasing selection is a terrible way of fighting premature convergence, because the variation it adds is random, unselected noise.
There are some better tricks, though unless your fitness function increases in complexity over time, they can only delay the end of evolution and/or decrease the odds of getting stuck in a local minimum. We can class them in 3 approaches: ones which change the evolution algorithm, ones which adjust the parameters, and the kind of direct intervention to increase variation that Robin is proposing (called "fitness sharing" in GA).
Changing the evolution algorithm imposes highly artificial modifications on evolution (eg simulated annealing), which is not easy when you're dealing with real-world evolution rather than a simulation. Adjusting the parameters means reducing migration (strongly disincentivising inter-cultural exchange and marriage: nationalism), increasing mutation (random cultural drift), and increasing selection (competition). Schopenhauer grasped the "competition" aspect dimly; combining Schopenhauer and Darwin led Nietzsche to attend to both migration and selection in /The Genealogy of Morals/, which led directly to Nazism. (I am having none of the standard "the Nazis misunderstood Nietzsche" bullshit; I consider it a dishonest position.)
So Robin is probably correct: our problem probably is, or will be, loss of variation. Our only alternatives seem to be bizarre centralized micromanagement of life to change the algorithm of life, Nazism, intervening directly with variation with subsidies (or punishments) as Robin proposes, or the end of cultural evolution.
(Note that Robin is talking about competition and selection between cultures. The parameters of evolution have changed on both the individual level (eg, members of every society on Earth find it easier to travel far now) and by group (eg some countries have mostly immigration while others have mostly emigration). I may have overlooked some subtlety in applying my experience fighting premature convergence in individual selection to fighting it in group selection.)
The Amish and Satmar maintain cultural boundaries without Nazism. "Reduced migration" can mean community formation under regimes that tolerate it, rather than top-down nationalist enforcement.
I agree if we were collectively trying to survive we'd be subsidizing or at least not actively suppressing skin-in-the-game cultural experimentation. There are some legit questions about how to reconcile that with the need to maintain and police some beneficial norms, but I see little point in discussing the details absent a context where there's some desire to *implement* such solutions if they become available.
Overall it seems like our current values as expressed are better modeled as emerging as an evolved memeplex, selected in part for their immune efficacy (propensity to suppress alternatives), than as a solution to a problem people had.
If we want to do better, we'd be better off trying to identify and appeal to subsets of the population whose culture is nearer the "solutions to a problem" attractor, who might therefore benefit from being better informed, and might be persuaded to reciprocate by doing things that you like, than to try to reason about survival value with something that isn't currently trying to think about how to survive.
This is a very bad idea all around. First we don't know enough about human DNA to do this safely and 2 e have too big an ego. As a Buddhist person I have become perceptive to the effects human ego has on society ans as a psychologist the language to express this. We are living in what one Sci-Fi opiinionator calls "egotastic fun time" Our political leadership is based entirely on the ego of one demented con man. But here are the fallacies I see in this. First, we only know what half the DNA in our genome does. We also know of the origins of less than half of that. is is H. Sapiens? H. Neandertlis? How about H. Erectus or Hidelbergensis? What about the inactive alleles? I don't think the European descended folk will want to be black again. What about gender, gender preference, and identity? Many untrained people think it's all in the X and Y chromosomes. It's not. Researchers have found over 46 alleles including those controlling the adrenal glands. Altering any one fo these to produce a pure heterosexual male or female may cause unforeseen complications in other parts of the organism. Also if science ever discovered a "gay gene" religious objections to abortion will quickly give way to religious bigotry. This is why I do not ascribe to the theory that aliens came here in our prehistoric ear, grabbed a bunch of apes, shaved them and augmented them to make a slave race. If that was the case, they did a piss poor job of it. Humanity is a colony animal. It is composed of a multitude of cells that were once once celled bacilli called Arcana. When they combined into a multi-celled being the cells specialized. To say we are a walking universe is a bit poetic but not so far from the truth. Each of those cells has DNA, not some DNA but the entire genome in it. CRISPER is a wonderful invention but it cannot change every cell in the body all at once. We are nowhere near that level of technology. Right now we could create a few Khan Noonian Singhs. But that's it.
But shouldn’t we say that some norms already proofed to be maladaptive (dictatorship etc), while others proadaptive (open inquiry) because the latter allow for more variance and adaptation, within a culture, others do not? I’m not getting making variability a value in itself, unless it drives towards something. Then, within, say, democracy, the question is how to make it less homogenous perhaps? And why exactly the culture as such should be a principal and crude evolutionary unit?
It seems to me that, in projecting the collapse of our civilization, you are projecting the trend over the last few centuries towards cultural uniformity, supposing it to continue indefinitely. But other factors are looming that may overwhelm the one on which you concentrate; for example, our growing mastery of biotechnology, and our growing ability to produce rational machines. The Singularity is near—perhaps several singularities in rapid succession—and the future will be unimaginably different from the past.
(Also, I think you underestimate the ability of mankind to respond rationally to the most obvious threats. It is far from perfect, but it is not totally nil.)
We should admit which way trends are moving without having absolute confidence they will continue.
'While we can often reason well about “facts”, we just don’t on “values”.'
I wish people could at least reason well about facts, but they don't.
Your proposal seems to depend on a view of Sam Harris’s “Moral Landscape” that is not static but dynamically reshaped by exogenous forces such as technological advance and cosmological events. If the moral landscape shifts by itself over time, then subsidizing high-variance cultural experiments might indeed hedge against locking humanity into a local but ephemeral peak. Your argument is effectively an exploration strategy for a non-stationary Harris landscape. We can see futures—AGI, your own emulated people—where Parfit’s Repugnant Conclusion and the possibility of large-scale slave-based or recreationally sadistic cultures can no longer be dismissed as mere dystopian fantasy.
A stress test of your idea is to consider the most extreme pariah-like objective conceivable: universal promortalism—the view that conscious experience is intrinsically bad and the morally correct endpoint is to prevent any conscious beings from existing, lest they suffer, so obliterating the universe to eliminate (and preclude the re-emergence of) consciousness is the moral imperative. What happens when such a group applies to the “pariah fund” for AI compute to explore manually collapsing the quantum state of vacuum to end the universe? Or maybe they just want a biolab to sterilize the Earth using self-destructing mirror life. Sharks, are you ready to hear the next pitch?
Back that insanity off to your own point of indecision about how much horror you're willing to subsidize in exchange for possibly averting human extinction. I venture that Harris’s landscape is more fixed than your argument would seem to require, but I’d be eager to hear the two of you discuss it.
Many here focus on the most extreme possible bad cultures they can imagine. But I didn't make an absolute claim, which wouldn't allow for any exceptions. I'm focused in the middle of an important distribution, not on every imaginable tail.
Thank you for the clarification! I agree this involves a distribution with subjective evaluation of trade-offs.