Apparently you didn't even read to the second half of the second sentence of the article you link - reproducibility relies on Ceteris paribus. It is only necessary that experiments be reproducible in principle.
It's not a nude photograph - the content of the image has nothing to do with TFA's point.
it is from a magazine that exists to objectify women and titillate men
Stipulated. The interesting questions are, given the above two elements, should the image, despite it's unobjectionable content, be considered 'bad' in some way because of its provenance, and if so, why.
So far all I've heard is appeals to emotion - with which I am inclined to agree - but no arguments.
I have to waste some mod points to give the reasons. The legislation bans consideration of research where all data is not publicly available without regard for which data is available - like public health studies with anonymized data.
This bill would make it impossible for the EPA to use many health studies, since they often contain private patient information that can’t and shouldn’t be revealed. Studies based on confidential business information would also be off-limits. Studies of human exposures to toxics over time and from a variety of locations likely cannot be reproduced. Neither can meta-analyses, looking at the results of hundreds of scientific studies to assess their conclusions. Such studies provide critical scientific evidence in many fields of research. This legislation wasn’t designed to promote good science—it was crafted to prevent public health and environmental laws from being enforced.
it may be a widespread problem, but it is self-correcting.
I'm not entirely convinced it is self-correcting. Some scholars have named the MBA phenomenon as a key component in the secular decline of the average rate of profit since mid century. (Yes, profits are higher than ever now, but not the rate of profit).
[1] Brenner, Robert. The economics of global turbulence: the advanced capitalist economies from long boom to long downturn, 1945-2005. Verso, 2006.
[2] Brenner, Robert. "What is Good for Goldman Sachs is Good for America The Origins of the Present Crisis." (2009).
Chrome is the new IE: Some pages only load on it...
Chrome is also the new IE because a bunch of other pages don't work on it at all. I just started switching back to Firefox because I was sick of so many compatibility issues with Chrome. (Other reasons like frequent brief lockups on one computer, the non-freeness, and the Eye of Mordor contributed somewhat).
When I look around me I see the prices of food, electricity, and a whole host of other things I might buy frequently increasing (or the sizes of things like food decreasing while the price stays the same) by many percent a year.
You simply aren't seeing that - instead your wages have been stagnating so the extremely modest inflation we've been experiencing seems increasingly onerous. It's the proportion of your budget these things require has been increasing significantly, not their prices.
it's genuine criticism of the use of an image from Playboy.
Yes - my post above is very terse, but I want it to be made clear that the problem with the image isn't its content, but solely its provenance.
If the point is that the Playboy empire is so nasty that anything related to it is indelibly tainted by association (a position I'm sympathetic to), then I want to hear arguments supporting that position.
Similarly, if the claim is that the cropped image is pornography itself, then I want to see reasoning to that effect. Are cropped subimages of pornographic images, which themselves do not have pornographic content, nevertheless pornography? The post I replied to above had stated this as fact, without argument (thus my "no it's not").
I am not personally offended by it...But surely you accept the empirical evidence that many other people do find even the idea of the use of a centerfold image lifted (even cropped) from Playboy to be inappropriate in an academic setting...I agree its inappropriate...not something I'd knowingly do...
That's all fine, but I have yet to hear an actual argument for why the provenance of the cropped image should be the deciding factor here, since rational assessment of information is usually based on content rather than provenance. I'm sympathetic to the position, but we are talking about recreating an image which is identical in every way, except for out-of-shot information which cannot be recovered.
If the only argument is that Playboy is so bad that the cropped image is indelibly tainted by association, then I guess I'm fine with that - but the logical foundation seems shaky.
Way too many computer vision papers get through while relying on just one or a few images, or without any real results (e.g. "look at the features!"). I can't count the times a promising-sounding segmentation algorithm turns out to be based on one or two easy images...IMHO if anything you should probably become even more stringent about lazy image selection in your reviews.
From Wikipedia: "An ad hominem (Latin for "to the man" or "to the person"), short for argumentum ad hominem, means responding to arguments by attacking a person's character, rather than to the content of their arguments."
The Lenna image is supposed to be "bad" solely because of its provenance (Playboy) rather than because of any content, so the analogy to an ad hominem argument is pretty clear.
Hell, I could recreate the pose with a volunteer model (wife), 10 minutes, and trip to the thrift shop for.
What's the point? The only difference would be the photographer. Just because the Playboy empire specifically is (maybe) bad? I don't see how continuing to use the image helps them, though.
How is it not blind puritanism to label anyone who's looked at pornography - or in this case what was once regarded as pornography - a "slimy scum bag?"
I think half of the problems today are because people assume if I use X in a discussion, I support X, and therefore we cant use or talk about X.
I encounter this all the time - people (especially my fellow millennials) are uncomfortable with any reference to subjectivity without loads of boilerplate ("just an opinion," "someone else might say," etc.). Even obvious statements of personal opinion/taste are frequently taken as offensive/arrogant assertions of objective truth, unless tiresome disclaimers are constantly issued.
For a visual algorithm test nobody would complain that the person on the image wasn't diverse enough.
Of course - for a single image. But computer vision research should employ a large number of images, and if they are not diverse, then I do object - on technical grounds. I'm working on a face processing algorithm now, where a lot of past work employs almost entirely Chinese face images. The results are noticeably worse for black and white (African/Caucasian, not monochrome) faces. So diversity in image databases for research use is important.
It has to do with the fact that it is an entirely outdated test image
Fair enough, it is pretty small by current standards.
Poor properties to visually assess the effects of image processing algorithms...Retrospectively, a variety of academics have justified its suitability (e.g., the fine detail of the feathers, the texture of the hat, contrasted with the smooth skin tone; as well as the uniquely human ability to perceive minute aberrations in facial structure)
All of those features are real. It also has a mirror and a partially occluded face. It has a bunch of regions which tend to oversegmentation by most algorithms. No computer vision result can be taken seriously unless many images are used (yes, that excludes much old research), but the Lenna image is certainly sufficient to be placed in standard test libraries.
but this is really a post-hoc rationalization not supported in the face of such facts as the image as it is frequently used is not even color balanced.
I disagree - I think that the relative suitability of the image is a large part of why it has persisted for so long.
I don't personally object to the image's content. But I absolutely understand why others would.
Almost no one is seriously objecting to the image's content (certainly not TFA), but only to its source. You should address that fact and argue why that is justified, since the content thing is a red herring.
Slashdot Mirror
I'm sure you'd object if people started copying and distributing the content of your emails without your permission.
Fortunately, such things are already illegal totally outside any form of intellectual property law.
taken without permission. The last 3 words in the previous sentence define theft.
No, they don't. Theft is taking scarce good without permission. You can keep using their newspeak if you want though.
If so, their morality is only dictated by laws and regulations
Yes, but don't forget that B-corps exist.
Here, you admit that only reproducibility in principle is required.
Apparently you didn't even read to the second half of the second sentence of the article you link - reproducibility relies on Ceteris paribus. It is only necessary that experiments be reproducible in principle.
It's a problem if they're only down to insufficient demand. It would be find if they were down and demand was adequate.
There has also been persistent oversupply - Brenner argues that that is the main driver in the decline.
If I look at a nude photograph
It's not a nude photograph - the content of the image has nothing to do with TFA's point.
it is from a magazine that exists to objectify women and titillate men
Stipulated. The interesting questions are, given the above two elements, should the image, despite it's unobjectionable content, be considered 'bad' in some way because of its provenance, and if so, why.
So far all I've heard is appeals to emotion - with which I am inclined to agree - but no arguments.
I have to waste some mod points to give the reasons. The legislation bans consideration of research where all data is not publicly available without regard for which data is available - like public health studies with anonymized data.
This bill would make it impossible for the EPA to use many health studies, since they often contain private patient information that can’t and shouldn’t be revealed. Studies based on confidential business information would also be off-limits. Studies of human exposures to toxics over time and from a variety of locations likely cannot be reproduced. Neither can meta-analyses, looking at the results of hundreds of scientific studies to assess their conclusions. Such studies provide critical scientific evidence in many fields of research. This legislation wasn’t designed to promote good science—it was crafted to prevent public health and environmental laws from being enforced.
it may be a widespread problem, but it is self-correcting.
I'm not entirely convinced it is self-correcting. Some scholars have named the MBA phenomenon as a key component in the secular decline of the average rate of profit since mid century. (Yes, profits are higher than ever now, but not the rate of profit).
[1] Brenner, Robert. The economics of global turbulence: the advanced capitalist economies from long boom to long downturn, 1945-2005. Verso, 2006.
[2] Brenner, Robert. "What is Good for Goldman Sachs is Good for America The Origins of the Present Crisis." (2009).
Microsoft itself is ditching IE for a new browser codenamed "Spartan"
I think they just announced that Spartan will be called Edge. </eyeroll>
Chrome is the new IE: Some pages only load on it...
Chrome is also the new IE because a bunch of other pages don't work on it at all. I just started switching back to Firefox because I was sick of so many compatibility issues with Chrome. (Other reasons like frequent brief lockups on one computer, the non-freeness, and the Eye of Mordor contributed somewhat).
The difference you see with Internet Explorer being "overcounted" shows that it occupies a long tail of many users who don't browse the web very often
And the users who occasionally need IE for something specific that still doesn't work in other browsers.
When I look around me I see the prices of food, electricity, and a whole host of other things I might buy frequently increasing (or the sizes of things like food decreasing while the price stays the same) by many percent a year.
You simply aren't seeing that - instead your wages have been stagnating so the extremely modest inflation we've been experiencing seems increasingly onerous. It's the proportion of your budget these things require has been increasing significantly, not their prices.
And when a page is loading in one tab, other tabs don't continue to update swiftly.
Chrome lazy-loads pages once the tabs are actually selected. In Firefox, it's an option you can set in the preferences dialog (turned on by default).
it's genuine criticism of the use of an image from Playboy.
Yes - my post above is very terse, but I want it to be made clear that the problem with the image isn't its content, but solely its provenance.
If the point is that the Playboy empire is so nasty that anything related to it is indelibly tainted by association (a position I'm sympathetic to), then I want to hear arguments supporting that position.
Similarly, if the claim is that the cropped image is pornography itself, then I want to see reasoning to that effect. Are cropped subimages of pornographic images, which themselves do not have pornographic content, nevertheless pornography? The post I replied to above had stated this as fact, without argument (thus my "no it's not").
I am not personally offended by it...But surely you accept the empirical evidence that many other people do find even the idea of the use of a centerfold image lifted (even cropped) from Playboy to be inappropriate in an academic setting...I agree its inappropriate...not something I'd knowingly do...
That's all fine, but I have yet to hear an actual argument for why the provenance of the cropped image should be the deciding factor here, since rational assessment of information is usually based on content rather than provenance. I'm sympathetic to the position, but we are talking about recreating an image which is identical in every way, except for out-of-shot information which cannot be recovered.
If the only argument is that Playboy is so bad that the cropped image is indelibly tainted by association, then I guess I'm fine with that - but the logical foundation seems shaky.
Way too many computer vision papers get through while relying on just one or a few images, or without any real results (e.g. "look at the features!"). I can't count the times a promising-sounding segmentation algorithm turns out to be based on one or two easy images...IMHO if anything you should probably become even more stringent about lazy image selection in your reviews.
From Wikipedia: "An ad hominem (Latin for "to the man" or "to the person"), short for argumentum ad hominem, means responding to arguments by attacking a person's character, rather than to the content of their arguments."
The Lenna image is supposed to be "bad" solely because of its provenance (Playboy) rather than because of any content, so the analogy to an ad hominem argument is pretty clear.
Hell, I could recreate the pose with a volunteer model (wife), 10 minutes, and trip to the thrift shop for.
What's the point? The only difference would be the photographer. Just because the Playboy empire specifically is (maybe) bad? I don't see how continuing to use the image helps them, though.
How is it not blind puritanism to label anyone who's looked at pornography - or in this case what was once regarded as pornography - a "slimy scum bag?"
I think half of the problems today are because people assume if I use X in a discussion, I support X, and therefore we cant use or talk about X.
I encounter this all the time - people (especially my fellow millennials) are uncomfortable with any reference to subjectivity without loads of boilerplate ("just an opinion," "someone else might say," etc.). Even obvious statements of personal opinion/taste are frequently taken as offensive/arrogant assertions of objective truth, unless tiresome disclaimers are constantly issued.
For a visual algorithm test nobody would complain that the person on the image wasn't diverse enough.
Of course - for a single image. But computer vision research should employ a large number of images, and if they are not diverse, then I do object - on technical grounds. I'm working on a face processing algorithm now, where a lot of past work employs almost entirely Chinese face images. The results are noticeably worse for black and white (African/Caucasian, not monochrome) faces. So diversity in image databases for research use is important.
It has to do with the fact that it is an entirely outdated test image
Fair enough, it is pretty small by current standards.
Poor properties to visually assess the effects of image processing algorithms...Retrospectively, a variety of academics have justified its suitability (e.g., the fine detail of the feathers, the texture of the hat, contrasted with the smooth skin tone; as well as the uniquely human ability to perceive minute aberrations in facial structure)
All of those features are real. It also has a mirror and a partially occluded face. It has a bunch of regions which tend to oversegmentation by most algorithms. No computer vision result can be taken seriously unless many images are used (yes, that excludes much old research), but the Lenna image is certainly sufficient to be placed in standard test libraries.
but this is really a post-hoc rationalization not supported in the face of such facts as the image as it is frequently used is not even color balanced.
I disagree - I think that the relative suitability of the image is a large part of why it has persisted for so long.
I don't personally object to the image's content. But I absolutely understand why others would.
Almost no one is seriously objecting to the image's content (certainly not TFA), but only to its source. You should address that fact and argue why that is justified, since the content thing is a red herring.
not so long ago 14 year olds were getting married and having children
And a bunch of US states currently have age of consent set to 16, with close-in-age exceptions going down to 14.