Yes because NASA never inventedanything we use every single day.
We get HUGE bang for our buck in NASA. If you want to cut wasteful spending, you could cut NASA's budget several dozen times over from the military and they'd barely feel it. NASA is probably the best example we have of a government organization gone right, and all people seem to want to do is cut it because they don't understand how science works. Things like NASA exist because all of their inventions came out of necessity of the incredibly complex things they were doing. Those inventions make billions of dollars for many companies. We probably wouldn't have invented half the stuff NASA has come out with because the current stuff we had was "good enough" for life down here on Earth.
To be fair to the AI researchers, they had no idea how the brain worked because neuropsychologists had no tools to figure it out back then. Now the psychologists have a slightly better understanding of how the brain works (though they are still in the dark about most things), and some of that has been put back into AI.
AI also wouldn't have seemed to fail so hard if not for Minsky and Papert publishing that there was no way a perceptron could solve the XOR problem, killing the field for ~25-30 years. Thankfully, it was solved by adding more perceptrons and reignited the field.
"Extraordinary claims require extraordinary evidence" is valid. If you tell me there is an invisible dragon in your garage, you better have some damn good evidence to back that claim up. In a slightly more realistic assertion, if you tell me that vaccines cause autism, then you also better have some damn good evidence to back that claim up. Extraordinary claims cannot be treated on equal footing as an ordinary claim because taking extraordinary claims at face value can be far more dangerous.
I haven't read the books, but that article is crap. The entire thing just says "evolution is clearly happening, so we should reinterpret the bible to say that God just got he ball rolling." It is an exercise in altering religious views to conform to modern science, not an exercise in scientific thought. It is just arguing that we should modify religion to become a "God of the gaps", which is a silly argument indeed.
That is true. It isn't religion that causes wars/genocides, it is religious institutions and the brainwashing that they enforce upon their members (governments who have committed such atrocities also are guilty of the exact same thing). Though I'd argue that power is more the goal than money.
My experience in the community is just that no one cares unless it starts effecting your science or hypotheses. Theist or atheist, if you're good at what you do no one cares. If you go around preaching to other scientists, yeah, you're opening yourself up for ridicule. But I think that is true in any field outside of the more religious areas of the US.
Free thinking expands science. Indoctrinating people (children) into ways of thinking stifles science. It is hard to break free of 18+ years of having a belief system drilled into your head.
That being said people should be able to believe what they want, but indoctrinating children or others by force is somewhat more iffy.
Science can tell you "why" also, such as why the Earth is round. I don't see why can can't just leave questions unanswered and we have to make up an answer for them. Perhaps in thousands of years science will progress to the point where it is possible to answer some questions previously thought impossible. A "God of the gaps" is a silly god.
'To remove the perceived stigma, we would need to have more scientists talking openly about issues of religion, where such issues are particularly relevant to their discipline.'
How often is religion relevant to a field of science where it needs to be discussed? Really, religion is inappropriate to discuss in scientific terms, as the entire point of religion is blind faith. Science has no place in religion, just as religion has no place in science (as in, "God did it" as a valid hypothesis, not as in the scientist's personal belief structures, which they are more than entitled to and I know many scientists who are also theists).
Yeah! All of that machine learning theory and heuristic problem solving really turned me into a Marxist! Man, thanks for pointing that out. It couldn't possibly be that, when taught critical thinking skills, someone realizes that a belief in the Earth being ~6000 years old is a particularly silly one?
Nope, vast Left wing conspiracy to indoctrinate the population. Totally much more likely.
In all of my schooling through undergraduate and (currently) graduate, in both political science and computer science, I have never heard a leftist viewpoint espoused that I need not meet after class specifically to talk about. I've had both left and right wing profs, but I only found out by talking to them about current issues after class. I'd always try to guess, but I was probably within error of chance. If anything, studying in a university has made me more conservative than when I started, as now I have the critical thinking skills and background knowledge to know how naieve and foolish my old views were.
This Marxist myth is silly and makes those on the Right who espouse it look like tin foil hat nutjobs. Maybe it was true in the 60s, I don't know. But it certainly isn't now.
Contribute in large, meaningful ways to open source projects. Write your own software and release it. There are plenty of ways to show programming competence that don't require a degree.
Except for that IQ tests are largely crap, I mostly agree with your point. They are measuring something that seems to be correlated with intelligence, but do not measure intelligence directly, thus there are going to outliers of people who just aren't good at whatever the IQ test is measuring, but excellent at other things (Savants might be an example here).
Also, throwing numbers around is largely meaningless, as the numbers are only valid when compared to others numbers in relatively small cultural groups. Things like primary language, the school's curriculum, interactions with objects, etc. will all change one's IQ score.
That being said, I think you are correct that some people just can't hack some of the higher level stuff, but I disagree with how black and white you make it seem.
Generally you want the other pair of eyes to also be an expert in your field, so they can criticize your work. If bad data is released into the wild, it is very hard to get back and make sure nothing gets published or released with it.
Also, if money was what researchers were after there are plenty of other fields where you can make tons more money for much less work.
You would be removing the main incentive for researchers; prestige amongst their peers and a feeling of having contributed to the furthering of human knowledge by not allowing them to publish their own data sets. You would also be removing many of the checks and balances internal to the system that help weed out a lot of crap.
Also, I don't get why "release when published" is so crazy. Do you go down to Washington and demand that your Senate rep opens up Word on his computer to show you the half completed/fleshed out draft of the bill he was working on? Do you want to go into the police station and demand to know their list of informants and undercover cops because they are paid with public money? Anything with the military?
Your position is unreasonable and counterproductive.
I call at least slightly bogus on that "report". I'm not trying to be an apologist for some of the bone headed mistakes in the IPCC (the WWF references are pretty inexcusable), but "Working Paper" doesn't mean that it hasn't been published. An alternative (and very European) definition is synonymous with "technical report", usually by some government institution.
While these may not be peer reviewed, despite what the IPCC originally said, they aren't "papers in progress" or "drafts".
Just a little nitpick in her article. I would also like her to publish a full list of the 1/3 of the the claims for others to check her work, until then, I see a handful of things she has posted on the web that I agree that many are dubious.
However, most of them are from the smaller sections of the report that were more addendums than anything else. The main information is still solid. Working group 1, by far the largest section of the report and containing almost all of the important information and findings, scored As and Bs on her "report card" for every chapter.
Generally by the time publication comes around, you've already prepared your data for presentation because you have a good idea whether or not the reviewers are going to ask for your raw data/processing code/whatnot. That varies depending on the data set and the reviewers.
The biggest part that is the problem of releasing the data right away is because you don't know if the data is crap without a lot of vetting. Once and a while, it doesn't get caught until a fresh pair of eyes (reviewers) take a look at it and pick up on something maybe your team hasn't thought of.
The other problem is then you would create a whole industry of research snipers who would poach papers from researchers, thus eliminating the benefits of gathering huge, difficult to collect data sets. There would be no incentive to spend 10 years gathering data, only to have your research sniped as you are writing up your papers.
Sure, the data belongs to the people (and rightfully so, information should be free), but you have to give some incentive to researchers who are already severely underpaid and overworked to painstakingly collect data. Pretty much the only thing they have is recognition.
Again, releasing data is fine. I am all for it. I think data should absolutely be released once the lab that generated is finished with it, and that that should be shortly after the initial batch of papers is published. This also gives the scientists time to fully understand and test the data, and understand the limitations of how it was collected.
Data should be released, yes. And like I said, you can always pay more taxes so labs can hire a grad student or someone to release data for them as soon as it is generated.
But then you'll just complain about the data you can't get because of privacy constraints or NDAs. Labs that generate data should have first dibs on that data, but I absolutely agree with you that is should be released on a reasonable time scale after their publications from it were made. However your demands of releasing it immediately are unreasonable and counter-productive to science.
It would probably be a dump of a large, fully annotated and understood database system. Or there are plenty of files I have that are several hundred megabytes where the columns are x y z and are motion tracking files, which are fully annotated in perfect order as far as protocol is concerned, but if I simply handed you that file with no preparation of what any of it actual is, then you have no idea what the hell to do with it.
When we analyze that data, we know exactly what the limitations of the system that collected it are, what parts of the data are good and which parts are bad, what went wrong during that study, etc. But do to privacy issues, we cannot turn over videos of the subjects performing the tasks (this is an IRB rule, not ours) without consent of the subjects. Handing you a text file without the video is worthless. Data without context, understanding of exactly how every piece of it was collected, or understanding of its limitations is meaningless at best. At worst, it is dangerous in the hands of someone with a larger megaphone who does not understand the data (as the AGW deniers love to do).
Like I said though, I think that data should be released, but releasing data as soon as it is generated rather than after the relevant papers have been published is an undue burden on the already heavily time constrained researchers.
I think he meant "indistinguishable to the layperson". That is a real issue. Jenny McCarthy can say all she wants, and to Joe Blow, it is all the same as what scientists say. Scientists know she is full of shit, but people not trained in science might not be able to tell.
Hire people to get the data to you then. I would imagine that most researchers would be fine releasing data if the time and money burden of doing so was lifted from them.
Slashdot Mirror
Which woud bolster the decision to remove Aspergers from the Autism Spectrum in the new DSV
Because we're the leading research country in the world and we like to keep it that way?
Yes because NASA never invented anything we use every single day.
We get HUGE bang for our buck in NASA. If you want to cut wasteful spending, you could cut NASA's budget several dozen times over from the military and they'd barely feel it. NASA is probably the best example we have of a government organization gone right, and all people seem to want to do is cut it because they don't understand how science works. Things like NASA exist because all of their inventions came out of necessity of the incredibly complex things they were doing. Those inventions make billions of dollars for many companies. We probably wouldn't have invented half the stuff NASA has come out with because the current stuff we had was "good enough" for life down here on Earth.
To be fair to the AI researchers, they had no idea how the brain worked because neuropsychologists had no tools to figure it out back then. Now the psychologists have a slightly better understanding of how the brain works (though they are still in the dark about most things), and some of that has been put back into AI.
AI also wouldn't have seemed to fail so hard if not for Minsky and Papert publishing that there was no way a perceptron could solve the XOR problem, killing the field for ~25-30 years. Thankfully, it was solved by adding more perceptrons and reignited the field.
"Extraordinary claims require extraordinary evidence" is valid. If you tell me there is an invisible dragon in your garage, you better have some damn good evidence to back that claim up. In a slightly more realistic assertion, if you tell me that vaccines cause autism, then you also better have some damn good evidence to back that claim up. Extraordinary claims cannot be treated on equal footing as an ordinary claim because taking extraordinary claims at face value can be far more dangerous.
Thank you for this post, this is why I read /.
You got hurt, it just hurt you less than the thing it was killing did.
I haven't read the books, but that article is crap. The entire thing just says "evolution is clearly happening, so we should reinterpret the bible to say that God just got he ball rolling." It is an exercise in altering religious views to conform to modern science, not an exercise in scientific thought. It is just arguing that we should modify religion to become a "God of the gaps", which is a silly argument indeed.
That is true. It isn't religion that causes wars/genocides, it is religious institutions and the brainwashing that they enforce upon their members (governments who have committed such atrocities also are guilty of the exact same thing). Though I'd argue that power is more the goal than money.
My experience in the community is just that no one cares unless it starts effecting your science or hypotheses. Theist or atheist, if you're good at what you do no one cares. If you go around preaching to other scientists, yeah, you're opening yourself up for ridicule. But I think that is true in any field outside of the more religious areas of the US.
Free thinking expands science. Indoctrinating people (children) into ways of thinking stifles science. It is hard to break free of 18+ years of having a belief system drilled into your head.
That being said people should be able to believe what they want, but indoctrinating children or others by force is somewhat more iffy.
Science can tell you "why" also, such as why the Earth is round. I don't see why can can't just leave questions unanswered and we have to make up an answer for them. Perhaps in thousands of years science will progress to the point where it is possible to answer some questions previously thought impossible. A "God of the gaps" is a silly god.
'To remove the perceived stigma, we would need to have more scientists talking openly about issues of religion, where such issues are particularly relevant to their discipline.'
How often is religion relevant to a field of science where it needs to be discussed? Really, religion is inappropriate to discuss in scientific terms, as the entire point of religion is blind faith. Science has no place in religion, just as religion has no place in science (as in, "God did it" as a valid hypothesis, not as in the scientist's personal belief structures, which they are more than entitled to and I know many scientists who are also theists).
Yeah! All of that machine learning theory and heuristic problem solving really turned me into a Marxist! Man, thanks for pointing that out. It couldn't possibly be that, when taught critical thinking skills, someone realizes that a belief in the Earth being ~6000 years old is a particularly silly one?
Nope, vast Left wing conspiracy to indoctrinate the population. Totally much more likely.
In all of my schooling through undergraduate and (currently) graduate, in both political science and computer science, I have never heard a leftist viewpoint espoused that I need not meet after class specifically to talk about. I've had both left and right wing profs, but I only found out by talking to them about current issues after class. I'd always try to guess, but I was probably within error of chance. If anything, studying in a university has made me more conservative than when I started, as now I have the critical thinking skills and background knowledge to know how naieve and foolish my old views were.
This Marxist myth is silly and makes those on the Right who espouse it look like tin foil hat nutjobs. Maybe it was true in the 60s, I don't know. But it certainly isn't now.
Contribute in large, meaningful ways to open source projects. Write your own software and release it. There are plenty of ways to show programming competence that don't require a degree.
Except for that IQ tests are largely crap, I mostly agree with your point. They are measuring something that seems to be correlated with intelligence, but do not measure intelligence directly, thus there are going to outliers of people who just aren't good at whatever the IQ test is measuring, but excellent at other things (Savants might be an example here).
Also, throwing numbers around is largely meaningless, as the numbers are only valid when compared to others numbers in relatively small cultural groups. Things like primary language, the school's curriculum, interactions with objects, etc. will all change one's IQ score.
That being said, I think you are correct that some people just can't hack some of the higher level stuff, but I disagree with how black and white you make it seem.
Do you live on the moon?
Generally you want the other pair of eyes to also be an expert in your field, so they can criticize your work. If bad data is released into the wild, it is very hard to get back and make sure nothing gets published or released with it.
Also, if money was what researchers were after there are plenty of other fields where you can make tons more money for much less work.
You would be removing the main incentive for researchers; prestige amongst their peers and a feeling of having contributed to the furthering of human knowledge by not allowing them to publish their own data sets. You would also be removing many of the checks and balances internal to the system that help weed out a lot of crap.
Also, I don't get why "release when published" is so crazy. Do you go down to Washington and demand that your Senate rep opens up Word on his computer to show you the half completed/fleshed out draft of the bill he was working on? Do you want to go into the police station and demand to know their list of informants and undercover cops because they are paid with public money? Anything with the military?
Your position is unreasonable and counterproductive.
I call at least slightly bogus on that "report". I'm not trying to be an apologist for some of the bone headed mistakes in the IPCC (the WWF references are pretty inexcusable), but "Working Paper" doesn't mean that it hasn't been published. An alternative (and very European) definition is synonymous with "technical report", usually by some government institution.
While these may not be peer reviewed, despite what the IPCC originally said, they aren't "papers in progress" or "drafts".
Just a little nitpick in her article. I would also like her to publish a full list of the 1/3 of the the claims for others to check her work, until then, I see a handful of things she has posted on the web that I agree that many are dubious.
However, most of them are from the smaller sections of the report that were more addendums than anything else. The main information is still solid. Working group 1, by far the largest section of the report and containing almost all of the important information and findings, scored As and Bs on her "report card" for every chapter.
Generally by the time publication comes around, you've already prepared your data for presentation because you have a good idea whether or not the reviewers are going to ask for your raw data/processing code/whatnot. That varies depending on the data set and the reviewers.
The biggest part that is the problem of releasing the data right away is because you don't know if the data is crap without a lot of vetting. Once and a while, it doesn't get caught until a fresh pair of eyes (reviewers) take a look at it and pick up on something maybe your team hasn't thought of.
The other problem is then you would create a whole industry of research snipers who would poach papers from researchers, thus eliminating the benefits of gathering huge, difficult to collect data sets. There would be no incentive to spend 10 years gathering data, only to have your research sniped as you are writing up your papers.
Sure, the data belongs to the people (and rightfully so, information should be free), but you have to give some incentive to researchers who are already severely underpaid and overworked to painstakingly collect data. Pretty much the only thing they have is recognition.
Again, releasing data is fine. I am all for it. I think data should absolutely be released once the lab that generated is finished with it, and that that should be shortly after the initial batch of papers is published. This also gives the scientists time to fully understand and test the data, and understand the limitations of how it was collected.
Data should be released, yes. And like I said, you can always pay more taxes so labs can hire a grad student or someone to release data for them as soon as it is generated.
But then you'll just complain about the data you can't get because of privacy constraints or NDAs. Labs that generate data should have first dibs on that data, but I absolutely agree with you that is should be released on a reasonable time scale after their publications from it were made. However your demands of releasing it immediately are unreasonable and counter-productive to science.
It would probably be a dump of a large, fully annotated and understood database system. Or there are plenty of files I have that are several hundred megabytes where the columns are x y z and are motion tracking files, which are fully annotated in perfect order as far as protocol is concerned, but if I simply handed you that file with no preparation of what any of it actual is, then you have no idea what the hell to do with it.
When we analyze that data, we know exactly what the limitations of the system that collected it are, what parts of the data are good and which parts are bad, what went wrong during that study, etc. But do to privacy issues, we cannot turn over videos of the subjects performing the tasks (this is an IRB rule, not ours) without consent of the subjects. Handing you a text file without the video is worthless. Data without context, understanding of exactly how every piece of it was collected, or understanding of its limitations is meaningless at best. At worst, it is dangerous in the hands of someone with a larger megaphone who does not understand the data (as the AGW deniers love to do).
Like I said though, I think that data should be released, but releasing data as soon as it is generated rather than after the relevant papers have been published is an undue burden on the already heavily time constrained researchers.
I think he meant "indistinguishable to the layperson". That is a real issue. Jenny McCarthy can say all she wants, and to Joe Blow, it is all the same as what scientists say. Scientists know she is full of shit, but people not trained in science might not be able to tell.
Hire people to get the data to you then. I would imagine that most researchers would be fine releasing data if the time and money burden of doing so was lifted from them.
Nope. Plenty of classified research goes on in university research labs.