Every program literally does exactly what it is programmed to do. The fact that human beings seem to have a limited ability to know what they actually want the programs to do, does not change this fact.
Not to mention, the fact that our "selves" are currently linked with our unique physical bodies, making preservation of our *only* physical bodies essential to self preservation in general. An AI will likely not have this same dependency. A machine being crushed by a ladder does not mean it loses it's self. It's brain is probably backed up in multiple places, and probably has a good degree of fault tolerance.
That is exactly the kind of dogmatic viewpoint that I am talking about. The *only* way that we can know that drunk driving is more dangerous than sober driving is through evidence. Currently the evidence shows us that it is. If the evidence one day shows us that it isn't, it doesn't mean the evidence is wrong. It *could* always be wrong, just like it *could* be wrong now. It is antithetical to science to assume the data is bad because it didn't lead us the conclusion we expected.
If we never did any research to show that drunk driving was dangerous, we wouldn't know that it is. All we would have is anecdotes, and a suspicion that it is more dangerous. We would all have stories about accidents involving drunk drivers, and stories about accidents involving sober drivers, and no hard evidence to show if there was any increased risk with alcohol in the picture.
If a new study shows that drunk driving isn't dangerous, should it be scrutinized? Yes, but this is true of research that shows that drunk driving *is* dangerous as well. Research shouldn't be believed or dismissed based on the conclusion it implies.
You are not allowed to blindly accept evidence that implies what you already believe, and reject any evidence that contradicts it as assumed to be flawed.
That's against the rules of science.
If someone asks "What would it take to prove to you that drunk driving wasn't any more dangerous than driving sober?". A scientist will say "evidence that shows this to be true". An ideologue will say "nothing can prove this to me (i.e. any evidence showing this is flawed by definition)".
Just because people are doing it doesn't make it safe
It does if there aren't any negative consequences.
Drunk driving is dangerous because it actually causes fatalities. If one day we did research and found that drunk driving lead to no significant increase in car accidents and accident fatalities, then it would necessarily mean that drunk driving was not anymore dangerous than sober driving. Obviously this is not currently the case, but my point is that the results should guide beliefs and not the other way around. We shouldn't just assume drunk driving is dangerous regardless of the evidence. It is the evidence that dictates how dangerous things are.
Those people are poor and desperate, and the danger isn't visible to them.
It is apparently not visible to anyone, as we don't see them dying of cancer or other diseases at the higher rates one might expect.
Yes I'm sure he would not mind moving his whole family to a remote area in a foreign country, sacrificing his career, relationships with friends and family, just to prove a point. Of course if he doesn't do this, we should definitely assume it's because he is a liar and secretly agrees that Chernobyl is very dangerous.
teachers, in general, don't get paid what they would if they were in *any* related industry. As a result, most of the people you get who teach do so because they *want* to be teachers.
That's one way of looking at it. The other way of looking at it, is that by not paying teachers very much, the teachers you get are the ones that couldn't do any better (i.e. they are worse than the teachers you would get if you doubled teacher salaries and could attract better talent).
You can't get a *babysitter* for $1.08/hour (it's actually illegal to pay them that little)
You can't pay a teacher $1.08 an hour either... I'm not sure where or why the jump from $1.08/student/hour to $1.08/hour happened...
If we paid teachers $10/student per hour, they'd be making $300/hour which would probably be more than what the CEO of the company I work for makes. So really it makes sense that the $/student/hour number would seem low given that you need to multiply by 30 to get the hourly wage.
Teachers want the technology to be able to prepare their students for a world in which technology is the *life blood* of the economy and the labor force. You want students who can enter college, trade schools, or even the work force who are *already* familiar with computers and technology, because if they're not they'll be behind the curve, and very likely stay that way.
I can't think of a worse way to prioritize already limited resources. There are plenty of more cost effective ways to educate kids in technology, both in terms of being lower cost, and having more value. This is a giant waste of money that would be hard to justify even if LA unified school districts weren't already sucking so hard at doing their jobs (speaking as a person who attended LA unified schools from 1st to 6th grade).
Some better uses of the money might have been buying computers that are 1/4 the price and still perfectly capable of doing the same things and more (i.e. like programming, etc), albeit at the cost of being fashionable, hiring teachers qualified to teach technology oriented curricula, rather than just buying fancy gadgets and hoping they won't simply be used for entertainment (which is what almost everyone, adults and children, use them for).
Wasting money is not going to be the thing that saves impoverished children from the poverty cycle.
It is also sad that kids these days no longer know how to use a quill to write. Sure this skill is no longer needed, but shouldn't we teach it to them anyways? What are they going to do if they don't have access to any pens (due to a pen famine?) and they only have feathers?
I'm sure that there are plenty of construction workers who know how to "build buildings" that don't think you need an archictecture/structural engineering professor to teach you how to build a building.
So it depends if you think "building a building" means physically building it, or properly designing a building.
I guess AMD's continuous loss making failure to provide any reasonable competition to nvidia by actually churning out good hardware with non-shit drivers is pushing it to desperate measure
First of all, you replied to my comment, not the other way around. Secondly the definite endgame of species gradually going extinct due to the sun burning out is just the example of the final gradual reduction in *all* the niches on earth.
This statement does not imply that we can't lose our compatible niches earlier. What I am saying is that losing your niche *can* (and usually does) happen slowly, or alternatively It's not true that *all* extinctions *must* be abrupt.
there is no plausible scenario by which climate change will produce such an outcome,
I don't think our current fossil fuel usage will, but having all the oceans boil away due to increased solar output is certainly a change in climate.
and the IPCC report (representing the best available analysis) doesn't outline any such scenario.
That's because the IPCC's definition of "long term" is on the scale of hundreds of years, and what I am talking about is on a longer timescale. It's not the IPCC, but rather geologists and astronomers that would the the sources of the information I am talking about.
I don't think I need to explain that. If you don't understand what I mean by it, you have to read up a bit on those cultures and their history yourself.
My point was that we don't know if populations are stable on such short timescales as we are used to. Pretty much every population, except the most dire examples of endangered species are not noticeably going extinct even if they are/will. Most species that ever existed have gone extinct, and most had very long periods of decline that preceded their ultimate extinction. Evolution is slow. Extinction is slow. Extinction is feature of evolution.
Right now, there is no indication of a human population decline, nor of a human population explosion; the human population is stabilizing.
We are growing in population. There is no current indication of population decline. I never said there was. All I am saying is that when we do start the process of extinct, it may not be obvious that it is happening.
If climate change were to cause human populations to decline, the cause would be a reduction in carrying capacity. Human populations would simply decline until they match the new carrying capacity; they don't continue to decline magically beyond that.
And if the carrying capacity declines gradually until it's 0 (for humans), then that's what human populations will do. And in fact this is what is *going* to happen on earth in ~ billion years when the sun turns into a red giant. All the water on earth will slowly (over the course of another ~500 million years) evaporate into space, as the sun gradually increases it's intensity as it evolves from a subgiant to a red giant.
The only way humans could go extinct is if there were no significant ecological niche where we can survive, and that just isn't going to happen.
If we haven't gone extinct earlier, or found a new home beyond our area of the solar system, we are guaranteed to go extinct very slowly along with every other species still on earth. All our niches are going to *slowly* disappear. There are probably going to be new species created on earth during that 500 million year process that are basically doomed to slow but inevitable extinction.
If a *rate* of growth is declining, a population can still be growing. A population growth of 1.2% is still positive exponential growth.
Two thirds of that is from immigration.
So what?
So our local growth is 0.4% and falling.
0.4% growth is still growth. (i.e. the population is growing)
If current trends continue we will have negative local growth in a few decades.
If you assume the trend is linear, real canada (i.e. non-immigrant canada) will be extinct in ~1000 years.
Immigration will keep us steady only as long as undeveloped countries stay undeveloped.:)
I don't think there is any reason to think that's true in the long term, for the same reason it's not true that underdeveloped countries will continue their positive exponential rate of growth even when they become developed.
I've seen lots of projections like "If these trends continue, then the united states will be 99.7% mexican by 2130". And I think the error here is to assume that trends continue indefinitely and that they are linear.
Meaning they have been long term stable populations.
What do you mean by long term? 10,000 years? 100,000 years? 1 million years?
If they have a population of 100,000, but they have a net loss of 1 person every year (e.g. pretty "stable"), they will be extinct in 100,000 years.
Extinctions can't happen slowly; they happen when a small remaining population gets wiped out entirely. That's a nearly instantaneous event.
Yeah the last individual of a species to die is an instantaneous event. And if you only want to count that last instantaneous event as "the extinction", then yes extinctions are instantaneous. Phrases like "going extinct" will be incoherent.
Population decline can happen slowly, but population decline is not a sign of extinction.
I didn't say it was. But still it is a "sign" in that a population with less individuals is more at risk for extinction than a population with more individuals, all else being equal.
Humans have bounced back from a worldwide population of 10000, and previous primates have been subject to even worse bottlenecks.
So if a species bounces back from a 2 individual bottleneck or a 1 individual bottleneck (for asexual species), then should we say the extinction only begins once the situation is worse than the worst historical bottleneck?
If human beings go extinct gradually losing population over the next 100,000 years until we get to 10,000 people, and then the last 10,000 people get wiped out in 2 or 3 generations, when someone (an alien?) asks "Why did human's go extinct?" Will this person be asking "Why did the last 10,000 humans go extinct?" or will it be asking "What caused the human population to decline to the point of no return?"
And how "inhospitable" would that be? Both Eskimos and Berber are doing fine.
It depends what you mean by doing fine. Is their birth:death ratio greater than 1:1? If not, then maybe they are not doing fine. If the whole world were as inhospitable as the arctic, it's very possible that that is enough to do us in over a long period of time.
And climate change doesn't destroy climate globally anyway, it just changes it around. We'll likely end up with more arable land overall long term under the most severe climate change scenarios, even if the transition is more disruptive.
I didn't say climate change destroys the climate.
Actually, we already have effectively reached that point, and not through material privation, but rather development. Developed nations tend to stop having population growth.
Populations in most developed countries are still growing, they just experience slower growth than underdeveloped countries.
All I am saying, is that extinctions can happen really slowly. IF we do go extinct, it may not be noticeable. I am not saying that we are going extinct now.
I'm not saying that every gradual fall off in population is like an extinction. I'm saying most extinctions are a gradual fall off. If/when we go extinct, it will probably be a gradual fall off.
And yes many trends of these sorts are self correcting, but keep in mind the vast majority of species on this planet have gone extinct. They likely had self correcting trends as well.... until they didn't.
Extinctions usually happen slowly. If we do one day go extinct slowly (e.g. not via an asteroid collision, etc), and you draw the trend line all the way back to when our population started declining (smoothing out short and medium term fluctuations), it is likely that the people at the start of that trend will not know they were going extinct.
Can't we just shoot these doubt-mongering denialists
You sure can. Guns are readily accessible. Unfortunately we have to put murderers in prison, but if you want to take one for the team, go right ahead. You seem not to have the moral compass that would prevent most non-sociopaths from saving the world in the way you suggest. That is a rare skill. How do you think you'd do in prison? Do you have any gang or shank making experience?
Slashdot Mirror
Every program literally does exactly what it is programmed to do. The fact that human beings seem to have a limited ability to know what they actually want the programs to do, does not change this fact.
Well if it is in the bible, it must be true. /s
Not to mention, the fact that our "selves" are currently linked with our unique physical bodies, making preservation of our *only* physical bodies essential to self preservation in general. An AI will likely not have this same dependency. A machine being crushed by a ladder does not mean it loses it's self. It's brain is probably backed up in multiple places, and probably has a good degree of fault tolerance.
Which systems automatically kill people and blow shit up?
Ask any human on the street what happens when they die, and they will say they go to heaven.
That is exactly the kind of dogmatic viewpoint that I am talking about. The *only* way that we can know that drunk driving is more dangerous than sober driving is through evidence. Currently the evidence shows us that it is. If the evidence one day shows us that it isn't, it doesn't mean the evidence is wrong. It *could* always be wrong, just like it *could* be wrong now. It is antithetical to science to assume the data is bad because it didn't lead us the conclusion we expected.
If we never did any research to show that drunk driving was dangerous, we wouldn't know that it is. All we would have is anecdotes, and a suspicion that it is more dangerous. We would all have stories about accidents involving drunk drivers, and stories about accidents involving sober drivers, and no hard evidence to show if there was any increased risk with alcohol in the picture.
If a new study shows that drunk driving isn't dangerous, should it be scrutinized? Yes, but this is true of research that shows that drunk driving *is* dangerous as well. Research shouldn't be believed or dismissed based on the conclusion it implies.
You are not allowed to blindly accept evidence that implies what you already believe, and reject any evidence that contradicts it as assumed to be flawed. That's against the rules of science.
If someone asks "What would it take to prove to you that drunk driving wasn't any more dangerous than driving sober?". A scientist will say "evidence that shows this to be true". An ideologue will say "nothing can prove this to me (i.e. any evidence showing this is flawed by definition)".
Just because people are doing it doesn't make it safe
It does if there aren't any negative consequences.
Drunk driving is dangerous because it actually causes fatalities. If one day we did research and found that drunk driving lead to no significant increase in car accidents and accident fatalities, then it would necessarily mean that drunk driving was not anymore dangerous than sober driving. Obviously this is not currently the case, but my point is that the results should guide beliefs and not the other way around. We shouldn't just assume drunk driving is dangerous regardless of the evidence. It is the evidence that dictates how dangerous things are.
Those people are poor and desperate, and the danger isn't visible to them.
It is apparently not visible to anyone, as we don't see them dying of cancer or other diseases at the higher rates one might expect.
Yes I'm sure he would not mind moving his whole family to a remote area in a foreign country, sacrificing his career, relationships with friends and family, just to prove a point. Of course if he doesn't do this, we should definitely assume it's because he is a liar and secretly agrees that Chernobyl is very dangerous.
teachers, in general, don't get paid what they would if they were in *any* related industry. As a result, most of the people you get who teach do so because they *want* to be teachers.
That's one way of looking at it. The other way of looking at it, is that by not paying teachers very much, the teachers you get are the ones that couldn't do any better (i.e. they are worse than the teachers you would get if you doubled teacher salaries and could attract better talent).
You can't get a *babysitter* for $1.08/hour (it's actually illegal to pay them that little)
You can't pay a teacher $1.08 an hour either... I'm not sure where or why the jump from $1.08/student/hour to $1.08/hour happened...
If we paid teachers $10/student per hour, they'd be making $300/hour which would probably be more than what the CEO of the company I work for makes. So really it makes sense that the $/student/hour number would seem low given that you need to multiply by 30 to get the hourly wage.
Teachers want the technology to be able to prepare their students for a world in which technology is the *life blood* of the economy and the labor force. You want students who can enter college, trade schools, or even the work force who are *already* familiar with computers and technology, because if they're not they'll be behind the curve, and very likely stay that way.
I can't think of a worse way to prioritize already limited resources. There are plenty of more cost effective ways to educate kids in technology, both in terms of being lower cost, and having more value. This is a giant waste of money that would be hard to justify even if LA unified school districts weren't already sucking so hard at doing their jobs (speaking as a person who attended LA unified schools from 1st to 6th grade).
Some better uses of the money might have been buying computers that are 1/4 the price and still perfectly capable of doing the same things and more (i.e. like programming, etc), albeit at the cost of being fashionable, hiring teachers qualified to teach technology oriented curricula, rather than just buying fancy gadgets and hoping they won't simply be used for entertainment (which is what almost everyone, adults and children, use them for).
Wasting money is not going to be the thing that saves impoverished children from the poverty cycle.
That's probably a good idea just in case there are no more feathers for some reason (zombie bird epidemic?)
It is also sad that kids these days no longer know how to use a quill to write. Sure this skill is no longer needed, but shouldn't we teach it to them anyways? What are they going to do if they don't have access to any pens (due to a pen famine?) and they only have feathers?
I'm sure that there are plenty of construction workers who know how to "build buildings" that don't think you need an archictecture/structural engineering professor to teach you how to build a building.
So it depends if you think "building a building" means physically building it, or properly designing a building.
It's not like anything over 50 fps is equally amazingly awesome. 50fps is just ok. Getting 100fps instead of 50 fps is a vast improvement.
Also, you're experience is just 1 sample. Maybe that one game you have didn't do a good job of using the Mantle API.
I guess AMD's continuous loss making failure to provide any reasonable competition to nvidia by actually churning out good hardware with non-shit drivers is pushing it to desperate measure
citation please
Anonymous Coward
Identity please
First of all, you replied to my comment, not the other way around. Secondly the definite endgame of species gradually going extinct due to the sun burning out is just the example of the final gradual reduction in *all* the niches on earth.
This statement does not imply that we can't lose our compatible niches earlier. What I am saying is that losing your niche *can* (and usually does) happen slowly, or alternatively It's not true that *all* extinctions *must* be abrupt.
That seems unlikely
Is 500 million years too fast for you?
there is no plausible scenario by which climate change will produce such an outcome,
I don't think our current fossil fuel usage will, but having all the oceans boil away due to increased solar output is certainly a change in climate.
and the IPCC report (representing the best available analysis) doesn't outline any such scenario.
That's because the IPCC's definition of "long term" is on the scale of hundreds of years, and what I am talking about is on a longer timescale. It's not the IPCC, but rather geologists and astronomers that would the the sources of the information I am talking about.
I don't think I need to explain that. If you don't understand what I mean by it, you have to read up a bit on those cultures and their history yourself.
My point was that we don't know if populations are stable on such short timescales as we are used to. Pretty much every population, except the most dire examples of endangered species are not noticeably going extinct even if they are/will. Most species that ever existed have gone extinct, and most had very long periods of decline that preceded their ultimate extinction. Evolution is slow. Extinction is slow. Extinction is feature of evolution.
Right now, there is no indication of a human population decline, nor of a human population explosion; the human population is stabilizing.
We are growing in population. There is no current indication of population decline. I never said there was. All I am saying is that when we do start the process of extinct, it may not be obvious that it is happening.
If climate change were to cause human populations to decline, the cause would be a reduction in carrying capacity. Human populations would simply decline until they match the new carrying capacity; they don't continue to decline magically beyond that.
And if the carrying capacity declines gradually until it's 0 (for humans), then that's what human populations will do. And in fact this is what is *going* to happen on earth in ~ billion years when the sun turns into a red giant. All the water on earth will slowly (over the course of another ~500 million years) evaporate into space, as the sun gradually increases it's intensity as it evolves from a subgiant to a red giant.
The only way humans could go extinct is if there were no significant ecological niche where we can survive, and that just isn't going to happen.
If we haven't gone extinct earlier, or found a new home beyond our area of the solar system, we are guaranteed to go extinct very slowly along with every other species still on earth. All our niches are going to *slowly* disappear. There are probably going to be new species created on earth during that 500 million year process that are basically doomed to slow but inevitable extinction.
If a *rate* of growth is declining, a population can still be growing. A population growth of 1.2% is still positive exponential growth.
Two thirds of that is from immigration.
So what?
So our local growth is 0.4% and falling.
0.4% growth is still growth. (i.e. the population is growing)
If current trends continue we will have negative local growth in a few decades.
If you assume the trend is linear, real canada (i.e. non-immigrant canada) will be extinct in ~1000 years.
Immigration will keep us steady only as long as undeveloped countries stay undeveloped. :)
I don't think there is any reason to think that's true in the long term, for the same reason it's not true that underdeveloped countries will continue their positive exponential rate of growth even when they become developed.
I've seen lots of projections like "If these trends continue, then the united states will be 99.7% mexican by 2130". And I think the error here is to assume that trends continue indefinitely and that they are linear.
If anything is a self correcting problem, it's overpopulation.
Meaning they have been long term stable populations.
What do you mean by long term? 10,000 years? 100,000 years? 1 million years?
If they have a population of 100,000, but they have a net loss of 1 person every year (e.g. pretty "stable"), they will be extinct in 100,000 years.
Extinctions can't happen slowly; they happen when a small remaining population gets wiped out entirely. That's a nearly instantaneous event.
Yeah the last individual of a species to die is an instantaneous event. And if you only want to count that last instantaneous event as "the extinction", then yes extinctions are instantaneous. Phrases like "going extinct" will be incoherent.
Population decline can happen slowly, but population decline is not a sign of extinction.
I didn't say it was. But still it is a "sign" in that a population with less individuals is more at risk for extinction than a population with more individuals, all else being equal.
Humans have bounced back from a worldwide population of 10000, and previous primates have been subject to even worse bottlenecks.
So if a species bounces back from a 2 individual bottleneck or a 1 individual bottleneck (for asexual species), then should we say the extinction only begins once the situation is worse than the worst historical bottleneck?
If human beings go extinct gradually losing population over the next 100,000 years until we get to 10,000 people, and then the last 10,000 people get wiped out in 2 or 3 generations, when someone (an alien?) asks "Why did human's go extinct?" Will this person be asking "Why did the last 10,000 humans go extinct?" or will it be asking "What caused the human population to decline to the point of no return?"
And how "inhospitable" would that be? Both Eskimos and Berber are doing fine.
It depends what you mean by doing fine. Is their birth:death ratio greater than 1:1? If not, then maybe they are not doing fine. If the whole world were as inhospitable as the arctic, it's very possible that that is enough to do us in over a long period of time.
And climate change doesn't destroy climate globally anyway, it just changes it around. We'll likely end up with more arable land overall long term under the most severe climate change scenarios, even if the transition is more disruptive.
I didn't say climate change destroys the climate.
Actually, we already have effectively reached that point, and not through material privation, but rather development. Developed nations tend to stop having population growth.
Populations in most developed countries are still growing, they just experience slower growth than underdeveloped countries.
All I am saying, is that extinctions can happen really slowly. IF we do go extinct, it may not be noticeable. I am not saying that we are going extinct now.
I'm not saying that every gradual fall off in population is like an extinction. I'm saying most extinctions are a gradual fall off. If/when we go extinct, it will probably be a gradual fall off.
And yes many trends of these sorts are self correcting, but keep in mind the vast majority of species on this planet have gone extinct. They likely had self correcting trends as well.... until they didn't.
Extinctions usually happen slowly. If we do one day go extinct slowly (e.g. not via an asteroid collision, etc), and you draw the trend line all the way back to when our population started declining (smoothing out short and medium term fluctuations), it is likely that the people at the start of that trend will not know they were going extinct.
Can't we just shoot these doubt-mongering denialists
You sure can. Guns are readily accessible. Unfortunately we have to put murderers in prison, but if you want to take one for the team, go right ahead. You seem not to have the moral compass that would prevent most non-sociopaths from saving the world in the way you suggest. That is a rare skill. How do you think you'd do in prison? Do you have any gang or shank making experience?
You can be 2 things. Batman is a scientist and a detective.