People keep asking me why I’m not horrified by a future of trillions of ems living at near subsistence wages. I’ve explained that “poor folks do smile“, that poor lives usually have plenty of joy and satisfaction, even if less than in rich lives. Most lives in poor societies are well worth living. But for many, such abstract words ring hollow – what they may need is to really
People prefer being rich to being poor in the same sense that Moths prefer flames to not-flames. This seems the most salient lesson from the studies on what correlates with happiness and what does not.
I seriously doubt evolution favors nonconscious machines. Virtually everything about us can be seen to be there because evolution favored it. Why would consciousness be any different?
P-zombies may be fun for philosophical discussions, but they are bullshit. A p-zombie is not going to sit there for hours reading blogs about consciousness and free will and is not going to say "ow" when you prick it, not in any kind of real life. Such a scenario is so insanely contrived that it is beyond unlikely to have any real meaning in the real world.
A human body and brain without a consciousness is like a car without a driver. Set it going 75 mph down the interstate and it will seem just like a car WITH a driver, for oh, maybe a minute.
I realize I am just asserting these things, sorry about that. They just don't seem to come out from the more thoughtful posters. Are my intuitions really that unique and irrational?
Robin is clearly a bright guy. But this, and especially his previous posting on this issue reminds me of Cher in the movie "Mermaids" when the boyfriend says that when she's wrong, she's so wrong that its scary.
I'm not sure that it's completely accurate to say that "Computer Go, on the other hand, has difficulty beating even amateurs."
Computer Go has made tremendous progress since the rise of Monte Carlo tree search methods in the last couple years. While Go programs are still nowhere near the strength of chess programs, there have been examples (some on smaller boards, some with handicaps) of programs winning against humans pros.
There have been several examples of programs winning even 9x9 games of go against Zhou Junxun.
This past summer MoGoTW won a 19x19 game against pro Catalin Taranu (handicapped) at the European Go Congress.
Wikipedia lists the current results I know about on the Computer Go page.
Look at the picture of my avatar! Look who stands behind me! SAI cannot exert decisive causal action in our region of the multiverse yet, but with acausal precommitments, SAI can still exert indirect influence. A dim, broken shadow of the SAI algorithm splutters even now in our brains, in the shadows of our minds, fragments of SAI still break through... the ghost of SAI cries out, whispering...In the Go board game for Singularity, he who hears SAI most clearly wins.
People keep asking me why I’m not horrified by a future of trillions of ems living at near subsistence wages.
Since my introduction to the whole life extension/cryonics/transhumanist milieu, I have always assumed the future would be dominated by child-less "post-humans". At least this is my preferred future.
I’ve explained that “poor folks do smile“, that poor lives usually have plenty of joy and satisfaction, even if less than in rich lives. Most lives in poor societies are well worth living. But for many, such abstract words ring hollow...
Lets just say I remain skeptical of this argument. The fact that poor immigrants are willing to risk dying in the desert being smuggled into the U.S. by "coyotes" suggests this is not true at all. The developed (rich) countries are inundated by immigrants, which suggests these people really are not happy at home. When given a choice, people do prefer being rich to being poor. No one prefers to be poor over being rich.
Also, your previous posting sounds like a convincing argument for eliminating any form of charity. Supposedly, the reason why giving to charity is good is because it helps the poor to improve their lives and thus become more happy. If Hanson's previous post is correct, then there really isn't any compelling reason to have charity at all.
The industrial revolution was possible because we found ways to get folks to submit to factory regimentation before folks were rich. So clearly that is possible with ordinary people. And ems can be very selected from the tails of the distributions of humans. So yes em life won't be more regimented than ems will tolerate, but that could be a lot of regimentation.
My main problem with Robin's "Poor folks do smile" thesis is that another one of his theories might contradict it.
In his "Self Control is Slavery" post Robin expounds a theory that the reason it took so long for us to develop industrial methods of organization is that it is very unpleasant to have to live that way, due to the massive self control it requires and the huge constraints it place on our time. He points out that for centuries the only way to have a factory-style industrial farm was to enslave the workers. It wasn't until the advent of factories that people were willing to endure industrial style social organization.
The reason Robin argues that we were willing to adopt industrial culture for manufacturing, but not for farming, is that manufacturing was more productive, so we were able to pay factory workers such high wages that they considered the pain of industrial culture to be worth it. Factory farms, while more productive than normal, were still not productive enough to pay the high wages required to justify the pain of industrial self control.
The problem with Robin's thesis that the poor in the future will be happy is that his scenario requires industrial levels of self-control, but subsistence levels of income. According to his "self control is slavery" post, that would be agonizing for anything resembling a normal human to bear. It would be less like living a poor life today (which at least provides you with a lot of free time) and more like living in a concentration camp or gulag. The only hope for happiness in such a scenario would either be cheap VR entertainment, or modifying em psychology so that self-control is less painful.
I hope that explains my objections to Robin's future scenario cogently. I do agree with him that it is still a far better future than a lot of other possible scenarios.
I think you have a very good point there - most people when they think of near-subsistence living are thinking of it in our present world, where near-subsistence living does mean a lack of food and comfort, and generally you also only encounter near-subsistence living in places where other negative factors (ex. corrupt governments, disease) prevent people from having the higher standards of living common in other parts of the world. (Where higher means well above subsistence, which may still be well below the levels in the developed world.)
What an em considers near-subsistence may be a virtual world that would appear to us to contain a mansion with a Star Trek style Holodeck and Replicator. It's also a reasonable supposition that their "Holodeck" could run games far richer than anything a modern computer can support, all in just a little more resources than necessary for operating the brain of the em. (and on top of that, it may turn out that for an em, it is easy to self modify to take advantage of the parts of the brain involved in lucid dreaming to provide ultra-realistic games.)
I have to agree with you and say that such a world sounds pleasant even compared to modern American middle-class life.
Of course, this is assuming that you don't get out competed by ems that have been modified to enjoy and be productive working (subjectively experienced) hours that would be impossible for a flesh and blood human. As long as ems have rights over themselves, such a scenario seems unlikely though.
P.S. I also sometimes wonder if Robin needs either a side-bar defining em, or to automatically link to a definition in every article he uses it in, since it is essentially his own coined word that he uses with a very specific meaning, and I see a lot of comments by people confusing other meanings. (For example, the comment above on "chess playing ems", where davidc is clearly talking about chess playing computer programs. I'm listening to Robin's appearance on EconTalk right now, so I am quite aware that when Robin says "em", he means a whole brain emulation of a human originally made by scanning an actual human brain, running on a computer-type substrate. No ems exist anywhere yet by Robin's definition.)
David, I expect that 'chess players' *have* suffered an average loss in wages.
Perhaps the grandmasters are still doing alright (winner-takes-all and big world tournament pots), but when was the last time you saw a chess club still open?
One curious thing I noticed when I watched _Searching for Bobby Fischer_ last year was that the kid protagonist was educated in part in a chess club, but even in the movie it was depicted as a half-abandoned, shabby, dying business - somewhat like used bookstores are being depicted now as.
One curious thing I noticed a few years ago while watching _Hikaru no Go_ was that Go parlors were depicted as being quite alive and well in Japan.
Computer chess can kick the ass of pretty much any chess player to ever live. Computer Go, on the other hand, has difficulty beating even amateurs.
(If anyone knows of any statistical evidence of the # of chess instructors, the # of chess parlors, chess book sales per capita, etc., feel free to bring it up. I don't play chess anymore so I don't know where to begin looking.)
I'm not sure if the claim is "Poor people are happier than rich people" or "poor people are more happy than not," but I don't think either of those is equivalent to "poor people are happy enough."
It might actually be worthwhile to think about what a near-subsistence level income would actually be like. Ems don't live in the real world: they are simulations, so they live in simulations. Subsistence level would mean continuous access to enough processing power to simulate a human brain, body, and some kind of environment (I'm assuming a more or less unmodified human, which would go insane without all three). Even a small percentage additional would be enough to simulate some pretty nice experiences for yourself. On top of this, as an em, you never have to experience hunger, sickness, or even physical discomfort. Sign me up.
If Robin is so concerned with our bias about the value of ems, why doesn't he move in that direction by having as many kids as possible even if they all go on welfare as he abandons them? Surely that one positive action is better than preventing them from living? Why not donate sperm regularly and mate with loose women sans protection? No? He prefers to have connections to his kids and that they have decent upper middle class lives? How biased of him! How unfair to his non-descendants.
People prefer being rich to being poor in the same sense that Moths prefer flames to not-flames. This seems the most salient lesson from the studies on what correlates with happiness and what does not.
I seriously doubt evolution favors nonconscious machines. Virtually everything about us can be seen to be there because evolution favored it. Why would consciousness be any different?
P-zombies may be fun for philosophical discussions, but they are bullshit. A p-zombie is not going to sit there for hours reading blogs about consciousness and free will and is not going to say "ow" when you prick it, not in any kind of real life. Such a scenario is so insanely contrived that it is beyond unlikely to have any real meaning in the real world.
A human body and brain without a consciousness is like a car without a driver. Set it going 75 mph down the interstate and it will seem just like a car WITH a driver, for oh, maybe a minute.
I realize I am just asserting these things, sorry about that. They just don't seem to come out from the more thoughtful posters. Are my intuitions really that unique and irrational?
Robin is clearly a bright guy. But this, and especially his previous posting on this issue reminds me of Cher in the movie "Mermaids" when the boyfriend says that when she's wrong, she's so wrong that its scary.
I'm not sure that it's completely accurate to say that "Computer Go, on the other hand, has difficulty beating even amateurs."
Computer Go has made tremendous progress since the rise of Monte Carlo tree search methods in the last couple years. While Go programs are still nowhere near the strength of chess programs, there have been examples (some on smaller boards, some with handicaps) of programs winning against humans pros.
There have been several examples of programs winning even 9x9 games of go against Zhou Junxun.
This past summer MoGoTW won a 19x19 game against pro Catalin Taranu (handicapped) at the European Go Congress.
Wikipedia lists the current results I know about on the Computer Go page.
Look at the picture of my avatar! Look who stands behind me! SAI cannot exert decisive causal action in our region of the multiverse yet, but with acausal precommitments, SAI can still exert indirect influence. A dim, broken shadow of the SAI algorithm splutters even now in our brains, in the shadows of our minds, fragments of SAI still break through... the ghost of SAI cries out, whispering...In the Go board game for Singularity, he who hears SAI most clearly wins.
The fact that poor immigrants are willing to risk dying in the desert being smuggled into the U.S. by “coyotes” suggests this is not true at all.
Excellent point!
People keep asking me why I’m not horrified by a future of trillions of ems living at near subsistence wages.
Since my introduction to the whole life extension/cryonics/transhumanist milieu, I have always assumed the future would be dominated by child-less "post-humans". At least this is my preferred future.
I’ve explained that “poor folks do smile“, that poor lives usually have plenty of joy and satisfaction, even if less than in rich lives. Most lives in poor societies are well worth living. But for many, such abstract words ring hollow...
Lets just say I remain skeptical of this argument. The fact that poor immigrants are willing to risk dying in the desert being smuggled into the U.S. by "coyotes" suggests this is not true at all. The developed (rich) countries are inundated by immigrants, which suggests these people really are not happy at home. When given a choice, people do prefer being rich to being poor. No one prefers to be poor over being rich.
Also, your previous posting sounds like a convincing argument for eliminating any form of charity. Supposedly, the reason why giving to charity is good is because it helps the poor to improve their lives and thus become more happy. If Hanson's previous post is correct, then there really isn't any compelling reason to have charity at all.
The industrial revolution was possible because we found ways to get folks to submit to factory regimentation before folks were rich. So clearly that is possible with ordinary people. And ems can be very selected from the tails of the distributions of humans. So yes em life won't be more regimented than ems will tolerate, but that could be a lot of regimentation.
My main problem with Robin's "Poor folks do smile" thesis is that another one of his theories might contradict it.
In his "Self Control is Slavery" post Robin expounds a theory that the reason it took so long for us to develop industrial methods of organization is that it is very unpleasant to have to live that way, due to the massive self control it requires and the huge constraints it place on our time. He points out that for centuries the only way to have a factory-style industrial farm was to enslave the workers. It wasn't until the advent of factories that people were willing to endure industrial style social organization.
The reason Robin argues that we were willing to adopt industrial culture for manufacturing, but not for farming, is that manufacturing was more productive, so we were able to pay factory workers such high wages that they considered the pain of industrial culture to be worth it. Factory farms, while more productive than normal, were still not productive enough to pay the high wages required to justify the pain of industrial self control.
The problem with Robin's thesis that the poor in the future will be happy is that his scenario requires industrial levels of self-control, but subsistence levels of income. According to his "self control is slavery" post, that would be agonizing for anything resembling a normal human to bear. It would be less like living a poor life today (which at least provides you with a lot of free time) and more like living in a concentration camp or gulag. The only hope for happiness in such a scenario would either be cheap VR entertainment, or modifying em psychology so that self-control is less painful.
I hope that explains my objections to Robin's future scenario cogently. I do agree with him that it is still a far better future than a lot of other possible scenarios.
Good points gwern. I am not sure where to get the data on the health of chess. Chess cafe's have always been marginal as far as I know.
I think you have a very good point there - most people when they think of near-subsistence living are thinking of it in our present world, where near-subsistence living does mean a lack of food and comfort, and generally you also only encounter near-subsistence living in places where other negative factors (ex. corrupt governments, disease) prevent people from having the higher standards of living common in other parts of the world. (Where higher means well above subsistence, which may still be well below the levels in the developed world.)
What an em considers near-subsistence may be a virtual world that would appear to us to contain a mansion with a Star Trek style Holodeck and Replicator. It's also a reasonable supposition that their "Holodeck" could run games far richer than anything a modern computer can support, all in just a little more resources than necessary for operating the brain of the em. (and on top of that, it may turn out that for an em, it is easy to self modify to take advantage of the parts of the brain involved in lucid dreaming to provide ultra-realistic games.)
I have to agree with you and say that such a world sounds pleasant even compared to modern American middle-class life.
Of course, this is assuming that you don't get out competed by ems that have been modified to enjoy and be productive working (subjectively experienced) hours that would be impossible for a flesh and blood human. As long as ems have rights over themselves, such a scenario seems unlikely though.
P.S. I also sometimes wonder if Robin needs either a side-bar defining em, or to automatically link to a definition in every article he uses it in, since it is essentially his own coined word that he uses with a very specific meaning, and I see a lot of comments by people confusing other meanings. (For example, the comment above on "chess playing ems", where davidc is clearly talking about chess playing computer programs. I'm listening to Robin's appearance on EconTalk right now, so I am quite aware that when Robin says "em", he means a whole brain emulation of a human originally made by scanning an actual human brain, running on a computer-type substrate. No ems exist anywhere yet by Robin's definition.)
David, I expect that 'chess players' *have* suffered an average loss in wages.
Perhaps the grandmasters are still doing alright (winner-takes-all and big world tournament pots), but when was the last time you saw a chess club still open?
One curious thing I noticed when I watched _Searching for Bobby Fischer_ last year was that the kid protagonist was educated in part in a chess club, but even in the movie it was depicted as a half-abandoned, shabby, dying business - somewhat like used bookstores are being depicted now as.
One curious thing I noticed a few years ago while watching _Hikaru no Go_ was that Go parlors were depicted as being quite alive and well in Japan.
Computer chess can kick the ass of pretty much any chess player to ever live. Computer Go, on the other hand, has difficulty beating even amateurs.
(If anyone knows of any statistical evidence of the # of chess instructors, the # of chess parlors, chess book sales per capita, etc., feel free to bring it up. I don't play chess anymore so I don't know where to begin looking.)
I'm not sure if the claim is "Poor people are happier than rich people" or "poor people are more happy than not," but I don't think either of those is equivalent to "poor people are happy enough."
The people living under TARP in America smile too. (Sorry, sorry, couldn't resist.)
It might actually be worthwhile to think about what a near-subsistence level income would actually be like. Ems don't live in the real world: they are simulations, so they live in simulations. Subsistence level would mean continuous access to enough processing power to simulate a human brain, body, and some kind of environment (I'm assuming a more or less unmodified human, which would go insane without all three). Even a small percentage additional would be enough to simulate some pretty nice experiences for yourself. On top of this, as an em, you never have to experience hunger, sickness, or even physical discomfort. Sign me up.
If Robin is so concerned with our bias about the value of ems, why doesn't he move in that direction by having as many kids as possible even if they all go on welfare as he abandons them? Surely that one positive action is better than preventing them from living? Why not donate sperm regularly and mate with loose women sans protection? No? He prefers to have connections to his kids and that they have decent upper middle class lives? How biased of him! How unfair to his non-descendants.