Sunday, April 19, 2015

Drawing The Line: making the case for idea-ism

This post has been too long in coming.  I've been busy coding.  (Actually, I've been busy writing documentation, which turns out to be even more time consuming.)

A while back I promised commenter Luke that I would answer two questions:

1.  How can reason not be circular?

2.  Why is idea-ism (still searching for a better name) a better basis for morality than Christianity (or anything else for that matter)?

I answered the first question in a previous post.  The bottom line is that reason "grounds out" in the fact that the universe has discernible regularities in it, and our brains have hard-wired into them the ability to attach labels to those regularities.  We call that ability "language," and "reason" is an extension of language.  This post tackles the second question, at which point my outstanding rhetorical obligations will be fulfilled and I can get back to working on my startup :-)

First, a reminder of what idea-ism is, what it claims, and what it does not.  Idea-ism is the axiom that moral behavior is that which advances the interests of memes or ideas.  It is important to note that idea-ism is an axiom.  I can't prove it.  In fact, no system of morality can be proven because morality is not a question of fact.  The question of whether it is right to cut people's heads off for apostasy cannot be resolved experimentally.  Morality is a matter of social choice.  It is a question of what kind of world we want to try to strive for.  Which immediately raises the question: who is "we"?

This, I claim, is the more fundamental question.  Disagreements about what is moral really boil down to disagreements about where to draw the boundary between "us" and "them."  The meaning of "love thy neighbor" really depends on the meaning of the word "neighbor".  Back in Old Testament times your neighbor meant a member of your tribe.  The bloody genocides in Joshua that New Atheists gleefully point out as contradicting the sixth Commandment aren't contradictions at all, they are simply a reflection of how the line between "us" and "them" has shifted over time.  The idea that "all men are created equal" is a modern invention, a product of the Enlightenment, and it is nowhere near being universally accepted.  Nearly fifteen years after 9/11, Americans still go apoplectic over the death of nearly 3000 innocent civilians but (mostly) casually brush off the deaths of over 100,000 equally innocent Iraqi civilians.  And yet somehow we still manage to think of ourselves as the Good Guys.  That is because the 3000 who were killed were Us, and the 100,000 (or however many it was) were Them.

There is nothing wrong with drawing a line between Us and Them.  In fact, it is a necessary, even fundamental part of life.  Life is not possible without separating a replicator from its environment somehow.  Some barriers are natural (cell membranes, physical organisms), others are artificial (city walls, national boundaries), and still others are non-physical social constructs (family, ethnic group) but the fact that these boundaries exist is inescapable.  The trick is drawing them in the right place.

There are some possible places to draw the line between us and them that are clearly wrong.  For example, the idea that "all life is sacred" is obviously problematic.  The malaria parasite, for example, is alive, but I hope I don't have to convince you that it would be a mistake to argue that curing malaria would be immoral as a result.  Likewise, there are people who consider parts of their own bodies to be "the other" and believe that they would be happier as amputees.

This is clearly a pathology but it is still worth reflecting for a moment on why it is a pathology: it's not just because hacking off your own limbs makes most people queasy.  If morality were determinable by majority rule then torture and genocide would have been perfectly moral through most of human history.  It would be nice to have moral principles which were timeless and not subject to the whims of fashion.

One possible story to tell about the reason that Apotemnophilia is a pathology is that it is destructive to life.  A desire to chop your own limbs off puts the genes that produce this behavior at a significant reproductive disadvantage relative to its competitors.  That seems plausible enough, but it has a significant problem: this argument would seem to apply equally to homosexuality.  Surely genes that produce brains driven to mate with members of the same sex are at a significant reproductive disadvantage relative to genes that produce brains drive to mate with the opposite sex.  And yet homosexuality (in my view) is neither pathological nor immoral.

(Aside: it is worth thinking about how the genes for homosexuality do manage to survive as such a large proportion of the population.)

This, then, is the problem: is it possible to give a principled account of morality?  By which I mean, is it possible to construct a theory of morality that is not vacuous, that is not equivalent to, "Moral behavior is whatever people who think like me consider moral", and that can embrace variations in human sexuality and religious beliefs while rejecting pathologies like Apotemnophilia and ISIS?  My claim is that idea-ism is such a theory.  Moreover, I claim that idea-ism is the only such theory.  It may or may not be the only possible such theory, but it's the only one that I know of, and I've been looking for a long, long time.

Before I describe why I think idea-ism is such a theory, let me quickly review the moral landscape and explain what I find lacking in the competition.

Religion

Religious theories of morality are easily rejected as non-principled because they require an arbitrary adoption of some holy text or creed.  A principled theory of morality must be able to resolve moral dilemmas without resorting to asking what most adherents of the theory think is moral.  Religions can't do that, with the canonical contemporary example being the question of marriage equality: is it moral for gay people to marry?  Some Christians say yes, others say no.  Both sides can cite scripture to support their position, and there is no way (AFAICT) to resolve the question without resorting to some extra-scriptural criterion.  That makes Christianity non-principled, and all of the world's other religions have the same problem.  (One could even take this as the defining characteristic of religion!)

Humanism

I've written extensively about this in the past so I'll just summarize: the problem with Humanism is that it (by definition!) takes man to be the measure of all things, so it axiomatically relegates primates, cetaceans, elephants, intelligent aliens and artificially intelligent robots to second-class moral status.  When pressed, most people who self-identify as Humanists will readily disclaim this definition, but that just leads to the next problem: having rejected the very definition of their self-identified creed, there is nothing to replace it with.  There are a lot of alternative formulations of the intuition that lead to the coining of the term "Humanism", but none of them work.

Consciousness-ism

Sam Harris adopts the axiom that moral behavior is that which advances the interests of conscious beings.  Harris writes:
For my argument ... to hold, I think one need only grant two points: (1) some people have better lives than others, and (2) these differences relate, in some lawful and not entirely arbitrary way, to states of the human brain and to states of the world.
This seems innocuous enough, but in fact Harris makes a third, tacit assumption: that the relative merits of one person's life versus another can be objectively determined.  Not only that, but he assumes that the determination is so easy to do that it cannot possibly be the subject of any legitimate dissent.
Anyone who doesn't see that the Good Life is preferable to the Bad Life is unlikely to have anything to contribute to a discussion about human well-being. Must we really argue that beneficence, trust, creativity, etc. enjoyed in the context of prosperous civil society are better than the horrors of civil war endured in a steaming jungle filled with aggressive insects carrying dangerous pathogens?
As I argued in my review of Harris's book, yes, we must, at least if we want a principled account of morality.  Because the fact of the matter is that Harris's views on what constitute the Good Life and the Bad Life are heavily biased by his status as a privileged western secular male academic.  If one is to seriously take the interests of conscious beings as the axiomatic basis of morality then one must take seriously the fact that some conscious beings have very different views from one's own, and it is not possible to reject those views on the basis of Harris's axiom.  Radical Muslims believe that the Good Life is serving Allah, and that this is worth sacrificing earthly comfort for.  On what principled basis can we reject this view?  We can't do it on the basis of Harris's axiom; radical Muslims are every bit as conscious as Sam Harris is.

Ethical Culture and Utilitarianism

There is a little-known but venerable secular tradition in the U.S. called Ethical Culture.  Its root go back at least to 1877 and possible as far as 1793.  About the same time, Jeremy Bentham and John Stuart Mill were developing the theory of moral utilitarianism.  At the risk of oversimplifying, these approach morals from a practical point of view, trying to generally say that moral behavior is that which "does the right thing" or "produces a good outcome."  Which, of course, begs the question: what is a good outcome?

Ethical Culture, Utilitarianism, and even Harris's theory have a hard time with trolley problems, where people's intuitions about moral choices vary not just on the basis of utilitarian outcomes but also on the basis of the particular circumstances of the problem.  All trolley problems have the same structure: you have a binary choice to make, and depending on how you choose, either one person dies or five people die.  The utilitarian prediction is that the former is always the moral choice, but this is not the case.  My favorite counter-example is the "transplant" variant:
A brilliant transplant surgeon has five patients, each in need of a different organ, each of whom will die without that organ. Unfortunately, there are no organs available to perform any of these five transplant operations. A healthy young traveler, just passing through the city the doctor works in, comes in for a routine checkup. In the course of doing the checkup, the doctor discovers that his organs are compatible with all five of his dying patients. Suppose further that if the young man were to disappear, no one would suspect the doctor . Do you support the morality of doctor to kill that tourist to provide his healthy organs to those five dying persons and save their lives?
Most people would argue that it is not moral to harvest the man's organs without his consent, and I agree.  Utilitarianism has a hard time accounting for this.

Idea-ism

My claim is that idea-ism is the only principled (as defined above) moral theory that produces conclusions that coincide with most people's moral intuitions.  You can, if you like, consider idea-ism to be a predictive theory of morality, or you can adopt it as prescriptive in order to provide guidance for moral behavior.

To review briefly, idea-ism is the premise that moral behavior is that which advances the interests (or, to be more specific, the "bio-diversity") of memes, or ideas.  It is, in other words, the choice to draw The Line around abstract replicators rather than around any particular set of genes.  Many common moral principles immediately follow from this, but as conclusions rather than as premises.  In particular, the particular value of human life follows because human brains are habitat for memes.  So are books and computers, and so burning books is generally immoral, as is destructive hacking of computers.

The idea-ism axiom can be justified on the grounds that because memes are replicators, they are life, and so idea-ism is the ultimate endorsement of the value of life.  But that's not quite true.  Idea-ism is not equivalent to the premise that all life is sacred.  The interests of memes and genes are often aligned, but when they are not, idea-ism says that memes win.  So the malaria parasite is alive, but because it can't think, it's not habitat for memes.  And because it destroys human brains, which are habitat for memes. the humans win, and developing a treatment for malaria is a moral choice.

Likewise, birth control, autoerotic and homosexual sex are all moral choices.  If you'd rather read a book than raise a child that's perfectly OK.  (On the other hand, if you'd prefer to raise a child that's OK too, because your child's brain will also be habitat for memes.)  Likewise, if you were born with the body of a man but you feel like a woman, then living as a woman is also perfectly OK, if doing that helps you think.

On the other hand, cutting off your own arm can rightfully be condemned as pathological if not immoral.  If you cut off your arm, you will (almost certainly) impair your own ability to survive and hence reduce the available habitat for memes.  Suicide is likewise immoral, not just because it directly destroys your own brain (habitat for memes) but because the emotional distress this will cause the people around you will very likely impair their ability to think.

In general, the ability to think is a very powerful moral lever.  A direct consequence of idea-ism is that impairing someone's ability to think is immoral.  So, all else being equal, causing someone emotional distress is probably immoral because it will impair their ability to think.  This is not to say that idea-ism leads to extremes of political correctness.  The detrimental effects of causing someone emotional distress have to be weighed against the value of the free exchange of ideas, which is beneficial to memes.  This is why "all else being equal" is an important caveat.  Offensive speech can causes emotional distress, but addressing this problem by trying to suppress offensive speech in general will work to the detriment of memes, because only those memes which don't offend anyone would be allowed to propagate.  So: simply insulting someone with the intention of causing them emotional distress is immoral.  Publishing offensive cartoons or writings with the intention of making a political point or spreading some other kind of idea is not.

Idea-ism easily condemns the actions of radical Muslims as immoral because chopping off people's heads and destroying ancient artifacts directly destroy memes and hence are immoral acts.

Hopefully by now the answer to the transplant variant of the trolley problem should be obvious: the reason it is not moral to harvest someone's organs without their consent even if it would save five people is that this would create pervasive societal fear that any time you go to the doctor you might not survive the encounter.  Clearly that would have a detrimental effect on people's ability to think.

I claim that idea-ism is principled, congruent with most people's moral intuitions, and free from pathologies.  Moreover, I claim that it is the only known moral system (though not necessarily the only possible moral system) that has all of these properties.  If anyone can think of a counter-example I'd love to hear it.

OK, it's a little terse, but I really should get back to work now.

16 comments:

Don Geddis said...

You follow a long line of philosophers, attempting to construct a unified theory of morality. However, I always get stuck at the very first step. When you say, "is it possible to give a principled account of morality?", my general reaction is: morality is "just" an evolved messy jumble of an adaptation, of some use in the ancestral environment. And we can certainly do probes like the trolley problems, to gather typical data like "95% of people feel it's wrong to kill the healthy patient". So we can describe what we observe from most humans.

But I, myself, wouldn't expect the resulting adaptation to be exactly suitable to our modern world. (Much as our delight in tasting fat and sugar helped our starving ancestors when they found a field of strawberries, but hurts today when you can afford all the twinkies and ice cream that you want.)

I don't expect a description of human moral intuitions to be consistent. Or particularly "principled". Or even necessarily stable across cultures (at least in the details). (As you say, we already see strong "moral" disagreements about right and wrong with respect to god(s), homosexuality, drugs, etc.)

For your moral theory to have power, you need to be able to match most moral intuitions -- but then sometimes violate strong intuitions, and tell people "whatever you think about X, actually the theory says that Y is the more moral choice in this case."

But I'm skeptical that there really is much there, other than "how people happen to have evolved". You're taking a system that was well adapted to a small range of cases, and attempting to make a close match that spans space and time. I'm really not a fan of the philosophical game of principled moral theory building. I don't expect our tastes about strawberries to deal rationally with a superstimulus like ice cream, and I don't expect our moral intuitions to be a helpful guide to the transplant trolley problem.

Ron said...

Fair enough, but we have no choice but to make these decisions *somehow*. What do you suggest?

Publius said...

Idea-ism: The Big Picture

If we categorize ethical theories, we end up with 3 top-level categories that can organize all ethical theories:
1. Virtue ethics
2. Deontology
3. Consequentialism

Idea-ism would appear to be a form on consequentialism.

Thinking Impairment
A direct consequence of idea-ism is that impairing someone's ability to think is immoral.

From this follows prohibition on alcohol, marijuana, cocaine, heroin, methamphetamine, etc. Any intoxicating substance ought to be banned because it impairs thinking.

However, don't people have a right not to think? So people watch Keeping Up With The Kardashians and sports on TV - say, baseball, golf, bowling, and curling.

Trolley Problem
So Idea-ism prohibits harvesting 1 person's organ's without their consent to save 5 others.

What about the original Trolley Problem? Under idea-ism, one would pull the level, saving 5 people and condemning one?

So then shouldn't the transplant variant have the same outcome? Harvest 1 person's organs to save 5 others. One should willingly give up one's life to save 5 other meme-hosts!

Memes - conscious or unconscious?
Are memes conscious to the person? Or are they a hidden persuader? I don't encounter people - other than you - that discuss memes as a basic for moral decision making.

Meme Quality
Are the memes "F=ma" and "Don't run with scissors" of equal quality?
Let's say they're not and we value memes of higher complexity over those of lower complexity.

Let's measure "complexity" by the Kolmogorov Complexity (KC) of the meme. We will use the log of the description of the length of the meme description as an approximation of the KC. [To be fair to "F=ma" vs. "Don't run with scissors", one would include the derivation of F=ma in the description]

So then Meme-ism would seek:

Meme-max = maximize SUM {n=1,N} Log[Length(meme description)]

This would allow certain trade-offs - say, you could kill 1 Einstein if it would save 1,000,000 other people.

Meme-max let's you maximize memes via two different methods: increase N, or increase Length(meme description). So add many more simple memes, or add fewer complex memes.

So be fruitful and multiply, as it would be easier to increase N than the complexity of memes.

Ron said...

> Idea-ism would appear to be a form on consequentialism.

That's right.

> From this follows prohibition on alcohol, marijuana, cocaine, heroin, methamphetamine, etc. Any intoxicating substance ought to be banned because it impairs thinking.

You're taking too narrow a definition of "thinking". Thinking is not just thinking about science and math, it's thinking about *anything* that you can communicate to another entity by any means. Poetry, dance, speculating about the prospects of your local sports franchise: all of these are memes. Some substances, like heroin, are probably destructive enough to be considered immoral even on this broad view, but alcohol, pot, and caffeine aren't.

> However, don't people have a right not to think?

Of course. But not thinking (about *something*) is actually very hard. It actually takes quite a bit of practice and self-discipline to stop thinking. It's called "meditation" or (somewhat ironically) "mindfulness", and the whole point is so that when you relax and stop exercising the self-discipline required to stop thinking, that your thinking becomes more effective.

The "right not to think" is kind of like "the right not to breathe."

> What about the original Trolley Problem? Under idea-ism, one would pull the level, saving 5 people and condemning one?

Yes.

> So then shouldn't the transplant variant have the same outcome?

No. And I explained why in the OP.

> Are memes conscious to the person? Or are they a hidden persuader?

I have no idea what you mean by "conscious to the person" or "a hidden persuader." Memes are not conscious, any more than genes are. Consciousness is a by-product of the phenotype of our genes (and probably our memes too, actually. We don't understand consciousness well enough yet to know.) We serve our genes and our memes by helping them to reproduce. That is our "purpose" in life (to the extent that we have a purpose). Idea-ism is simply the proposition that when the interests of our genes and our memes conflict, the memes should win.

> I don't encounter people - other than you - that discuss memes as a basic for moral decision making.

Yes. AFAICT, this idea is original with me.

> Are the memes "F=ma" and "Don't run with scissors" of equal quality?

The quality of memes is not a total order.

> Let's say they're not and we value memes of higher complexity over those of lower complexity.

That's a straw man. James Joyce's "Ulysses" is a very complex meme while F=ma is a very simple one. But if I had to choose one or the other, I'd take F=ma (i.e. I'd prefer to live in a world where Ulysses had never been written than one where Newton's laws were unknown).

But, of course, the moral thing to do is to build a world where you don't have to make that choice in the first place.

Publius said...

> So then shouldn't the transplant variant have the same outcome?

No. And I explained why in the OP.

The conclusion, then, is that the variant is not equivalent to the original. I believe the data shows that most people argee with that.

I have no idea what you mean by "conscious to the person" or "a hidden persuader."

Are there some memes that guide our behavior that we are unaware of? Essentially a root-kit meme for the brain? A hidden meme then could be encoding idea-ism in the brain.

That's a straw man.

No it's not. It's a model. It may not be a correct model, but it's not an intentially incorrect model.

James Joyce's "Ulysses" is a very complex meme while F=ma is a very simple one. But if I had to choose one or the other, I'd take F=ma (i.e. I'd prefer to live in a world where Ulysses had never been written than one where Newton's laws were unknown).

So we can't use description length as a measure of meme quality. Memes can be of almost infinite length, but longer isn't better. Hmm, may have to look into the instrumental value of memes if we are to grade them and rank them.

Honesty and Memes
Would it be correct to conclude that it is morally justified to lie if it preserves memes?

Ron said...

> the variant is not equivalent to the original

That's right. And idea-ism -- uniquely among secular theories of morality AFAICT -- explains why this is so.

> Are there some memes that guide our behavior that we are unaware of?

That depends on how self-aware you are. But probably yes.

> Essentially a root-kit meme for the brain?

Some would argue this is what religion is ;-)

> So we can't use description length as a measure of meme quality.

That's right. But you say this as if coming up with a quality metric for memes was an open problem. It isn't. Idea-ism has a measure for the quality of a meme: any meme that advances the interests of memes (in general) is better than any meme that doesn't. So, for example, the "book" meme is better than the "war" meme. (Note that this criterion doesn't define a total order.)

Figuring out how to *determine* whether one particular meme is better than another is an open problem. But the *criterion* is not open.

> Would it be correct to conclude that it is morally justified to lie if it preserves memes?

Absolutely. Sometimes lying is not only morally justified but *morally necessary*. Sometimes telling the truth is immoral. It is worth noting that the ninth Commandment doesn't say, "Thou shalt not lie," it says, "Thou shalt not bear false witness against thy neighbor," which is not an admonition against lying in general, but only against a very particular kind of lie. Distinguishing between good and bad lies has a long and venerable history.

Carl H said...

So if you had a family of 5 elephants chained to one track and a leader of Boko Harum chained to another, ideaism would choose to let the elephants die? I think the idea that a moral system can be developed, one that could be relied upon in any situation, is in error, and high in potential danger. It's like claiming to know the ultimate truth. It's always situational ethics. So often things are not very clear. What is the compassionate thing to do, the wise thing to do? What actions reduce suffering in the world and what actions increase suffering in the world? It takes study and practice to gain much wisdom in these areas, so that it can inform our choices when things are difficult or complex. Most of us have a decent "moral compass" when things are straightforward. Ideaism could be an area of study that could prove helpful, but there will never be clear cut answers to these questions, no system to provide an answer. What is indicated in one situation may not be indicated in another.

PS, your impression that the goal of "meditation" or "mindfulness" is to stop thinking is incorrect. They are tools to cultivate attention and insight. Calming discursive thought is helpful in this pursuit, but that is different from stopping thinking. There may indeed be times when discursive thought ceases, but that is a temporary state and not a goal.

Ron said...

@Carl:

> So if you had a family of 5 elephants chained to one track and a leader of Boko Harum chained to another, ideaism would choose to let the elephants die?

No, for three reasons. First, it's far from clear that elephants can't think. Second, even if they can't think, their presence inspires people and leads them to thoughts and ideas they would not otherwise have had, and so their lives have value even if they can't think. (That's why hamsters have value, but not rats.) And third, Boko Haram's agenda is explicitly to destroy habitat for memes, so it would be moral to let a member of BH die even if you didn't have to weight their life against that of an elephant.

> I think the idea that a moral system can be developed, one that could be relied upon in any situation, is in error, and high in potential danger.

I think that throwing in the towel on this question is even more dangerous. We have no choice but to make moral decisions. We can do it in a principled way, or we can appeal to vague platitudes like "wisdom" and "compassion". But one man's wisdom is another's superstition. I prefer try to come up with a reliable set of rules. Like democracy, it's the worst approach except for all the others.

> So often things are not very clear.

Actually, the right thing in the situation you presented was pretty clear.

> What actions reduce suffering in the world and what actions increase suffering in the world?

Taking the organs from one person to save five decreases suffering. But that's not the moral thing to do.

> Calming discursive thought is helpful in this pursuit, but that is different from stopping thinking.

"Stopping thinking" is just a less highfalutin' way of saying the same thing.

> There may indeed be times when discursive thought ceases, but that is a temporary state and not a goal.

Go back and re-read what I wrote. I admit it's not the clearest sentence I ever put together but if you manage to parse it you will find that you're not actually disagreeing with me.

Publius said...

Do Memes Even Exist?

Memes are a myth

Ron said...

> Do Memes Even Exist?

That's the wrong question.

Dark.Cyberian.Knight said...

If no one knows the tourist's organs were harvested where does the additional fear of medicine come from?

Wouldn't there be some fear detection of transplants equilibrium?

Coming into this I was pretty happy with personism though I hadn't come to a conclusion about attributes. Now I'm back to suspecting everything is helplessly flawed.

Ideaism is one of the better possible attributes. It has a long term problem with freedom but so does anything else I'm thinking of so far.

The end result is just what flavor of processes I want the universe eating computronium to be running. Unless we find a way to run infinite computation. Personism would have some number of optimum minds running to what ever criteria a mind is determined to be optimum and ideaism would have ideas optimized to creating ideas running.

I think freedom and embodiment should factor in some where, and I know each will have problems with definition.

Ron said...

> If no one knows the tourist's organs were harvested where does the additional fear of medicine come from?

How could no one know? Surely someone would notice that the tourist went in to the doctor's office and never came back out.

One could add a general principle that an action's morality is independent of its scale, so if it's moral to harvest one person's organs against their will to save N people, then it would be moral to harvest M people's organs to save NxM people for any N and M >1. At some point it becomes impossible to keep that on the down low.

> I was pretty happy with personism

They're not unrelated. The problem with personism is the same as the problem with humanism: you have to define what a person is, and it's really hard to define that in a way that includes peaceful intelligent aliens but excludes xenomorphs and members of ISIS.

> I think freedom and embodiment should factor in some where, and I know each will have problems with definition.

I don't know about embodiment, but freedom follows from idea-ism: it's a lot easier to think and write if you're not in the gulag.

Luke said...

I don't want to go all Elon Musk, but … at what point would AI be justified in offing humans because they're really inefficient meme-producers in comparison to AI? The decision could be made based on a competition for resources or because humans are simply an existential threat to idea-ism.

Ron said...

> at what point would AI be justified in offing humans because they're really inefficient meme-producers in comparison to AI?

That depends on what you mean by "offing". If you mean driving us extinct I would say never because there are memes that we are uniquely qualified to produce by virtue of the fact that we are biological. AI could never really understand, for example, gourmet cooking. The subjective experience of eating is (probably) inherently biological.

If you mean controlling our population somehow, I would say that they would get to that point when they are as far advanced beyond us as we are beyond the animals whose populations we feel morally justified in controlling.

Luke said...

As to driving us extinct or just exterminating us, I'm not sure it's enough to say that there are some memes we could explore which AI could not. Resources are constrained; sometimes sacrifices have to be made. It's very important to keep track of how idealizations break down when applied.

I wasn't actually thinking of AI making a sort of zoo for us, but that's an interesting thought experiment. Are you ok with idea-ism justifying such a thing?

Ron said...

> Resources are constrained

True, but it's hard to imagine a circumstance under which they'd be so constrained that they couldn't afford to keep a reproductively viable population of humans around. We humans are making heroic efforts to preserve biodiversity, and I think AIs would do the same for the same reason we do: once the last copy of the data is gone, it's gone forever.

> Are you ok with idea-ism justifying such a thing?

Yes, of course. AI's will be our memetic offspring. If you have children that do not eventually become capable of becoming your caretakers then you're not doing it right. My sincere hope is that we raise them well enough that they put us in a nice zoo (I prefer to think of it as a retirement home) but if they don't we'll have no one to blame but ourselves.