Yep, I would say this is likely another case of reporting doing more damage than good. People are going to react by not trusting the blood pressure monitor without understanding that it was never supposed to be that accurate in the first place. At what point is it just unethical to go for the scare headline at the cost of putting people's health in danger.
My impression from reading about blood pressure measurement is that the actual value, not just its uncertainty, also depends on the rate of release. The rate of release is supposed to be part of the measurement standard. This is one of the biggest reasons that the manual method is very uncertain, it is hard for a human to do it with a constant, fixed release rate. If you "slow down" near the expected pressure like some suggest --- I think you're doing it wrong. Also, as others have said, a single BP measurement isn't SUPPOSED to be accurate to 5mmHg. Presumably the article is not totally stupid and is taking average measurements.
It also seems like poorly constructed samples might fail this test, so it is possible that some fraction of the 43 are not intentional fraud, but just poor science. My impression is that it can be tricky to construct a good sample for a lot of medical studies, so it's not surprising that some of them are imperfect. I can also think of situations where correctly constructed samples might fail this test, but it's possible that the analysis is accounting for that.
Well, I think this is a bit more than "just an algorithm" but even so, I agree that machine learning is not AI. That being said, when PR people and reporters say "AI" they often mean "machine learning", so the battle for the proper use of this term is likely already lost. This really should be "Google's AlphaGo Trained Machine Defeats...". I believe that is a much better description anyhow. It is a lot more like a trained animal than an intelligent machine.
By the standards of Soyuz and the current SpaceX effort, the shuttle wasn't even a man-ratable launch system. It was designed without the required abort modes. Also the Russians can't raise their prices too much or it will become cost-effective to accelerate SpaceX's effort. They only a have a handful of tests to complete, and while they are by far the furthest along, they are not the only one. If you want the scientists and engineers to figure it out -- let NASA procure launch services rather than rocket components.
It is almost trivial to kill cookies (self-destructing cookies firefox plugin) and use an email provider that doesn't profile you. Yes many people do this now. Setting up a VPN just to keep your own ISP from selling your profile is ridiculous, but now apparently necessary.
Well, if you are searching for the false article about presidents in the clan that everyone is talking about, then it is just what you wanted! Search engines are just a map between search terms and results, where rank is largely determined by popularity. So there is always the effective filter of "perform this search assuming I want the result most often chosen in pop culture". If you wanted a "only include verified claims" you're using the wrong tool. And who would believe something on the internet from a source with no reputation? People talk about how wikipedia is bad. A random site has even less chance of being true. The problem is not that fake news exists, the problem is that the general population appears to have lost their ability to distinguish truth from fiction. I think the issue is that many people don't have a conceptual framework for "unverified" statements. i.e. they automatically force-classify everything as either true or false. This is in conflict with how "news" is currently run, where the first reporting of something is always when it is in the "unverified" state. Often by the time it transitions to the "confirmed" or "falsified" state, nobody pays any attention anymore. (It's actually worse than that, as some news outlets will willfully delay reporting the information that moves it to the "falsified" state so as to prolong the controversy. The surest way to get a viewer to switch to another channel is to resolve the issue!)
My opinion is that git and svn have largely different purposes. The centralized/distributed one is the most obvious. But also git is a revision history manager, whereas svn is mostly a revision history. And I am convinced that simplicity is in the eye of the beholder.
Well, sort of. The problem is that facebook (et al) is not a news aggregator, it is an advertising and publicity company paid to keep people on the app and revealing information about themselves that allows them to be put in easily exploited marketing groups. It is a meme-generation machine for rent. Why is it any surprise that the constant attempt to hack people's judgement and social interactions to trick them into wanting the latest crappy crap leads to collateral damage to their social and mental health?
Well if that's your point of view, Obama did great. One of his first executive orders was to close gitmo, keeping a campaign promise. That worked... not.
Sorry, but if you want something like that there's this thing called the legislature that makes laws. I know it sucks to not live in a dictatorship, but that's how it is here.
Well according to your buddy Trump, in order to require computer programming you'll have to remove two other graduation requirements! Math and science? Reading and writing?
Used appropriately, H1B holders actually contribute expertise and productivity to US economy, which is a plus. Teaching and research positions at Universities are examples where this is obvious -- the value of having the foreign worker's expertise is direct and clear because he or she was compared directly to US candidates in the selection process and one of their major roles is to educate US students. But my understanding is that educational institutions have different rules from businesses. Educational institutions are able to choose a foreigner because they are the best applicant. Businesses must show that there are NO qualified domestic applicants. So the headline doesn't make sense to me, it's already required. They just aren't enforcing it very well.
But it would not surprise me at all for an executive order to be oblivious to the fact that this is already required. The track record so far indicates that the Trump administration does not actually understand how the government that they are now in charge of actually works. The most likely thing seems to be that they will phrase the same requirement in some way that is stupidly incompatible with the actual laws and it will just make a huge mess with so many loopholes that it will get worse instead of better. Competence is so underrated.
The point is that if WhatsApp is not blocked and Signal is, using WhatsApp is better than other options. You say yourself that the single block-able route is not the difference, its that one is blocked and the other isn't. As for the article, I would say that if someone's life or freedom depends on whether WhatsApp is secure -- they better well understand how this vulnerability applies to them based on their specific usage pattern, not based on some generalization from a newspaper article.
Um, Facebook (the company) owns whatsapp, so facebook owns their data. Transfer complete. The rest is just a shell game subject to how much money they want to spend on litigation.
I would just remind you that Putin also denies Russian involvement in Ukraine. The intelligence agencies have said they will release more detailed information in the next month, and we can judge it then.
Given the nature of the hack, it seems like if Podesta had just enabled two-factor like he was told, the typo wouldn't have mattered and even giving the hackers his password wouldn't have mattered. The IT guy says right there that two-factor should be enabled as soon as possible, and even implies that it already should have been. Actually this level of person not using two-factor is just madness. And how does the first part of the email even make sense? Why would he use mfisher's two-step verification codes?
I won't say there aren't districts where there is waste, there are plenty, but one does have to build and maintain buildings, which can be quite expensive. There are also pensions to consider, which are not always managed sensibly. So a lot depends on what is in that $18k figure. If it includes facilities and the district made a really bad facilities decision 15-20 years ago, that can be a huge chunk right there that the teachers don't see one penny of. And in that situation, blind cuts will just hurt the people who are just doing their job because the overhead of that bad decision is probably impossible to change. (e.g. a bond issue or similar)
And this idea that parents pay taxes for the education of their own children is madness. Taxes that go to schools principally pay for the education of *other people's* children. That is and always has been what public schools are for. It seems like this should be obvious, since people don't just pay those taxes while their children are in school. So the thing you should argue for is better schools at the same cost. The community has already decided it is worth $18k per year per student, but that value for that money can be improved.
But the problem is that, as you say, the $30M gift probably cost Alphabet half that but they probably wrote off the full amount. i.e. it's just a scam to get a $15M tax write-off. And it's probably less than half so they get an even bigger fraction as write-off. If they gave that hardware to employees, they could still write it off as a business expense, but only at cost. So this is the employee being told "instead of giving you a gift, we are giving a poor child's school your gift (which they will probably never use), and ourselves an even bigger gift as a result." Because Alphabet itself is clearly the neediest party in this situation.
A few reporters getting duped is NOT the same as intentionally and repeatedly manufacturing fake news. This is actually a strategy of the fake news pushers... try to dress something up well enough to get picked up by the more reputable sources. Then declare them irriputable when one of dozens of false-flag operations gets through. But the point is that this story was (a) discredited by other news sources and (b) will be or already has been retracted by the WP and (c) reporters will lose their jobs or be demoted, not promoted like they would at a fake news operation.
Also, I'm sure that Rolling Stone has been scouring the Washington Post for something poorly sourced like this, since it was the Post that eviscerated Rolling Stone for the expose on gang rape on college campuses, which turned out to be fabricated by the victim (though actually in a very convincing way).
The problem isn't fake news, it's a public who can't pay attention longer than one 24-hour news cycle. The real facts are almost never known within 24 hours. If people drew their conclusions after there had been enough time for consideration and cross-checking, fake news would have no power. Just like superpacs would have no power if people would just not believe a word of what someone paid for them to hear, duh.
It's funny to me that they say the Earth's rotation is "determined" by IERS. No, it is *measured*. A subtle implied difference, but a critical one. The problem here is actually with unix time, not UTC. TAI as an alternative to unix time actually makes pretty decent sense. When was the last time you manually converted from seconds since epoch to day/month/year?
I think the point is that since time zones are actually updated more frequently for political purposes than leap seconds occur, it makes sense, for network-connected computers at least, to just stuff the leap seconds in the same distribution channels (already done actually) and abandon trying to hack around the clumsy (non-monotonically increasing) definition of unix time.
Slashdot Mirror
Yep, I would say this is likely another case of reporting doing more damage than good. People are going to react by not trusting the blood pressure monitor without understanding that it was never supposed to be that accurate in the first place. At what point is it just unethical to go for the scare headline at the cost of putting people's health in danger.
My impression from reading about blood pressure measurement is that the actual value, not just its uncertainty, also depends on the rate of release. The rate of release is supposed to be part of the measurement standard. This is one of the biggest reasons that the manual method is very uncertain, it is hard for a human to do it with a constant, fixed release rate. If you "slow down" near the expected pressure like some suggest --- I think you're doing it wrong. Also, as others have said, a single BP measurement isn't SUPPOSED to be accurate to 5mmHg. Presumably the article is not totally stupid and is taking average measurements.
It also seems like poorly constructed samples might fail this test, so it is possible that some fraction of the 43 are not intentional fraud, but just poor science. My impression is that it can be tricky to construct a good sample for a lot of medical studies, so it's not surprising that some of them are imperfect. I can also think of situations where correctly constructed samples might fail this test, but it's possible that the analysis is accounting for that.
"so long as they are turned off" is unrealistic. Some people barely even know the difference between suspend and off. And plenty are likely to forget.
Well, I think this is a bit more than "just an algorithm" but even so, I agree that machine learning is not AI. That being said, when PR people and reporters say "AI" they often mean "machine learning", so the battle for the proper use of this term is likely already lost. This really should be "Google's AlphaGo Trained Machine Defeats...". I believe that is a much better description anyhow. It is a lot more like a trained animal than an intelligent machine.
By the standards of Soyuz and the current SpaceX effort, the shuttle wasn't even a man-ratable launch system. It was designed without the required abort modes. Also the Russians can't raise their prices too much or it will become cost-effective to accelerate SpaceX's effort. They only a have a handful of tests to complete, and while they are by far the furthest along, they are not the only one. If you want the scientists and engineers to figure it out -- let NASA procure launch services rather than rocket components.
It is almost trivial to kill cookies (self-destructing cookies firefox plugin) and use an email provider that doesn't profile you. Yes many people do this now. Setting up a VPN just to keep your own ISP from selling your profile is ridiculous, but now apparently necessary.
Well, if you are searching for the false article about presidents in the clan that everyone is talking about, then it is just what you wanted! Search engines are just a map between search terms and results, where rank is largely determined by popularity. So there is always the effective filter of "perform this search assuming I want the result most often chosen in pop culture". If you wanted a "only include verified claims" you're using the wrong tool. And who would believe something on the internet from a source with no reputation? People talk about how wikipedia is bad. A random site has even less chance of being true. The problem is not that fake news exists, the problem is that the general population appears to have lost their ability to distinguish truth from fiction. I think the issue is that many people don't have a conceptual framework for "unverified" statements. i.e. they automatically force-classify everything as either true or false. This is in conflict with how "news" is currently run, where the first reporting of something is always when it is in the "unverified" state. Often by the time it transitions to the "confirmed" or "falsified" state, nobody pays any attention anymore. (It's actually worse than that, as some news outlets will willfully delay reporting the information that moves it to the "falsified" state so as to prolong the controversy. The surest way to get a viewer to switch to another channel is to resolve the issue!)
My opinion is that git and svn have largely different purposes. The centralized/distributed one is the most obvious. But also git is a revision history manager, whereas svn is mostly a revision history. And I am convinced that simplicity is in the eye of the beholder.
Well, sort of. The problem is that facebook (et al) is not a news aggregator, it is an advertising and publicity company paid to keep people on the app and revealing information about themselves that allows them to be put in easily exploited marketing groups. It is a meme-generation machine for rent. Why is it any surprise that the constant attempt to hack people's judgement and social interactions to trick them into wanting the latest crappy crap leads to collateral damage to their social and mental health?
So use news.google.com without being logged in and with cookies deleted. Everyone uses the firefox self-destructing cookies plugin right?
Ha Ha... the most obvious exporters of terrorism... you mean Saudi Arabia and Pakistan right?
Well if that's your point of view, Obama did great. One of his first executive orders was to close gitmo, keeping a campaign promise. That worked... not.
Sorry, but if you want something like that there's this thing called the legislature that makes laws. I know it sucks to not live in a dictatorship, but that's how it is here.
Well according to your buddy Trump, in order to require computer programming you'll have to remove two other graduation requirements! Math and science? Reading and writing?
Used appropriately, H1B holders actually contribute expertise and productivity to US economy, which is a plus. Teaching and research positions at Universities are examples where this is obvious -- the value of having the foreign worker's expertise is direct and clear because he or she was compared directly to US candidates in the selection process and one of their major roles is to educate US students. But my understanding is that educational institutions have different rules from businesses. Educational institutions are able to choose a foreigner because they are the best applicant. Businesses must show that there are NO qualified domestic applicants. So the headline doesn't make sense to me, it's already required. They just aren't enforcing it very well.
But it would not surprise me at all for an executive order to be oblivious to the fact that this is already required. The track record so far indicates that the Trump administration does not actually understand how the government that they are now in charge of actually works. The most likely thing seems to be that they will phrase the same requirement in some way that is stupidly incompatible with the actual laws and it will just make a huge mess with so many loopholes that it will get worse instead of better. Competence is so underrated.
The point is that if WhatsApp is not blocked and Signal is, using WhatsApp is better than other options. You say yourself that the single block-able route is not the difference, its that one is blocked and the other isn't. As for the article, I would say that if someone's life or freedom depends on whether WhatsApp is secure -- they better well understand how this vulnerability applies to them based on their specific usage pattern, not based on some generalization from a newspaper article.
Um, Facebook (the company) owns whatsapp, so facebook owns their data. Transfer complete. The rest is just a shell game subject to how much money they want to spend on litigation.
It sounds like these are known issues that just aren't fixed yet on some distributions. That's not a zero day.
I would just remind you that Putin also denies Russian involvement in Ukraine. The intelligence agencies have said they will release more detailed information in the next month, and we can judge it then.
Given the nature of the hack, it seems like if Podesta had just enabled two-factor like he was told, the typo wouldn't have mattered and even giving the hackers his password wouldn't have mattered. The IT guy says right there that two-factor should be enabled as soon as possible, and even implies that it already should have been. Actually this level of person not using two-factor is just madness. And how does the first part of the email even make sense? Why would he use mfisher's two-step verification codes?
I won't say there aren't districts where there is waste, there are plenty, but one does have to build and maintain buildings, which can be quite expensive. There are also pensions to consider, which are not always managed sensibly. So a lot depends on what is in that $18k figure. If it includes facilities and the district made a really bad facilities decision 15-20 years ago, that can be a huge chunk right there that the teachers don't see one penny of. And in that situation, blind cuts will just hurt the people who are just doing their job because the overhead of that bad decision is probably impossible to change. (e.g. a bond issue or similar)
And this idea that parents pay taxes for the education of their own children is madness. Taxes that go to schools principally pay for the education of *other people's* children. That is and always has been what public schools are for. It seems like this should be obvious, since people don't just pay those taxes while their children are in school. So the thing you should argue for is better schools at the same cost. The community has already decided it is worth $18k per year per student, but that value for that money can be improved.
But the problem is that, as you say, the $30M gift probably cost Alphabet half that but they probably wrote off the full amount. i.e. it's just a scam to get a $15M tax write-off. And it's probably less than half so they get an even bigger fraction as write-off. If they gave that hardware to employees, they could still write it off as a business expense, but only at cost. So this is the employee being told "instead of giving you a gift, we are giving a poor child's school your gift (which they will probably never use), and ourselves an even bigger gift as a result." Because Alphabet itself is clearly the neediest party in this situation.
A few reporters getting duped is NOT the same as intentionally and repeatedly manufacturing fake news. This is actually a strategy of the fake news pushers... try to dress something up well enough to get picked up by the more reputable sources. Then declare them irriputable when one of dozens of false-flag operations gets through. But the point is that this story was (a) discredited by other news sources and (b) will be or already has been retracted by the WP and (c) reporters will lose their jobs or be demoted, not promoted like they would at a fake news operation.
Also, I'm sure that Rolling Stone has been scouring the Washington Post for something poorly sourced like this, since it was the Post that eviscerated Rolling Stone for the expose on gang rape on college campuses, which turned out to be fabricated by the victim (though actually in a very convincing way).
The problem isn't fake news, it's a public who can't pay attention longer than one 24-hour news cycle. The real facts are almost never known within 24 hours. If people drew their conclusions after there had been enough time for consideration and cross-checking, fake news would have no power. Just like superpacs would have no power if people would just not believe a word of what someone paid for them to hear, duh.
It's funny to me that they say the Earth's rotation is "determined" by IERS. No, it is *measured*. A subtle implied difference, but a critical one. The problem here is actually with unix time, not UTC. TAI as an alternative to unix time actually makes pretty decent sense. When was the last time you manually converted from seconds since epoch to day/month/year?
I think the point is that since time zones are actually updated more frequently for political purposes than leap seconds occur, it makes sense, for network-connected computers at least, to just stuff the leap seconds in the same distribution channels (already done actually) and abandon trying to hack around the clumsy (non-monotonically increasing) definition of unix time.