12 ms·

When you say "survival" in terms of Darwinian evolution, you're really talking about survival of specific genes. Why should we choose to be beholden to the prim

by feoren 3y ago

When you say "survival" in terms of Darwinian evolution, you're really talking about survival of specific genes. Why should we choose to be beholden to the primal "desires" (if you'll grant me the intentional stance) of our genes? We are the AI that our DNA invented that turned around and overthrew it. The planet is ours now, not our DNA's. There are many other things that evolve too: ideas, memes, viruses, and many more if you relax "evolution" to simply mean "progress", like my own knowledge and the capabilities of the human race. It's myopic to focus on genes alone. > eugenics Eugenics takes for granted the idea that some "genetic purity" is good for its own sake. Why? Why is a world with fewer, but "genetically superior" people, (and more genetically similar), better than a world with more people and more genetic variety? That's not even necessarily good even if you myopically focus on the actual gene pool, much less if you take all the other aspects of humanity into account. > So, how exactly do we know which is good, which is bad? Is there even an objective good or bad without invoking the idea that some higher beings say so? Yes, there is: we all collectively decide properties of the world we'd like to live in, and we try to guess what kind of world future generations would also like to live in, and then we call those actions that ultimately improve those properties "good" and those that lessen those properties "bad", and we can even partially rank actions based on the integral of their effect on these properties from now until the end of the universe (allowing a Net Present Value calculation to avoid infinities, to acknowledge our imprecision of future outcomes and preferences, and even to simply prefer utility now over the same utility later). We try to maximize happiness (in its various forms, some more important than others) and minimize suffering (ditto), taking diminishing returns into account so we don't just make one dude trillions-happy at the expense of everyone else. This has infinite degrees of freedom and requires perfect knowledge of the future, so we know it's not actually possible to achieve, but it gives us a direction to orient our moral compasses and strive to get as close to this perfection as we can.

vacuity 3y ago
Objective, huh. Anyways, here's a problem in your last paragraph: We?
- feoren 3y ago
  Yes, we. See my sibling comment for my response to this.
HKH2 3y ago
> Eugenics takes for granted the idea that some "genetic purity" is good for its own sake. Why? Why is a world with fewer, but "genetically superior" people, (and more genetically similar), better than a world with more people and more genetic variety? That's not even necessarily good even if you myopically focus on the actual gene pool, much less if you take all the other aspects of humanity into account. 'Genetic purity' is not a charitable way of looking at eugenics; how about 'genetic quality'? That's what most of us seek. Nobody asks for a disabled baby. Should we edit our genes to allow for more diversity? How about all the diverse genetic mutations we haven't dreamed of yet! We could rediscover a lot of weird and wonderful maladaptations that our ancestors have spent millions of years weeding out. We'll have X-Men in no time.
beltsazar 3y ago
> Yes, there is: we all collectively decide properties of the world we'd like to live in, and we try to guess what kind of world future generations would also like to live in, and then we call those actions that ultimately improve those properties "good" and those that lessen those properties "bad" First of all, who are "we"? Second of all, you just proved my point, unless you want to redefine "objective" as something that the majority accept. Also, why are those properties desirable? Because the majority prefer it? There would be no objective good or bad, then—it's just the preference of the majority.
- feoren 3y ago
  See my sibling comment for my response to this.
feoren 3y ago
People are responding that I'm using the word "we" and therefore I'm not being objective. I think I need to emphasize a point better: there are infinitely many objective measures. Here's one: how many total paperclips humanity produces over the lifespan of the universe. That's a truly objective measure; we can just count them. We can analyze every action any human ever takes and estimate whether it increases or decreases the total number of paperclips any human will ever create. There's no subjectivity at all -- it's completely objective. While objective, paperclip count is notorious for not really aligning with our intuition of good vs. bad; I'm pretty sure we can do better. But this illustrates the point that the problem is really about picking from the infinitely many objective measures that, objectively, do already exist. Humans exist; that is objectively true. Humans have preferences; that is objectively true. Humans take pleasure from some experiences and experience pain from others; that is objectively true, albeit a big simplification. It's objectively true that humans dislike having rusty nails jammed into their eyes. It's objectively true that humans enjoy eating nice meals when they're hungry. We should seek to perform those actions which minimize the number of rusty nails jammed into eyes and maximize the number of times humans can enjoy nice meals when they're hungry. Obviously there are trillions more considerations, but nowhere has subjectivity even entered the picture yet. What I'm saying is that we can pick, or define, or refine, a truly objective measure; one that is much better than paperclip count. You could say "well there's subjectivity in picking which measure we use " -- only because we have imperfect knowledge of the universe and its future. If we had perfect knowledge of all the likes and dislikes of all humans that will ever exist, all sources of pain and pleasure and suffering and contentment and sadness and happiness, all their values and goals and desires, and how all actions affect all of that, then we could simply integrate their effects over time. Yes, we have to weight them appropriately, and find a good NPV discount rate, and all that, but with perfect knowledge, we could easily find one that's at the very least on the Pareto surface. That's our metric. So our subjectivity comes from imperfect knowledge. But that's OK -- we accept that we have imperfect knowledge. There's a huge difference between "there's no objective measure of good vs. evil" and "it's really hard to pick an optimal measure of good vs. evil, and to know which actions maximize the net present value of that metric over the course of the universe." The former gives you nothing; the latter at least orients your moral compass toward something that humanity can all work toward together.

by feoren 3y ago

vacuity 3y ago

Objective, huh. Anyways, here's a problem in your last paragraph: We?

feoren 3y ago
Yes, we. See my sibling comment for my response to this.

HKH2 3y ago

> Eugenics takes for granted the idea that some "genetic purity" is good for its own sake. Why? Why is a world with fewer, but "genetically superior" people, (and more genetically similar), better than a world with more people and more genetic variety? That's not even necessarily good even if you myopically focus on the actual gene pool, much less if you take all the other aspects of humanity into account. 'Genetic purity' is not a charitable way of looking at eugenics; how about 'genetic quality'? That's what most of us seek. Nobody asks for a disabled baby. Should we edit our genes to allow for more diversity? How about all the diverse genetic mutations we haven't dreamed of yet! We could rediscover a lot of weird and wonderful maladaptations that our ancestors have spent millions of years weeding out. We'll have X-Men in no time.

beltsazar 3y ago

> Yes, there is: we all collectively decide properties of the world we'd like to live in, and we try to guess what kind of world future generations would also like to live in, and then we call those actions that ultimately improve those properties "good" and those that lessen those properties "bad" First of all, who are "we"? Second of all, you just proved my point, unless you want to redefine "objective" as something that the majority accept. Also, why are those properties desirable? Because the majority prefer it? There would be no objective good or bad, then—it's just the preference of the majority.

feoren 3y ago
See my sibling comment for my response to this.

feoren 3y ago

People are responding that I'm using the word "we" and therefore I'm not being objective. I think I need to emphasize a point better: there are infinitely many objective measures. Here's one: how many total paperclips humanity produces over the lifespan of the universe. That's a truly objective measure; we can just count them. We can analyze every action any human ever takes and estimate whether it increases or decreases the total number of paperclips any human will ever create. There's no subjectivity at all -- it's completely objective. While objective, paperclip count is notorious for not really aligning with our intuition of good vs. bad; I'm pretty sure we can do better. But this illustrates the point that the problem is really about picking from the infinitely many objective measures that, objectively, do already exist. Humans exist; that is objectively true. Humans have preferences; that is objectively true. Humans take pleasure from some experiences and experience pain from others; that is objectively true, albeit a big simplification. It's objectively true that humans dislike having rusty nails jammed into their eyes. It's objectively true that humans enjoy eating nice meals when they're hungry. We should seek to perform those actions which minimize the number of rusty nails jammed into eyes and maximize the number of times humans can enjoy nice meals when they're hungry. Obviously there are trillions more considerations, but nowhere has subjectivity even entered the picture yet. What I'm saying is that we can pick, or define, or refine, a truly objective measure; one that is much better than paperclip count. You could say "well there's subjectivity in picking which measure we use " -- only because we have imperfect knowledge of the universe and its future. If we had perfect knowledge of all the likes and dislikes of all humans that will ever exist, all sources of pain and pleasure and suffering and contentment and sadness and happiness, all their values and goals and desires, and how all actions affect all of that, then we could simply integrate their effects over time. Yes, we have to weight them appropriately, and find a good NPV discount rate, and all that, but with perfect knowledge, we could easily find one that's at the very least on the Pareto surface. That's our metric. So our subjectivity comes from imperfect knowledge. But that's OK -- we accept that we have imperfect knowledge. There's a huge difference between "there's no objective measure of good vs. evil" and "it's really hard to pick an optimal measure of good vs. evil, and to know which actions maximize the net present value of that metric over the course of the universe." The former gives you nothing; the latter at least orients your moral compass toward something that humanity can all work toward together.