Why is everyone talking about masking or disabling the hypothetical red light. What are you supposed to do if you walk into a cafe and there are 3 or 4 red lights around? And why is it now my responsibility to check? Even it's it's not actually hidden, it's awful.
But as they get closer in size a pocket calculator than to a refrigerator
Actually I've seen two instances recently of refrigerators (one basic consumer level, one high-end consumer) where one part failed and the whole refrigerator had to be disposed of and replaced. Current manufacturing is generally heading down this road, optimizing for cheapness up to the moment of sale.
What really makes me upset about this is that I bet these are Energy-Star certified, because I doubt the certification takes into account the expected lifetime, repairability, and environmental costs of manufacture.
I think the essense of the problem is not the fact that everyone's opinion counts equally, which it does in democracy, but the way that those opinions are being informed. That's what's changing for the worse.
First of all nearly all social media now is guiding people towards opinions in line with their own. It's a technically beautiful idea, but the more successful the technology, the worse for democracy.
Slashdotters at Google, Facebook, even Amazon, everywhere else directing eyes: do you see how you're seriously subverting and polarizing our democracy. As a media organization you need to take responsibility for making sure you are not hiding alternative viewpoints. By doing this you're squeezing out the center, both in terms of how voters identify themselves and discuss issues, and in terms of politicians' loyalties. The incredible shrinking center
And there's also the continued bite-sized-ification that prevents construction of thought-out arguments. This has been progressing for a long time, but keeps getting worse.
In both cases, we need to compare to what we had when the technology for communication was largely print and newspapers, and think about how to replace what's been eroded away.
That's probably because no one uses a web search that isn't Google.
... and while I basically never use anything but google day-to-day, I'd be pretty comfortable using it to mean general web-searching, whatever the tool.
(maybe I should go altavista whether someone's got a more in depth discussion of google as a verb... heh. I had to go check if altavista still exists after I wrote the joke. It does. Sorry if that comment touches a nerve for someone.)
While Dr. Jacobson makes some good points, most of them are pretty obvious to anyone who has taken a good class on creating experiments with humans
I did not find the article trivial like that. There's a lot of useful thinking the author has put into Turk specifically---and how it intersects with research methods, IRB, etc. For instance his own experiments to help think about what to pay people. (Not a lot of detail given on how he did this, but helpful information nonetheless.) Also helpful technical information on how, within the structure of MTurk, to maintain contact with subjects for follow-ups, make payments efficiently, etc.
Being a techie (and lefty) person with many techy and mostly lefty associates, I've seen the people around me move gradually from a prevalance to use various unixes to more macs as the mac under OSX became better. This isn't as insulting to conservative minded people as it sounds. It's conservative in the good sense to use something well-tested before jumping along to the next hot thing.
Yes, somestimes a close race is just a close race, but we happen to be at a moment where I and lots of other people don't feel comfortable trusting the system. It seems worthwhile to think about whether there are any ways to make sure we can trust the results, although I've never read anything suggesting there's much hope for added confidence (whether things are close or not.)
It's not crazy to be paranoid when the stakes are incredibly high. (NASA, the pentagon, etc. aren't crazy when they're careful.) I think the last line in the recent arstechnica e-voting article sums it up well:
"The clock is ticking on this issue, because a party that can use these techniques to gain control of the government can also use them to maintain control in perpetuity."
Eliminate all dangerous states and replace them with democracies structured in such a way to make it very difficult for them to wage war. Then advocate free trade, with disputes mediated by an international organization and you take away a reason for other powerful states to make war on you. Then you can get buy with a minimal army.
Just look at history and you can see the value of such a plan.
One thing I'd be curious to hear about is horror stories of switching some aspect of a business over to open source and getting burned. Is there a wrong way/wrong situation to do it?
I'm not so worried about my old pdf. I'm worried about the viewer. This is a pdf that looked fine in acrobat 4 when I made it, so why would I have done it differently? I assume I'll come across other pdf's on the web that will break in Preview if mine did.
On a simple latex document (converted with ps2pdf) Preview misrendered some plain text subheadings into garbage, and also where it was working, the font just looked worse (It was computer modern, I think.) Acrobat looked fine. I may have had something set up wrong in Preview perhaps.
Slashdot Mirror
Why is everyone talking about masking or disabling the hypothetical red light. What are you supposed to do if you walk into a cafe and there are 3 or 4 red lights around? And why is it now my responsibility to check? Even it's it's not actually hidden, it's awful.
Actually I've seen two instances recently of refrigerators (one basic consumer level, one high-end consumer) where one part failed and the whole refrigerator had to be disposed of and replaced. Current manufacturing is generally heading down this road, optimizing for cheapness up to the moment of sale.
What really makes me upset about this is that I bet these are Energy-Star certified, because I doubt the certification takes into account the expected lifetime, repairability, and environmental costs of manufacture.
I think the essense of the problem is not the fact that everyone's opinion counts equally, which it does in democracy, but the way that those opinions are being informed. That's what's changing for the worse.
First of all nearly all social media now is guiding people towards opinions in line with their own. It's a technically beautiful idea, but the more successful the technology, the worse for democracy.
Slashdotters at Google, Facebook, even Amazon, everywhere else directing eyes: do you see how you're seriously subverting and polarizing our democracy. As a media organization you need to take responsibility for making sure you are not hiding alternative viewpoints. By doing this you're squeezing out the center, both in terms of how voters identify themselves and discuss issues, and in terms of politicians' loyalties.
The incredible shrinking center
And there's also the continued bite-sized-ification that prevents construction of thought-out arguments. This has been progressing for a long time, but keeps getting worse.
In both cases, we need to compare to what we had when the technology for communication was largely print and newspapers, and think about how to replace what's been eroded away.
That's probably because no one uses a web search that isn't Google.
... and while I basically never use anything but google day-to-day, I'd be pretty comfortable using it to mean general web-searching, whatever the tool.
(maybe I should go altavista whether someone's got a more in depth discussion of google as a verb... heh. I had to go check if altavista still exists after I wrote the joke. It does. Sorry if that comment touches a nerve for someone.)
I did not find the article trivial like that. There's a lot of useful thinking the author has put into Turk specifically---and how it intersects with research methods, IRB, etc. For instance his own experiments to help think about what to pay people. (Not a lot of detail given on how he did this, but helpful information nonetheless.) Also helpful technical information on how, within the structure of MTurk, to maintain contact with subjects for follow-ups, make payments efficiently, etc.
I don't see the relevance of the number 17,000. The only question is whether the 35 is big enough.
Being a techie (and lefty) person with many techy and mostly lefty associates, I've seen the people around me move gradually from a prevalance to use various unixes to more macs as the mac under OSX became better. This isn't as insulting to conservative minded people as it sounds. It's conservative in the good sense to use something well-tested before jumping along to the next hot thing.
Yes, somestimes a close race is just a close race, but we happen to be at a moment where I and lots of other people don't feel comfortable trusting the system. It seems worthwhile to think about whether there are any ways to make sure we can trust the results, although I've never read anything suggesting there's much hope for added confidence (whether things are close or not.)
It's not crazy to be paranoid when the stakes are incredibly high. (NASA, the pentagon, etc. aren't crazy when they're careful.) I think the last line in the recent arstechnica e-voting article sums it up well:
"The clock is ticking on this issue, because a party that can use these techniques to gain control of the government can also use them to maintain control in perpetuity."
What are you referring to?
One thing I'd be curious to hear about is horror stories of switching some aspect of a business over to open source and getting burned. Is there a wrong way/wrong situation to do it?
I'm not so worried about my old pdf. I'm worried about the viewer. This is a pdf that looked fine in acrobat 4 when I made it, so why would I have done it differently? I assume I'll come across other pdf's on the web that will break in Preview if mine did.
On a simple latex document (converted with ps2pdf) Preview misrendered some plain text subheadings into garbage, and also where it was working, the font just looked worse (It was computer modern, I think.) Acrobat looked fine. I may have had something set up wrong in Preview perhaps.