While I am not quite as vehement about it as you, I agree to some extent. Firefox has been trying to Chrome-ify its interface, and it sucks. It needs to go back to its roots.
And goddamn Google for harming it. Firefox is our last best hope for a non-intrusive, "independent" browser. Firefox needs to start looking -- HARD -- for better outside funding.
I use cash or checks for 99% of my purchases. That way I avoid this issue. I'm also an old guy so "Get off my lawn!"
I also tend to stay away from places with "Bistro" in the name. You can generally count on a 50% or more higher price, with no commensurate increase in quality.
Actually, this is a quite silly, rather paranoid policy.
Google Glass shoots only low-resolution video. And put that together with the fact that peoples' heads move around a lot, a Google Glass copy of a film would be one that enjoyed practically zero popularity.
About the worst they would do is supply free advertising for the actual theater movie. Banning them was an ignorant thing to do.
That just shows the need of Turing tests in other languages. Intelligence isn't bound to a single language.
Nobody claimed that it is. This is all beside the point.
Turing's test was supposed to involve a "natural language" conversation. The presumption is that this is the same language spoken by the judges. I don't believe it involves any assumption that it is a particular language, only that it be the same on both sides of the conversation.
This bot failed that challenge. It deliberately simulated broken English in order to conceal its inability to actually use proper English. Therefore it did not meet the "natural language" criterion. It has nothing to do with the particular language chosen... had the test taken place in Chinese, and the bot had simulated broken Chinese, it would still have failed the natural language requirement.
You certainly can't say, given all the various variables that were changing, that greater police presence reduces crime.
But that is exactly the mistake here, which others have made as well. This situation has not controlled for other variables.
In the case of firearms availability vs violent crime, not only do we have millions of data points over many decades, but nobody has found other correlations or influencing factors. There have been lots of theories, but despite many many years of trying to find other correlations, nobody has managed to make any stick.
Students in a chemistry class study an average of 6 hours for a test and get a C for their class average. The next week, the same student study an average of 1 hour for their Interpretive Dance class and average an A..
Such a small sample doesn't prove anything, granted. But when you have millions of data points over periods of many decades, it is far harder to dismiss the evidence.
That is true only *IF* you can identify other influencing factors. If no other influencing factors can be identified then the presumption that your null hypothesis has been disproved must hold.
If other factors are identified later, then this presumption would have to be re-examined. But in their absence, the evidence is clear.
In the case of firearms availability vs violent crime (in the US), nobody has found other influencing factors. All other potential correlations have been pretty thoroughly shot down, with only one possible exception.
The ONE other societal variable that may have been shown to correlate with reduced crime (availability of contraception) has been widely criticized. Whether it stands remains to be seen. And even if so, it has not been shown to affect the negative firearms/crime relationship.
...with this rapid release schedule. Firefox is trying to update more often than Java nowadays.
For the most part I haven't minded, and for the most part, the changes have been appreciated.
However, version 29 revamped the entire toolbar customization scheme. Which has caused problems. Not only did it force me to move my refresh button (which for many years I kept on the left where it belongs next to the other navigation buttons), but it also eliminated the "addon bar" (which was historically the "status bar" at the bottom). That change broke the interfaces of a couple of add-ons I use.
Also, version 29 broke a web-crawling tool I use frequently. I got that fixed, but I should not have had to.
It has nothing to do with actual artificial intelligence and everything to do with writing deceptive scripts. It's not just this incident, it's a problem with the goal of the Turing test itself. I always found the Turing test a kind of stupid exercise due to this.
Yes. TechDirt's points 3 and 6 are basically the same thing I wrote here the other day:
First, that the "natural language" requirement was gamed. It deliberately simulated someone for whom English is not their first language, in order to cover its inability to actually hold a good English conversation. Fail.
Second, that we have learned over time that the Turing test doesn't really mean much of anything. We are capable of creating a machine that holds its own in limited conversation, but in the process we have learned that it has little to do with "AI".
I think some of TechDirt's other points are also valid. In point 4, for example, they explain that this wasn't even the real Turing test.
No, no, no. If you went through college, your statistics professor needs to be fired. Please, please stop posting about statistics before you break something.
I have had 3 statistics professors, and 4 textbooks, and 2 professors of logic. And they all said the same thing.
Which is more likely? That THEY were all wrong, or that some Anonymous Coward is?
Town A has 5 police per thousand people, and 3 crimes reported per thousand people every day. The next year, they increase the number of police to 7 per thousand people, but crime rates go up to 5 crimes reported per day.
Despite the negative correlation, this doesn't disprove the idea that having a greater police presence reduces crime.
Ahem... I hate to have to tell you this, but yes it did. The simple fact is: you had greater police presence, but crime went up. Your hypothesis has been disproved.
IN THIS CASE. It doesn't prove anything for, say, the county or the state or the country. But for the local situation, which is what you were talking about, sorry but your theory IN THIS LOCAL CASE was disproved.
But country-wide, in the U.S., we have 80+ YEARS of statistics, from across the country and for all the states. And we can say, with no fear of contradiction because the numbers are what they are, that crime has reduced (by more than HALF) as per-capita firearm ownership and frequency of carry have gone up. And yes, it does disprove the theory that more guns cause more crime.
If you have further argument with that, I strongly suggest you borrow somebody's statistics textbook, or take a class in formal logic.
I'm pro-gun (or at least anti gun restriction), but it's hardly indisputable disproof.
Yes, it is. Leaving guns aside, it's the way logic and statistics work together.
A correlation (we see lots of accidents involving cell phones, for example) does not imply cause-and-effect. There is often some outside factor (or even many factors) that influence the things correlated. A classic example, from Darrell Huff's "How To Lie With Statistics" is: the salary of Protestant ministers in the U.S. is strongly correlated with price of rum in Jamaica.
Does one cause the other? Of course not. The more likely answer is that general inflation (an outside factor) has affected both of them similarly.
But while a correlation does not prove cause-and-effect, a lack of correlation -- or more properly, a negative correlation -- can DISprove cause-and-effect. Example: something -- all evidence points to one animal -- has been killing your chickens. You suspect the neighbor's dog. So you start keeping tabs on when the dog is let out, and when it is in the house. It turns out, after examination, that whatever it is has been killing your chickens when the dog was locked up in the house. There is no dispute... it is indisputable that the dog wasn't there when the chickens died. This negative correlation between the dog being out and dead chickens has DISproved your theory that the dog was killing the chickens.
It gets a bit more complicated when the numbers go up but the same principle still holds. If your theory is that X causes Y, and you find a negative correlation, for example X goes up while Y goes down, you have DISproved that X causes Y. Otherwise, barring other outside influences, you would have (no dispute) observed that Y went up as X went up. Anything else contradicts your theory.
And in the gun-control debate, we have in fact had ample time and opportunity to control for other factors. And it is extremely important to note that try as we might, we have found no other causal factors that apply to the situation. Yet even so, as X (per-capita gun ownership and frequency of carry) has gone up, Y (violent crime of all sorts) has continued to go down. Therefore: X does not cause Y. Q.E.D.
There has been only ONE societal factor that has been found to satisfactorily correlate with the reduction in crime (see the movie Freakonomics, and that has been widely disputed.
Lacking any other evidence of outside factors, and even allowing for the one that (maybe) was found, we are still left with the simple mathematical fact that in the U.S., prevalence of gun ownership DOES NOT cause crime.
It isn't an opinion. It's as scientific as it gets.
And not just a little. FAR safer. Violent crime is less than half what it was 20 years ago. And even less compared to 30 years ago.
The only "increasing" violence is news-media propaganda. Because chicks hatching on the farm does not sell news.
In fact, some recent studies have concluded that it was news media coverage, and not guns, which led to copy-cat "mass" shootings on college and other school campuses. (But even so, and even though they are splashed all over the news, THOSE are way down, too, compared to 2-3 decades ago.)
American does not have "increasing" internal violence. It has decreasing violence.
Statistics do not prove cause-and-effect. But a negative correlation can DISprove cause-and-effect.
We have more guns. (Per person!) According to our own government's statistics. Yet we have less violent crime. This is a direct, indisputable DISproof of the idea that "more guns equals more crime".
[Sources: U.S. DOJ, and for more recent years: U.S. Federal Bureau of Justice Statistics]
You mean our atmosphere, the thing that keeps us alive and will hopefully continue to do so in the future as long as we don't fuck it up? That thing? Yeah, totally not worth the sliver of NASA's small budget.
I, for one, celebrate Hansen's stepping down. Maybe for a change we will get responsible science as opposed to Hansen's zealotry.
Given the content of your reply, I am going to give you some credit for relevance. But I do so only very cautiously, in light of your past behavior.
I say up front: if you have science to present, then present it. Facts and figures, with references. Otherwise, you have nothing to say to me. I have been very tolerant, and even so I do not like you, or your behavior, or your methods. But if you can produce real science, I will look at it.
Now, if the parent poster has evidence that coral TODAY can survive and thrive under the changing conditions discussed
In fact I did. I mentioned that the DAILY variability of pH is far higher than any change that can be honestly attributed to pH. If you were to do your research (which you obviously have not done), you would already know that these animals have been through conditions far "worse" (in CO2 terms) not just once but at least several times, and not just for 100 years but for thousands and millions of years.
Yet, they're still here.
If you don't think that's relevant, then you don't know what's relevant.
What does that have to do with anything? An ancestor of a lifeform from the Cambrian era 500 million years ago evolved to survive in the environment of 500 million years ago. That has no bearing on whether a descendant
WHOOSH...
It was presumed that participants in this discussion would possess both a certain amount of scientific literacy, AND some background on the issue.
So, to supply that deficiency (which I am not obligated to do): the basic argument that has been presented is that the acidification of the oceans via dissolved CO2 would harm sea life that depends on calciferous body parts. (Bivalves, for example, with their calciferous shells, and coral with its calciferous communal "skeleton".)
The reality, however, is that these creatures have survived long periods (thousands or even millions of years) of both higher CO2 levels and higher oceanic pH levels. Without noticeable damage.
I am NOT arguing that human changes to the environment are favorable. I am simply pointing out that these Earth creatures have been through far worse (for humans) conditions in the past and seem little the worse for wear. So if we want to be concerned, we should probably point our concerns elsewhere.
Turing explicitly said the test must be done in "natural language". While it violated the spirit, it violated the letter too because it did not use "natural language" of the speakers of that language. It used this as a trick to try to subvert the process.
It wasn't just "the spirit" (even though that is the most important part. It was also the letter.
"Natural language" must be considered to be general level of language and understanding of whoever is giving the test. Otherwise the test itself admits to meaninglessness.
In the U.S., laws are required to follow the logic of a "reasonable person", presumed to be both common and an adult. If you presume other standards you can "prove" anything you like, but again, as I originally stated: in the real world you have proved nothing.
Pardon me. NOAAs figures are that ocean rise has been about 0.9mm per year since 2010. Projecting backward (which I am not sure is a valid technique, Mann nothwithstanding) would give a rise of about 3".
So by projection only, without researching actual historical data this Sunday evening, I concede that 3" may be a reasonable figure.
Slashdot Mirror
Firefox's share is plummeting
While I am not quite as vehement about it as you, I agree to some extent. Firefox has been trying to Chrome-ify its interface, and it sucks. It needs to go back to its roots.
And goddamn Google for harming it. Firefox is our last best hope for a non-intrusive, "independent" browser. Firefox needs to start looking -- HARD -- for better outside funding.
On the other hand...
I use cash or checks for 99% of my purchases. That way I avoid this issue. I'm also an old guy so "Get off my lawn!"
I also tend to stay away from places with "Bistro" in the name. You can generally count on a 50% or more higher price, with no commensurate increase in quality.
"Bob's Chinese" is more likely to try harder.
Actually, this is a quite silly, rather paranoid policy.
Google Glass shoots only low-resolution video. And put that together with the fact that peoples' heads move around a lot, a Google Glass copy of a film would be one that enjoyed practically zero popularity.
About the worst they would do is supply free advertising for the actual theater movie. Banning them was an ignorant thing to do.
That just shows the need of Turing tests in other languages. Intelligence isn't bound to a single language.
Nobody claimed that it is. This is all beside the point.
Turing's test was supposed to involve a "natural language" conversation. The presumption is that this is the same language spoken by the judges. I don't believe it involves any assumption that it is a particular language, only that it be the same on both sides of the conversation.
This bot failed that challenge. It deliberately simulated broken English in order to conceal its inability to actually use proper English. Therefore it did not meet the "natural language" criterion. It has nothing to do with the particular language chosen... had the test taken place in Chinese, and the bot had simulated broken Chinese, it would still have failed the natural language requirement.
You certainly can't say, given all the various variables that were changing, that greater police presence reduces crime.
But that is exactly the mistake here, which others have made as well. This situation has not controlled for other variables.
In the case of firearms availability vs violent crime, not only do we have millions of data points over many decades, but nobody has found other correlations or influencing factors. There have been lots of theories, but despite many many years of trying to find other correlations, nobody has managed to make any stick.
Students in a chemistry class study an average of 6 hours for a test and get a C for their class average. The next week, the same student study an average of 1 hour for their Interpretive Dance class and average an A..
Such a small sample doesn't prove anything, granted. But when you have millions of data points over periods of many decades, it is far harder to dismiss the evidence.
We DO have such.
In an uncontrolled study, it disproves nothing.
That is true only *IF* you can identify other influencing factors. If no other influencing factors can be identified then the presumption that your null hypothesis has been disproved must hold.
If other factors are identified later, then this presumption would have to be re-examined. But in their absence, the evidence is clear.
In the case of firearms availability vs violent crime (in the US), nobody has found other influencing factors. All other potential correlations have been pretty thoroughly shot down, with only one possible exception.
The ONE other societal variable that may have been shown to correlate with reduced crime (availability of contraception) has been widely criticized. Whether it stands remains to be seen. And even if so, it has not been shown to affect the negative firearms/crime relationship.
...with this rapid release schedule. Firefox is trying to update more often than Java nowadays.
For the most part I haven't minded, and for the most part, the changes have been appreciated.
However, version 29 revamped the entire toolbar customization scheme. Which has caused problems. Not only did it force me to move my refresh button (which for many years I kept on the left where it belongs next to the other navigation buttons), but it also eliminated the "addon bar" (which was historically the "status bar" at the bottom). That change broke the interfaces of a couple of add-ons I use.
Also, version 29 broke a web-crawling tool I use frequently. I got that fixed, but I should not have had to.
Unless they have some way to distinguish "free" wifi access via your router from your own traffic, then this scheme is incompatible with usage caps.
I am all for open access points. I have been running one for years. BUT, I do not condone Comcast doing it to people without their prior consent.
I use my own DOCSIS 3 cable adapter, and my own router, with my own private network and an open public network. My ISP has no say in the matter.
It has nothing to do with actual artificial intelligence and everything to do with writing deceptive scripts. It's not just this incident, it's a problem with the goal of the Turing test itself. I always found the Turing test a kind of stupid exercise due to this.
Yes. TechDirt's points 3 and 6 are basically the same thing I wrote here the other day:
First, that the "natural language" requirement was gamed. It deliberately simulated someone for whom English is not their first language, in order to cover its inability to actually hold a good English conversation. Fail.
Second, that we have learned over time that the Turing test doesn't really mean much of anything. We are capable of creating a machine that holds its own in limited conversation, but in the process we have learned that it has little to do with "AI".
I think some of TechDirt's other points are also valid. In point 4, for example, they explain that this wasn't even the real Turing test.
No, no, no. If you went through college, your statistics professor needs to be fired. Please, please stop posting about statistics before you break something.
I have had 3 statistics professors, and 4 textbooks, and 2 professors of logic. And they all said the same thing.
Which is more likely? That THEY were all wrong, or that some Anonymous Coward is?
Statistically (and for other reasons), you lose.
Town A has 5 police per thousand people, and 3 crimes reported per thousand people every day. The next year, they increase the number of police to 7 per thousand people, but crime rates go up to 5 crimes reported per day.
Despite the negative correlation, this doesn't disprove the idea that having a greater police presence reduces crime.
Ahem... I hate to have to tell you this, but yes it did. The simple fact is: you had greater police presence, but crime went up. Your hypothesis has been disproved.
IN THIS CASE. It doesn't prove anything for, say, the county or the state or the country. But for the local situation, which is what you were talking about, sorry but your theory IN THIS LOCAL CASE was disproved.
But country-wide, in the U.S., we have 80+ YEARS of statistics, from across the country and for all the states. And we can say, with no fear of contradiction because the numbers are what they are, that crime has reduced (by more than HALF) as per-capita firearm ownership and frequency of carry have gone up. And yes, it does disprove the theory that more guns cause more crime.
If you have further argument with that, I strongly suggest you borrow somebody's statistics textbook, or take a class in formal logic.
I'm pro-gun (or at least anti gun restriction), but it's hardly indisputable disproof.
Yes, it is. Leaving guns aside, it's the way logic and statistics work together.
A correlation (we see lots of accidents involving cell phones, for example) does not imply cause-and-effect. There is often some outside factor (or even many factors) that influence the things correlated. A classic example, from Darrell Huff's "How To Lie With Statistics" is: the salary of Protestant ministers in the U.S. is strongly correlated with price of rum in Jamaica.
Does one cause the other? Of course not. The more likely answer is that general inflation (an outside factor) has affected both of them similarly.
But while a correlation does not prove cause-and-effect, a lack of correlation -- or more properly, a negative correlation -- can DISprove cause-and-effect. Example: something -- all evidence points to one animal -- has been killing your chickens. You suspect the neighbor's dog. So you start keeping tabs on when the dog is let out, and when it is in the house. It turns out, after examination, that whatever it is has been killing your chickens when the dog was locked up in the house. There is no dispute... it is indisputable that the dog wasn't there when the chickens died. This negative correlation between the dog being out and dead chickens has DISproved your theory that the dog was killing the chickens.
It gets a bit more complicated when the numbers go up but the same principle still holds. If your theory is that X causes Y, and you find a negative correlation, for example X goes up while Y goes down, you have DISproved that X causes Y. Otherwise, barring other outside influences, you would have (no dispute) observed that Y went up as X went up. Anything else contradicts your theory.
And in the gun-control debate, we have in fact had ample time and opportunity to control for other factors. And it is extremely important to note that try as we might, we have found no other causal factors that apply to the situation. Yet even so, as X (per-capita gun ownership and frequency of carry) has gone up, Y (violent crime of all sorts) has continued to go down. Therefore: X does not cause Y. Q.E.D.
There has been only ONE societal factor that has been found to satisfactorily correlate with the reduction in crime (see the movie Freakonomics, and that has been widely disputed.
Lacking any other evidence of outside factors, and even allowing for the one that (maybe) was found, we are still left with the simple mathematical fact that in the U.S., prevalence of gun ownership DOES NOT cause crime.
It isn't an opinion. It's as scientific as it gets.
the US is safer now than ever before.
And not just a little. FAR safer. Violent crime is less than half what it was 20 years ago. And even less compared to 30 years ago.
The only "increasing" violence is news-media propaganda. Because chicks hatching on the farm does not sell news.
In fact, some recent studies have concluded that it was news media coverage, and not guns, which led to copy-cat "mass" shootings on college and other school campuses. (But even so, and even though they are splashed all over the news, THOSE are way down, too, compared to 2-3 decades ago.)
American does not have "increasing" internal violence. It has decreasing violence.
And during the same period, it is interesting to not, per-capita gun ownership in the U.S. has gone steadily up. And also during that same period, concealed-carry laws have become much more common.
Statistics do not prove cause-and-effect. But a negative correlation can DISprove cause-and-effect.
We have more guns. (Per person!) According to our own government's statistics. Yet we have less violent crime. This is a direct, indisputable DISproof of the idea that "more guns equals more crime".
[Sources: U.S. DOJ, and for more recent years: U.S. Federal Bureau of Justice Statistics]
"causal relation as well as correlation". They are not mutually exclusive. On the contrary: the former requires the latter.
I'm not so sure, it would be hard to pin down the number of people that played along after they worked out it was a bot from transcripts alone.
Please read the relevant literature. It was studied in a responsible manner and measurable results were obtained.
You mean our atmosphere, the thing that keeps us alive and will hopefully continue to do so in the future as long as we don't fuck it up? That thing? Yeah, totally not worth the sliver of NASA's small budget.
I, for one, celebrate Hansen's stepping down. Maybe for a change we will get responsible science as opposed to Hansen's zealotry.
Given the content of your reply, I am going to give you some credit for relevance. But I do so only very cautiously, in light of your past behavior.
I say up front: if you have science to present, then present it. Facts and figures, with references. Otherwise, you have nothing to say to me. I have been very tolerant, and even so I do not like you, or your behavior, or your methods. But if you can produce real science, I will look at it.
s/attributed to pH/attributed to pH changes due to CO2
Now, if the parent poster has evidence that coral TODAY can survive and thrive under the changing conditions discussed
In fact I did. I mentioned that the DAILY variability of pH is far higher than any change that can be honestly attributed to pH. If you were to do your research (which you obviously have not done), you would already know that these animals have been through conditions far "worse" (in CO2 terms) not just once but at least several times, and not just for 100 years but for thousands and millions of years.
Yet, they're still here.
If you don't think that's relevant, then you don't know what's relevant.
Pardon me. Higher pH should have been higher acidity = lower pH.
What does that have to do with anything? An ancestor of a lifeform from the Cambrian era 500 million years ago evolved to survive in the environment of 500 million years ago. That has no bearing on whether a descendant
WHOOSH...
It was presumed that participants in this discussion would possess both a certain amount of scientific literacy, AND some background on the issue.
So, to supply that deficiency (which I am not obligated to do): the basic argument that has been presented is that the acidification of the oceans via dissolved CO2 would harm sea life that depends on calciferous body parts. (Bivalves, for example, with their calciferous shells, and coral with its calciferous communal "skeleton".)
The reality, however, is that these creatures have survived long periods (thousands or even millions of years) of both higher CO2 levels and higher oceanic pH levels. Without noticeable damage.
I am NOT arguing that human changes to the environment are favorable. I am simply pointing out that these Earth creatures have been through far worse (for humans) conditions in the past and seem little the worse for wear. So if we want to be concerned, we should probably point our concerns elsewhere.
Turing explicitly said the test must be done in "natural language". While it violated the spirit, it violated the letter too because it did not use "natural language" of the speakers of that language. It used this as a trick to try to subvert the process.
It wasn't just "the spirit" (even though that is the most important part. It was also the letter.
"Natural language" must be considered to be general level of language and understanding of whoever is giving the test. Otherwise the test itself admits to meaninglessness.
In the U.S., laws are required to follow the logic of a "reasonable person", presumed to be both common and an adult. If you presume other standards you can "prove" anything you like, but again, as I originally stated: in the real world you have proved nothing.
Pardon me. NOAAs figures are that ocean rise has been about 0.9mm per year since 2010. Projecting backward (which I am not sure is a valid technique, Mann nothwithstanding) would give a rise of about 3".
So by projection only, without researching actual historical data this Sunday evening, I concede that 3" may be a reasonable figure.