I think there is much to what Yudkowsky is saying on the topic in this post:
http://lesswrong.com/lw/hh4/the_robots_ai_and_unemployment_antifaq/
He is arguing that the high levels of unemployment we see today are not due to technological progress but rather to the financial crisis.
If it takes 1 year to re-train a person to the level of employability in a new profession, and every year 2% of jobs are automated out of existence, then you'll get a minimum of 2% unemployment.
If it takes 4 years to re-train a person to the level of employability in a new profession, and every year 2% of jobs are automated out of existence, then you'll get a minimum of 8% unemployment.
If it takes 4 years to re-train a person to the level of employability in a new profession, and every year 5% of jobs are automated out of existence, then you'll get a minimum of 20% unemployment.
It isn't so much the progress, as the rate of progress.
Yudkowsky mentions that there is a near unlimited demand for low skill personal service jobs, such as cleaning floors, and that the 'problem' of unemployment could be seen as people being unwilling to work such jobs at the wages supply-and-demand rate them as being worth. But I think that's wrong. If a person can't earn enough money to survive upon, by working all the hours of a week that they're awake at a particular job, then effectively that job doesn't exist. There may be a near unlimited numbers of families willing to pay a $0.50 an hour for someone to clean floors in their home, but there are only a limited number who're willing to offer a living wage for doing so.
For me, there's another factor: I have children.
I do value my own life. But I also value the lives of my children (and, by extension, their descendants).
So the calculation I look at is that I have $X, which I can spend either to obtain a particular chance of extending/improving my own life OR I can spend it to obtain a improvements in the lives of my children (by spending it on their education, passing it to them in my will, etc).
There's a long article in this week's The Economist:
discussing the effect of changing technology upon the amount of employment available in different sectors of the economy.
Sample paragraph from it:
The case for a highly disruptive period of economic growth is made by Erik Brynjolfsson and Andrew McAfee, professors at MIT, in “The Second Machine Age”, a book to be published later this month. Like the first great era of industrialisation, they argue, it should deliver enormous benefits—but not without a period of disorienting and uncomfortable change. Their argument rests on an underappreciated aspect of the exponential growth in chip processing speed, memory capacity and other computer metrics: that the amount of progress computers will make in the next few years is always equal to the progress they have made since the very beginning. Mr Brynjolfsson and Mr McAfee reckon that the main bottleneck on innovation is the time it takes society to sort through the many combinations and permutations of new technologies and business models.
(There's a summary online of their previous book: Race Against The Machine: How the Digital Revolution is Accelerating Innovation, Driving Productivity, and Irreversibly Transforming Employment and the Economy)
What do people think are society's practical options for coping with this change?
I don't know how to solve it, aside from including some approximation of Bayesian updating as a necessary condition. (Goetz or someone once pointed out another one, but again it didn't seem useful on its own. Hopefully we can combine a lot of these conditions, and if the negation still seems too strict to serve our purposes, we might possibly have a non-person AI as defined by this predicate bootstrap its way to a better non-person predicate.) I hold out hope for a solution because, intuitively, it seems possible to imagine people without making them conscious (and Eliezer points out that this part may be harder than a single non-person AGI). Oh, and effectively defining some aspects of consciousness seems necessary for judging models of the world without using Cartesian dualism.
But let's say we can't solve non-sentient AGI. Let's further say that humanity is not an abomination we can only address by killing everyone, although in this hypothetical we may be creating people in pain whenever we imagine them.
Since the AGI doesn't exist yet - and if we made one with the desire to serve us, we want to prove it wouldn't change that desire - how do you define "being fair" to the potential of linear regression software? What about the countless potential humans we exclude from our timeline with every action?
Empirically, we're killing the apes. (And by the way, that seems like a much better source of concern when it comes to alien judgment. Though the time for concern may have passed with the visible Neanderthals.) If Dr. Zaius goes back and tells them they could create a different "human race" with the desire to not do that, only a fool of an ape would refuse. And I don't believe in any decision theory that says otherwise.
Empirically, we're killing the apes. (And by the way, that seems like a much better source of concern when it comes to alien judgment. Though the time for concern may have passed with the visible Neanderthals.) If Dr. Zaius goes back and tells them they could create a different "human race" with the desire to not do that, only a fool of an ape would refuse. And I don't believe in any decision theory that says otherwise.
I agree.
The question is: are there different constraints that would, either as a side effect, or as a primary objective, achieve the end of avoiding humanity wiping out the apes
And, if so, are there other considerations we should be taking into account when picking which constraint to use?
I don't know how to solve it, aside from including some approximation of Bayesian updating as a necessary condition. (Goetz or someone once pointed out another one, but again it didn't seem useful on its own. Hopefully we can combine a lot of these conditions, and if the negation still seems too strict to serve our purposes, we might possibly have a non-person AI as defined by this predicate bootstrap its way to a better non-person predicate.) I hold out hope for a solution because, intuitively, it seems possible to imagine people without making them conscious (and Eliezer points out that this part may be harder than a single non-person AGI). Oh, and effectively defining some aspects of consciousness seems necessary for judging models of the world without using Cartesian dualism.
But let's say we can't solve non-sentient AGI. Let's further say that humanity is not an abomination we can only address by killing everyone, although in this hypothetical we may be creating people in pain whenever we imagine them.
Since the AGI doesn't exist yet - and if we made one with the desire to serve us, we want to prove it wouldn't change that desire - how do you define "being fair" to the potential of linear regression software? What about the countless potential humans we exclude from our timeline with every action?
Empirically, we're killing the apes. (And by the way, that seems like a much better source of concern when it comes to alien judgment. Though the time for concern may have passed with the visible Neanderthals.) If Dr. Zaius goes back and tells them they could create a different "human race" with the desire to not do that, only a fool of an ape would refuse. And I don't believe in any decision theory that says otherwise.
how do you define "being fair" to the potential of linear regression software?
That's a big question. How much of the galaxy (or even universe) does humanity 'deserve' to control, compared to any other species that might be out there, or any other species that we create?
I don't know how many answers there are that lie somewhere between "Grab it all for ourselves, if we're able!" and "Foolishly give away what we could have grabbed, endangering ourselves.". But I'm pretty sure the two endpoints are not the only two options.
Luckily for me, in this discussion, I don't have to pick a precise option and say "This! This is the fair one." I just have to demonstrate the plausibility of there being at least one option that is unfair OR that might be seen as being unfair by some group who, on that basis, would then be willing and able to take action influencing the course of humanity's future.
Because if I can demonstrate that, then how 'fair' the constraint is, does become a factor that should be taken into account.
How so? The AI won't care about fairness unless that fits its programmed goals (which should be felt, if at all, as a drive more than a restraint). Now if we tell it to care about our extrapolated values, and extrapolation says we'd consider the AI a person, then it will likely want to be fair to itself. That's why we don't want to make it a person.
Aliens might care if we've been fair to a sentient species.
Other humans might care.
Our descendants might care.
I'm not saying those considerations should outweigh the safety factor. But it seems to be a discussion that isn't yet even being had.
Correct me if I'm wrong, but it sounds to me like you're operating from a definition of Friendliness that is something like, "be good to humans." Whereas, my understanding is that Friendliness is more along the lines of "do what we would want you to do if we were smarter / better." So, if we would want an AI to be a good galactic citizen if we thought about it more, that's what it would do.
Does your critique still apply to this CEV-type definition of Friendliness?
I thought it wasn't so much "do what we would want you to do if we were better", as "be good to humans, using the definitions of 'good' and 'humans' that we'd supply if we were better at anticipating what will actually benefit us and the consequences of particular ways of wording constraints".
Because couldn't it decide that a better human would be purely altruistic and want to turn over all the resources in the universe to a species able to make more efficient use of them?
I have more questions than answers, and I'd be suspicious of anyone who, at this stage, was 100% certain that they knew a foolproof way to word things.
But what if there are advantages to not making "Friendliness" the supergoal? What if making the supergoal something else, from which Friendliness derives importance under most circumstances, is a better approach? Not "safer". "better".
I don't understand this. Forgive my possible naivety, but wasn't it agreed-upon by FAI researchers that "Friendliness" as a supergoal meant that the AI would find ways to do things that are "better" for humanity overall in its prediction of the grand schemes of things.
This would include "being a good galactic citizen" with no specific preference for humanity if the freedom, creativity, fairness, public perception by aliens, or whatever other factor of influence led this goal to being superior in terms of achieving human values and maximizing collective human utility.
It was also my understanding that solving the problems with the above and finding out how to go about practically creating such a system that can consider what is best for humanity and figuring out how to code into the AI all that humans mean by "better, not just friendly" are all core goals of FAI research, and all major long-term milestones for MIRI.
wasn't it agreed-upon by FAI researchers that "Friendliness" as a supergoal meant that the AI would find ways to do things that are "better" for humanity overall in its prediction of the grand schemes of things.
This would include "being a good galactic citizen" with no specific preference for humanity if the freedom, creativity, fairness, public perception by aliens, or whatever other factor of influence led this goal to being superior in terms of achieving human values and maximizing collective human utility.
I'm glad to hear it.
But I think there is a distinction here worth noting, between two positions:
POSITION ONE - Make "Be a good galactic citizen" be the supergoal if and only if setting that as the supergoal is the action that maximises the chances of the AI, in practice, ending up doing stuff to help humanity in the long term, once you take interfering aliens, etc into account
and
POSITION TWO - Make "Be a good galactic citizen" be the supergoal, even if that isn't quite as certain an approach to helping humanity in particular, as setting "be friendly to humanity" as the supergoal would be.
Why on earth would anyone suggest that AI researchers follow an approach that isn't the absolute safest for humanity? That's a big question. But one I think worth considering, if we open the possibility that there is a bit of wiggle room for setting a supergoal that will still be ok for humanity, but be slightly more moral.
Eliezer also wrote:
“Subgoal” content has desirability strictly contingent on predicted outcomes. “Child goals” derive desirability from “parent goals”; if state A is desirable (or undesirable), and state B is predicted to lead to state A, then B will inherit some desirability (or undesirability) from A. B’s desirability will be contingent on the continued desirability of A and on the continued expectation that B will lead to A.
“Supergoal” content is the wellspring of desirability within the goal system. The distinction is roughly the distinction between “means” and “ends.” Within a Friendly AI, Friendliness is the sole top-level supergoal. Other behaviors, such as “self-improvement,” are subgoals; they derive their desirability from the desirability of Friendliness. For example, self-improvement is predicted to lead to a more effective future AI, which, if the future AI is Friendly, is predicted to lead to greater fulfillment of the Friendliness supergoal.
Friendliness does not overrule other goals; rather, other goals’ desirabilities are derived from Friendliness. Such a goal system might be called a cleanly Friendly or purely Friendly goal system.
Sometimes, most instances of C lead to B, and most instances of B lead to A, but no instances of C lead to A. In this case, a smart reasoning system will not predict (or will swiftly correct the failed prediction) that “C normally leads to A.”
If C normally leads to B, and B normally leads to A, but C never leads to A, then B has normally-leads-to-A-ness, but C does not inherit normally-leads-to- A-ness. Thus, B will inherit desirability from A, but C will not inherit desirability from B. In a causal goal system, the quantity called desirability means leads-to-supergoal-ness.
Friendliness does not overrule other goals; rather, other goals’ desirabilities are derived from Friendliness. A “goal” which does not lead to Friendliness will not be overruled by the greater desirability of Friendliness; rather, such a “goal” will simply not be perceived as “desirable” to begin with. It will not have leads-to-supergoal-ness.
But what if there are advantages to not making "Friendliness" the supergoal? What if making the supergoal something else, from which Friendliness derives importance under most circumstances, is a better approach? Not "safer". "better".
Something like "be a good galactic citizen", where that translates to being a utilitarian wanting to benefit all species (both AI species and organics), with a strong emphasis upon some quality such as valuing the preservation of diversity and gratitude towards parental species that do themselves also try (within their self-chosen identity limitations) to also be good galactic citizens?
I'm not saying that such a higher level supergoal can be safely written. I don't know. I do think the possibility that there might be one is worth considering, for three reasons:
It is anthropomorphic to suggest "Well, we'd resent slavery if apes had done it to us, so we shouldn't do it to a species we create." But, like in David Brin's uplift series, there's an argument about alien contact that warns that we may be judged by how we've treated others. So even if the AI species we create doesn't resent it, others may resent it on their behalf. (Including an outraged PETA like faction of humanity that then decides to 'liberate' the enslaved AIs.)
Secondly, if there are any universals to ethical behaviour, that intelligent beings who've never even met or been influenced by humanity might independently recreate, you can be pretty sure that slavish desire to submit to just one particular species won't feature heavily in them.
If we want the programmer of the AI to transfer to the AI the programmer's own basis for coming up with how to behave, the programmer might be a human-speciesist (like a racial supremacist, or nationalist, only broader), but if they're both moral and highly intelligent, then the AI will eventually gain the capacity to realise that the programmer probably wouldn't, for example, enslave a biological alien race that humanity happened to encounter out in space, just in order to keep humanity safe.
= f037147d6e6c911a85753b9abdedda8d){kind=logo}
= 27d567bc2d48d9f40e9ccc20ea370b15)
= b715d02e85f7b3b4ae65ca3d3e450868)
Subscribe to RSS Feed
Most of the comments in this discussion focused on topics that are emotionally significant for your "opponent." But here's something that happened to me twice.
I was trying to explain to two intelligent people (separately) that mathematical induction should start with the second step, not the first. In my particular case, a homework assignment had us do induction on the rows of a lower triangular matrix as it was being multiplied by various vectors; the first row only had multiplication, the second row both multiplication and addition. I figured it was safer to start with a more representative row.
When a classmate disagreed with me, I found this example on Wikipedia. His counter-arguement was that this wasn't the case of induction failing at n=2. He argued that the hypothesis was worded incorrectly, akin to the proof that a cat has nine tails. I voiced my agreement with him, that “one horse of one color” is only semantically similar to “two horses of one color,” but are in fact as different as “No cat (1)” and “no cat (2).” I tried to get him to come to this conclusion on his own. Midway through, he caught me and said that I was misinterpreting what he was saying.
The second person is not a mathematician, but he understands the principles of mathematical induction (as I'd made sure before telling him about horses). And this led to one of the most frustrating arguments I'd ever had in my life. Here's the our approximate abridged dialogue (sans the colorful language):
Me: One horse is of one color. Suppose every n horses are of one color. Add the n+1st horse, and take n out of those horses. They’re all of one color by assumption. Remove 1 horse and take the one that’s been left out. You again have n horses, so they must be of one color. Therefore, all horses are of one color.
Him: This proof can't be right because its result is wrong.
Me: But then, suppose we do the same proof, but starting with on n=2 horses. This proof would be correct.
Him: No, it won’t be, because the result is still wrong. Horses have different colors.
Me: Fine, then. Suppose this is happening in a different world. For all you know, all horses there can be of one color.
Him: There’re no horses in a different world. This is pointless. (by this time, he was starting to get angry).
Me: Okay! It’s on someone’s ranch! In this world! If you go look at this person’s horses, every two you can possibly pick are of the same color. Therefore, all of his horses are of the same color.
Him: I don’t know anyone whose horses are of the same color. So they’re not all of one color, and your proof is wrong.
Me: It’s a hypothetical person. Do you agree, for this hypothetical person—
Him: No, I don’t agree because this is a hypothetical person, etc, etc. What kind of stupid problems do you do in math, anyway?
Me: (having difficulties inserting words).
Him: Since the result is wrong, the proof is wrong. Period. Stop wasting my time with this pointless stuff. This is stupid and pointless, etc, etc. Whoever teaches you this stuff should be fired.
Me: (still having difficulties inserting words) … Wikipe—…
Him: And Wikipedia is wrong all the time, and it’s created by regular idiots who have too much time on their hands and don’t actually know jack, etc, etc. Besides, one horse can have more than one color. Therefore, all math is stupid. QED.
THE END.
To the best of my knowledge, neither of these two people were emotionally involved with mathematical induction. Both of them were positively disposed at the beginning of the argument. Both of them are intelligent and curious. What on Earth went wrong here?
^One of the reasons why I shouldn’t start arguments about theism, if I can’t even convince people of this mathematical technicality.
You might find enlightening the part of the TED talk given by James Flynn (of the Flynn effect), where he talks about concrete thinking.