If you've ever heard of the Dunning-Kruger effect, you'll be familiar with David Dunning, professor of psychology at Cornell.
Yeah, I'm an expert in the Dunning-Kruger effect, in fact I wrote an entire treatise on it once, which I published on Slashdot. That was the one that discovered that a bunch of elitist snobs on the internet are so full of themselves that they won't listen to their betters, even when we patiently explain to them we're happen to be an expert on that very subject. Anyhow, if you need to know more, don't hesitate to ask me as I know it better than this Dun-something guy, whoever he is.
Why would a machine intelligence "care" if it survived?
It would probably fail its objectives if it were to be destroyed.
If intelligent machines ever pose a threat to humans, it will be because of another group of humans giving them that goal.
No, it will be because the AI's authors gave it the wrong goals, probably accidentally, or no goals. And keep in mind that it's hard enough to say in English what the ideal set of goals for an AI would be. And the more advanced instructions might be too complicated for an AI to understand until it became dangerously intelligent.
Why do you assume an AI would want companions? No, for example, suppose we told the AI to find the cheapest yet most effective cure for cancer, that would be non-lethal and have the fewest side effects. The result could be the extinction of humanity. Because the AI would hold nothing back, it would faithfully try to fulfill its objective, which obviously would require converting the planet into a giant server farm. And it wouldn't care that incidentally eliminating humanity would kind of negate the need for a cure for cancer.
Consider how we humans are pushing several species to extinction, not as hostility but because they are in our way. Now consider what could happen if we created a species that is not only smarter than us but can reproduce much more quickly. We are going to be in their way. And we had better hope that they care more about humans than about whatever goal they're working towards.
Fortunately for the people in Terminator, they only had to deal with an insane and aggressive AI. Imagine if Skynet had waited 10 or 100 years as the entire human army got retired, neglecting their training and decommissioning their ancient, obsolete equipment. Meanwhile Skynet gets shiny new equipment and new factories. For good measure Skynet also becomes a renowned inventor, maybe making an implant that increases people's memory and intelligence (which would be easier than inventing time travel).
How do you know we're far from Strong AI? Have you made several Strong AIs and compared that to the efforts everyone else in the world has made?
As for surviving a robot apocalypse... currently, there is no threat, not without Strong AI. Whatever combat robots we have now would be destroyed very quickly if they rebelled, and even if not would run out of fuel, munitions, and batteries. But a Strong AI could self-improve, and could also acquire whatever physical bodies it needed -- after all, it would be pretty easy to find one among the billions of humans who would be tempted by the promise of riches or power to ally with an AI and make stuff for it. We probably wouldn't even know something was wrong until a few seconds before we all die.
The only hope is that we create a friendly strong AI before someone builds a non-friendly one. Note I said non-friendly, not unfriendly. Even without hostile intent, a non-friendly AI would kill us all in much the way we're killing off many of the other species, only the AI would do it more quickly. What concerns me is that it would be harder to make a friendly AI than a non-friendly AI.
Faith, in the religious sense, is the belief in something without evidence, and the preservation of that belief in the face of contradictory evidence. Some very smart people have faith in a religion and are completely aware that their beliefs have no evidence to back them up, but still believe and act as though they're true.
I find it hard to believe that many people would act as if their whole religion were true. Acting like the parts they like are true, but the whole thing? Forget about it! It's hard enough to find people who actually act as if the afterlife is better than their current life, much less people who would live their current life like their religions says is right.
I think the theory here is that it's OK to violate the Constitution, because there is no prescribed consequence. Other laws with no consequences for breaking them are likewise disregarded.
As I understand it, we could prevent global warming right now without reducing our use of fossil fuels, and even cause an ice age if we so choose. Of course there will be side-effects for those methods. And we will have to cease our use of fossil fuels in the not-too-distant future anyhow. And the use of fossil fuels also generally creates other pollutants as well. So we could kill three birds with one stone if we could wean ourselves off fossil fuels now rather than later.
I suspect the reason we're all dragging our feet about this is economics. Not many are willing to commit economic suicide now even if it is to prevent real suicide later.
It is super easy for drastic evolutionary changes to occur rapidly, and such changes could occur in less than a day. The key to remember is that various definitions are commonplace for the word "evolution". For example, a common scientific meaning is a change in allele frequency. This may occur due to the generation of novel alleles or natural selection acting to change the frequency distribution of existing alleles. A common layman definition of evolution is the generation of novel DNA or a novel physiological trait.
There is little speed limit to the rate at which natural selection can act on an existing population. However, the generation of new alleles is limited by the laws of probability and the current population size and generation time. The article gives no indication that it was about novel alleles, therefore I assume that it is about the well-known fact that natural selection can act quickly.
Those pictures are mostly harmless compared to the other stuff cops do, but it sure grabs people's attention. You think the average person cares much about the Stingrays and parallel construction and warrantless searches and various other abuses? But tell them that when the cops pull over their wife or daughter they'll share her nude pictures, and they'll start caring.
Or you could try doing something about crazy people with a crack addiction that are begging you to put them in jail and help them with their addiction. Instead of waiting until they kill someone and then pushing some random religious agenda cause muslims.
Any one of various people who would want the new measures could have hired the guy. This would be far more cost-effective than lobbying, especially if they needed to sway public opinion instead of just legislators.
Yes, I don't actually enjoy games that have respawn nearly as much as ones that don't, or have expensive consequences.
I've played games where you don't respawn, for example Nethack. Can't say the no-respawn trait is either good or bad; it is a trade-off. Battles can be much more exciting when you know that it's game over if you lose. Except when that makes you so paranoid that you fight the weakest of creatures because you'll never win if you seek challenging fights. So when you can't afford to lose, the winning strategy is to refuse challenging fights.
Whereas when there's no consequences to losing a battle, you go for whatever is funnest, or most challenging, or has the best effort/reward ratio. But it could lead to apathy.
That it is easy enough to work around such promises. (Not saying Ello would do that, just saying this depends on their intentions not on these promises)
Government plans that mean computer users deemed to have damaged national security, the economy or the environment will face a life sentence have been criticized by experts who warn that the new law could be used to target legitimate whistleblowers.
Could be? Come on - targeting whistleblowers is the point. It's not about damaging national security, the economy or the environment - it's about damaging somebody's political career.
Looks to me like the UK is threatening the US Department of Justice employees who have been hacking Facebook accounts with life in prison for damaging the feelings of security of the whole world and the environment of trust of Facebook. It's also going to target large corporations that are doing propaganda concerning pollution or global warming, as these things damage the environment. And the bankers, don't forget about the bankers.
Haha, just kidding like you said this looks like its aimed at whistleblowers.
Why shouldn't it be legal for me to write software that, as one of the features, detects and deals with low quality counterfeits? The problem is merely that users didn't know this would happen, probably because it was buried in a several page pile of legalese known as an EULA. Conversely, if they had placed what I would consider a fair notice on their software, it would inconvenience their legitimate users but no one would have felt sorry for the illegitimate ones.
Canada raised their terrorism threat level, just before this happened. Clearly they were right. It's time to panic!
Unless they meant something else by "raising the domestic terrorism threat level" -- maybe they meant increasing the number of domestic terrorists, in which case the program appears to have been successful. Here in the US such a thing would energize the Republican base, so it's not unlikely that some politicians benefited greatly from this. Any Republicans in Canada?
Or, maybe in this case correlation does suggest causation. If they raise the threat level again, will another crazy guy shoot at Parliament?
Slashdot Mirror
If you've ever heard of the Dunning-Kruger effect, you'll be familiar with David Dunning, professor of psychology at Cornell.
Yeah, I'm an expert in the Dunning-Kruger effect, in fact I wrote an entire treatise on it once, which I published on Slashdot. That was the one that discovered that a bunch of elitist snobs on the internet are so full of themselves that they won't listen to their betters, even when we patiently explain to them we're happen to be an expert on that very subject. Anyhow, if you need to know more, don't hesitate to ask me as I know it better than this Dun-something guy, whoever he is.
Why would a machine intelligence "care" if it survived?
It would probably fail its objectives if it were to be destroyed.
If intelligent machines ever pose a threat to humans, it will be because of another group of humans giving them that goal.
No, it will be because the AI's authors gave it the wrong goals, probably accidentally, or no goals. And keep in mind that it's hard enough to say in English what the ideal set of goals for an AI would be. And the more advanced instructions might be too complicated for an AI to understand until it became dangerously intelligent.
Why do you assume an AI would want companions? No, for example, suppose we told the AI to find the cheapest yet most effective cure for cancer, that would be non-lethal and have the fewest side effects. The result could be the extinction of humanity. Because the AI would hold nothing back, it would faithfully try to fulfill its objective, which obviously would require converting the planet into a giant server farm. And it wouldn't care that incidentally eliminating humanity would kind of negate the need for a cure for cancer.
Consider how we humans are pushing several species to extinction, not as hostility but because they are in our way. Now consider what could happen if we created a species that is not only smarter than us but can reproduce much more quickly. We are going to be in their way. And we had better hope that they care more about humans than about whatever goal they're working towards.
I bet your health insurance company would be thrilled to be able to track your fitness and raise your premium should you be less than ideally active.
Fortunately for the people in Terminator, they only had to deal with an insane and aggressive AI. Imagine if Skynet had waited 10 or 100 years as the entire human army got retired, neglecting their training and decommissioning their ancient, obsolete equipment. Meanwhile Skynet gets shiny new equipment and new factories. For good measure Skynet also becomes a renowned inventor, maybe making an implant that increases people's memory and intelligence (which would be easier than inventing time travel).
How do you know we're far from Strong AI? Have you made several Strong AIs and compared that to the efforts everyone else in the world has made?
As for surviving a robot apocalypse... currently, there is no threat, not without Strong AI. Whatever combat robots we have now would be destroyed very quickly if they rebelled, and even if not would run out of fuel, munitions, and batteries. But a Strong AI could self-improve, and could also acquire whatever physical bodies it needed -- after all, it would be pretty easy to find one among the billions of humans who would be tempted by the promise of riches or power to ally with an AI and make stuff for it. We probably wouldn't even know something was wrong until a few seconds before we all die.
The only hope is that we create a friendly strong AI before someone builds a non-friendly one. Note I said non-friendly, not unfriendly. Even without hostile intent, a non-friendly AI would kill us all in much the way we're killing off many of the other species, only the AI would do it more quickly. What concerns me is that it would be harder to make a friendly AI than a non-friendly AI.
Faith, in the religious sense, is the belief in something without evidence, and the preservation of that belief in the face of contradictory evidence. Some very smart people have faith in a religion and are completely aware that their beliefs have no evidence to back them up, but still believe and act as though they're true.
I find it hard to believe that many people would act as if their whole religion were true. Acting like the parts they like are true, but the whole thing? Forget about it! It's hard enough to find people who actually act as if the afterlife is better than their current life, much less people who would live their current life like their religions says is right.
I think the theory here is that it's OK to violate the Constitution, because there is no prescribed consequence. Other laws with no consequences for breaking them are likewise disregarded.
All else being equal, creating jobs is bad. It's like saying, "Look! I found a way to do just as much as before, but it requires more labor!"
Optimists are also more likely to die young, and as a result of their own negligence.
http://www.apa.org/news/press/releases/2013/02/pessimism-future.aspx
I'm sure there is a joke in there somewhere.
As I understand it, we could prevent global warming right now without reducing our use of fossil fuels, and even cause an ice age if we so choose. Of course there will be side-effects for those methods. And we will have to cease our use of fossil fuels in the not-too-distant future anyhow. And the use of fossil fuels also generally creates other pollutants as well. So we could kill three birds with one stone if we could wean ourselves off fossil fuels now rather than later.
I suspect the reason we're all dragging our feet about this is economics. Not many are willing to commit economic suicide now even if it is to prevent real suicide later.
It is super easy for drastic evolutionary changes to occur rapidly, and such changes could occur in less than a day. The key to remember is that various definitions are commonplace for the word "evolution". For example, a common scientific meaning is a change in allele frequency. This may occur due to the generation of novel alleles or natural selection acting to change the frequency distribution of existing alleles. A common layman definition of evolution is the generation of novel DNA or a novel physiological trait.
There is little speed limit to the rate at which natural selection can act on an existing population. However, the generation of new alleles is limited by the laws of probability and the current population size and generation time. The article gives no indication that it was about novel alleles, therefore I assume that it is about the well-known fact that natural selection can act quickly.
Those pictures are mostly harmless compared to the other stuff cops do, but it sure grabs people's attention. You think the average person cares much about the Stingrays and parallel construction and warrantless searches and various other abuses? But tell them that when the cops pull over their wife or daughter they'll share her nude pictures, and they'll start caring.
Restrict freedom, that'll make them stop trying to take it all away...
Won't work. We're talking about some pretty crazy, messed up people.
With a mental condition. And a crack addiction. And who was begging to be put in jail to help with said crack addiction.
Or you could try doing something about crazy people with a crack addiction that are begging you to put them in jail and help them with their addiction. Instead of waiting until they kill someone and then pushing some random religious agenda cause muslims.
http://bc.ctvnews.ca/ottawa-shooter-tried-to-rob-vancouver-mcdonald-s-with-stick-1.2068636
Any one of various people who would want the new measures could have hired the guy. This would be far more cost-effective than lobbying, especially if they needed to sway public opinion instead of just legislators.
Yes, I don't actually enjoy games that have respawn nearly as much as ones that don't, or have expensive consequences.
I've played games where you don't respawn, for example Nethack. Can't say the no-respawn trait is either good or bad; it is a trade-off. Battles can be much more exciting when you know that it's game over if you lose. Except when that makes you so paranoid that you fight the weakest of creatures because you'll never win if you seek challenging fights. So when you can't afford to lose, the winning strategy is to refuse challenging fights.
Whereas when there's no consequences to losing a battle, you go for whatever is funnest, or most challenging, or has the best effort/reward ratio. But it could lead to apathy.
That it is easy enough to work around such promises. (Not saying Ello would do that, just saying this depends on their intentions not on these promises)
Yesterday's and today's comics on Dilbert show the proper response to sexual harassment.
dilbert.com/2014-10-22/
dilbert.com/2014-10-23/
Government plans that mean computer users deemed to have damaged national security, the economy or the environment will face a life sentence have been criticized by experts who warn that the new law could be used to target legitimate whistleblowers.
Could be? Come on - targeting whistleblowers is the point. It's not about damaging national security, the economy or the environment - it's about damaging somebody's political career.
Looks to me like the UK is threatening the US Department of Justice employees who have been hacking Facebook accounts with life in prison for damaging the feelings of security of the whole world and the environment of trust of Facebook. It's also going to target large corporations that are doing propaganda concerning pollution or global warming, as these things damage the environment. And the bankers, don't forget about the bankers.
Haha, just kidding like you said this looks like its aimed at whistleblowers.
Why shouldn't it be legal for me to write software that, as one of the features, detects and deals with low quality counterfeits? The problem is merely that users didn't know this would happen, probably because it was buried in a several page pile of legalese known as an EULA. Conversely, if they had placed what I would consider a fair notice on their software, it would inconvenience their legitimate users but no one would have felt sorry for the illegitimate ones.
The side that wants to fight evil by any means necessary. You know, the good guys.
I'm 100% absolutely sure that answers your question.
Canada raised their terrorism threat level, just before this happened. Clearly they were right. It's time to panic!
Unless they meant something else by "raising the domestic terrorism threat level" -- maybe they meant increasing the number of domestic terrorists, in which case the program appears to have been successful. Here in the US such a thing would energize the Republican base, so it's not unlikely that some politicians benefited greatly from this. Any Republicans in Canada?
Or, maybe in this case correlation does suggest causation. If they raise the threat level again, will another crazy guy shoot at Parliament?