The most advanced "electronic noses" actually use trained insects with electrodes attached. All the power of the biological nose, but less hungry than dogs. Also quicker to train with lower volumes of scent sample required. Pure Pavlovian conditioning.
It's not that complicated: it's really hard to make a profit selling people their own mind's ability to heal them.
The placebo only works if you don't think it's a placebo. So if people believe in weird shit, yet it increase to success of the placebo effect then where's the harm?
The harm is that it discourages patients from seeking other treatments that will give both placebo benefits and genuine medical benefits.
People make money by encouraging people to believe that sugar that has touched water that has touched water that has touched water that has touched water that has touched water that has touched arsenic is more effective than a proven medical treatment. That's immoral. Your justification could be applied to any and all fraud.
Except that many homeopathic treatments are actually water that has no active ingredients dropped onto sugar pills then dried. So no, not even dehydration. Just hypoglycaemia.
Microsoft are losing ground to Google, Amazon Web Services etc in the cloud computing market, so they've decided a free course is the best way to get people using their product
Ah, the old.. market share = competence argument. I guess Linux is the single most worst desktop operating system ever built.
It's nothing to do with how good or bad Azure or its competitors are -- all I said is that Microsoft are not currently doing in a commercially strong position in the cloud marketplace, and they want to change that.
A teacher cannot be a billboard.
By that logic, nobody should be teaching mechanics how to repair a Subaru, cuz.. y know all mechanics should know how every single car works.
You people are deluded. People *want* to learn about/get trained on MS products, because the entire fucking world uses them. Like GP said, I'm glad if even a handful of people get trained and improve their life by finding better jobs. If you want to offer them something better, put your money where your mouth is.
There is a big difference between teaching and training. You teach someone about how cars work and how to repair them. You train someone in using Subaru's engine management system. (Of course, in many teaching courses, you simultaneously train in the use of particular technologies or techniques, but that is only because you need a concrete environment for practise.)
edX (and Coursera) is supposed to be a teaching platform, but there was never any money in that. Udacity claimed to be teaching, but was to all intents and purposes a training platform from day one. It was no surprise that it very quickly started offering vendor-specific sponsor (training) courses, but I'm disappointed that edX are following suit.
Teaching vs training is not just a problem in online education -- CS departments the world over are under constant pressure to train students in industry standard tools so that new employees can hit the ground running, rather than having to be sent on training courses for their employers' choice of tools. You would not teach a two-year diploma in Subaru maintenance -- the Subaru garage would expect to have to train new staff, but if they can find someone with Subaru experience, so much the better!
And the problem extends beyond universities, because in computing more than most fields, you need constant training to keep up with new technologies, and employers are loathe to do that. They expect people to come in with the training already, which means people are now expected to use their own time and money throughout their career just to keep their jobs. By supporting this, edX are actually threatening their own revenue streams, as they already have a "training" business in the form of edX Professional Education, an all-paid-for brand which they're trying to sell to employers, and the existence of free training courses on the teaching site dilutes the apparent value of these.
Then you're also in favour of demolishing the William Gates building at several universities
No, but I'd take his f#@king name off the wall.
Hear, hear.
Remember when you celebrate "philanthropy" that you're celebrating the guys who insisted that their name be splashed across everything done with their money. Real charity asks nothing in return, or at most asks for a building to be named after someone else. If I was a billionaire funding a university CS building, I'd ask for it to be named after Grace Hopper.
I know almost everyone here will totally dis this news and make fun of it. Whatever. I welcome free education for those who can't afford it. I was lucky enough to get university education, but not everyone is that lucky. Even if 20 people learn something from these Microsoft courses, and it helps them land better jobs, I will be happy.
It's a vendor-specific training course for a vendor-specific development/operational environment. Over the course of history, many enlightened salespeople have understood that free training courses (note: free "training courses", not free "education") improve brand awareness and market share. On the flipside, if you have a popular product anyway, you can make a lot of money by selling official training materials.
Microsoft are losing ground to Google, Amazon Web Services etc in the cloud computing market, so they've decided a free course is the best way to get people using their product. And they picked as their provider a company that has a list of many thousands of students, but who are themselves playing second fiddle to their competitors -- ie. Coursera and Udacity.
I do not believe in the corporate sponsorship of education. A teacher cannot be a billboard.
Not Always Necessary. Sometimes you know from the very beginning that your list will be mostly ordered, and you choose your sorting and searching algorithms accordingly. In some cases it really is pretty trivial. In general, though, I do agree with the principle of "optimise late". We just have to remember that there are no absolutes.
People reinventing the wheel always say their approach is more efficient. They forget that all the extra features they don't use (or don't understand) are just left out at compile time. So what they do is just add code bloat.
The problem with relying on compiler optimisations is that the compiler rarely knows the valid or most likely forms of input. This is why anyone concerned with performance will explicitly specify quicksort/bubblesort/whatever rather than just calling "sort" and letting the compiler choose.
I agree that some people take this too far, and basically "micromanage" their code, but then people also go too far the other way at times.
The thing is, there's a whole spectrum that you're dismissing with "reinventing the wheel". Sometimes you're not actually "reinventing", just building an appropriatel sized one for the task at hand, rather that grafting on an oversized wheel with unneccesary chrome alloy rims etc, just because it's already there. The line between these things is fuzzy, and people often get it wrong, but no system will ever be fool-proof.
There's a huge gap between "not investigating someone" and "going to the judge for a warrant before even talking to a guy who has a non-zero statistical possibility of being potentially involved." Seek the middle ground.
For instance if you find the murder weapon and it has the suspect's DNA on it, you only have direct evidence than his DNA was in fact on the knife. It is circumstantial evidence of the suspect actually committing the murder, since you don't know for a fact how the DNA got on the knife.
But that's a world apart from "someone probably (but not definitely) related to him did it, and he happens to know a few people who lived somewhere in the sort of vicinity, even though he didn't".
The problem is that it was ridiculously heavy-handed. Think about it - if they lied to him, if they got a warrant to take his DNA before he even knew the truth of what he was being investigated for, then they were clearly convinced he did it... with only tiny slivers of circumstancial evidence. It was disproportionate action. Yes, if he had been guilty, he would have been a flight risk, but that's the case with all serious crimes. No-one treats a potential suspect as a flight risk unless they are convinced its the right man, and once the investigators are convinced they've found the right man, there's a lot of history that shows they're not liable to let go quickly, even when the evidence is against them.
He was identified on circumstancial evidence then treated as a prime suspect. Look:
Detectives traveled to New Orleans in December and persuaded a magistrate judge to sign a search warrant ordering Usry to provide his DNA for comparison.
The caller identified himself as a law enforcement official in New Orleans and said he was investigating a hit-and-run in City Park.
They got a court order without ever asking the guy to help with their enquiries. They brought him in with an outright lie and forced him to give a DNA sample. That's treatment unwarranted by the evidence, whatever the judge thinks.
This isn't an abuse. This is the system working (as well as it can, though getting him in to the office on a lie is kind of skeevy). And since even the cleared suspect is considering the branches of his family tree even he's buying into the validity of the technique.
There was no concrete evidence, only circumstancial. I find it extremely worrying that he was brought in under false pretenses and that they granted a court order for the DNA sample before any request had been made to him for a voluntary submission.
Even if someone is your only suspect, he shouldn't be treated as a "prime suspect" without evidence. Usry says in the article that he was worried that they were going to push for charges even when the sample came back negative, and it was the police and FBI's treatment of him that made him worried. It wouldn't have been the first time that an overzealous law enforcement officer had got so fixated on "getting a result" that they would have pushed on, ignoring the evidence.
The biggest warning light in the article is actually an innocent-looking remark from the Sgt in the case: " “I wish it wasn’t a dead end, but it was.”" I know it's important to solve crimes, but no police officer should ever be disappointed to find out a suspect is innocent. Do you think that one-month wait was lab time? I doubt it. During all that time, some of the officers and agents involved may well have been pushing to continue, on the grounds that he probably passed through the town once, and they think it was a cousin of his, so they would be arguing that he "probably" drove his cousin up and was an accessory.
There is always the danger with these speculative searches that you end up identifying a suspect then convincing yourself he's the guy, then confirmation bias leads the officers to build a case around lots of selective evidence. In a cold case, you'd be essentially building a "loser edit" for the poor sod you've picked.
Why are people still wasting their money on this? there are plenty, more advanced embedded boards at about the same price....
The Pi has the advantage of near uiquity -- you don't have to write anything new or faff about too much with configuration, as someone has already done it for you.
The Pi should have become a reference platform for low-processor-power computing, but then they decided to go for a non-commodity part and locked out interoperability at the hardware level....
And all calculus is based on arithmetic. This doesn't mean calculus is arithmetic.
Natural cause-and-effect is kind of fuzzy and can lead to incorrect generalisations, which we call "superstitions"; but process thinking has to be rigid, complete and unambiguous.
The problem with many laws and corporate policies is that the letter and spirit are not properly aligned. People write rules that can be gamed. "You know what I mean" doesn't work for computers, and it doesn't work for big organisations.
The reason the Occulus Rift still isn't available as a consumer device is that they still have problems with lag. If you know the solution, maybe you should offer them your services as a well-paid consultant.
It's ideal for highlighting the often-missed initial concepts, such as the facts that statements are executed in order, variables can store information and change, and that certain statements can change the program flow.
Or you could use a traditional shell-scripting language (eg BASH), and then start introducing control-flow based on the output of other programs. "But wouldn't it be cool if we could have multiple programs in a single file...?" -- et voilà, everyone wants to know about procedural programming, and the notion of a master procedure (main) is a natural consequence.
Slashdot Mirror
If it wasn't for all the additives and fillers, I'd be interested in its properties as part of a dry-cure seasoning....
The most advanced "electronic noses" actually use trained insects with electrodes attached. All the power of the biological nose, but less hungry than dogs. Also quicker to train with lower volumes of scent sample required. Pure Pavlovian conditioning.
It's not that complicated: it's really hard to make a profit selling people their own mind's ability to heal them.
The placebo only works if you don't think it's a placebo. So if people believe in weird shit, yet it increase to success of the placebo effect then where's the harm?
The harm is that it discourages patients from seeking other treatments that will give both placebo benefits and genuine medical benefits.
People make money by encouraging people to believe that sugar that has touched water that has touched water that has touched water that has touched water that has touched water that has touched arsenic is more effective than a proven medical treatment. That's immoral. Your justification could be applied to any and all fraud.
Except that many homeopathic treatments are actually water that has no active ingredients dropped onto sugar pills then dried. So no, not even dehydration. Just hypoglycaemia.
Microsoft are losing ground to Google, Amazon Web Services etc in the cloud computing market, so they've decided a free course is the best way to get people using their product
Ah, the old .. market share = competence argument. I guess Linux is the single most worst desktop operating system ever built.
It's nothing to do with how good or bad Azure or its competitors are -- all I said is that Microsoft are not currently doing in a commercially strong position in the cloud marketplace, and they want to change that.
A teacher cannot be a billboard.
By that logic, nobody should be teaching mechanics how to repair a Subaru, cuz.. y know all mechanics should know how every single car works.
You people are deluded. People *want* to learn about/get trained on MS products, because the entire fucking world uses them. Like GP said, I'm glad if even a handful of people get trained and improve their life by finding better jobs. If you want to offer them something better, put your money where your mouth is.
There is a big difference between teaching and training. You teach someone about how cars work and how to repair them. You train someone in using Subaru's engine management system. (Of course, in many teaching courses, you simultaneously train in the use of particular technologies or techniques, but that is only because you need a concrete environment for practise.)
edX (and Coursera) is supposed to be a teaching platform, but there was never any money in that. Udacity claimed to be teaching, but was to all intents and purposes a training platform from day one. It was no surprise that it very quickly started offering vendor-specific sponsor (training) courses, but I'm disappointed that edX are following suit.
Teaching vs training is not just a problem in online education -- CS departments the world over are under constant pressure to train students in industry standard tools so that new employees can hit the ground running, rather than having to be sent on training courses for their employers' choice of tools. You would not teach a two-year diploma in Subaru maintenance -- the Subaru garage would expect to have to train new staff, but if they can find someone with Subaru experience, so much the better!
And the problem extends beyond universities, because in computing more than most fields, you need constant training to keep up with new technologies, and employers are loathe to do that. They expect people to come in with the training already, which means people are now expected to use their own time and money throughout their career just to keep their jobs. By supporting this, edX are actually threatening their own revenue streams, as they already have a "training" business in the form of edX Professional Education, an all-paid-for brand which they're trying to sell to employers, and the existence of free training courses on the teaching site dilutes the apparent value of these.
Then you're also in favour of demolishing the William Gates building at several universities
No, but I'd take his f#@king name off the wall.
Hear, hear.
Remember when you celebrate "philanthropy" that you're celebrating the guys who insisted that their name be splashed across everything done with their money. Real charity asks nothing in return, or at most asks for a building to be named after someone else. If I was a billionaire funding a university CS building, I'd ask for it to be named after Grace Hopper.
I know almost everyone here will totally dis this news and make fun of it. Whatever. I welcome free education for those who can't afford it. I was lucky enough to get university education, but not everyone is that lucky. Even if 20 people learn something from these Microsoft courses, and it helps them land better jobs, I will be happy.
It's a vendor-specific training course for a vendor-specific development/operational environment. Over the course of history, many enlightened salespeople have understood that free training courses (note: free "training courses", not free "education") improve brand awareness and market share. On the flipside, if you have a popular product anyway, you can make a lot of money by selling official training materials.
Microsoft are losing ground to Google, Amazon Web Services etc in the cloud computing market, so they've decided a free course is the best way to get people using their product. And they picked as their provider a company that has a list of many thousands of students, but who are themselves playing second fiddle to their competitors -- ie. Coursera and Udacity.
I do not believe in the corporate sponsorship of education. A teacher cannot be a billboard.
Not Always Necessary. Sometimes you know from the very beginning that your list will be mostly ordered, and you choose your sorting and searching algorithms accordingly. In some cases it really is pretty trivial. In general, though, I do agree with the principle of "optimise late". We just have to remember that there are no absolutes.
People reinventing the wheel always say their approach is more efficient. They forget that all the extra features they don't use (or don't understand) are just left out at compile time. So what they do is just add code bloat.
The problem with relying on compiler optimisations is that the compiler rarely knows the valid or most likely forms of input. This is why anyone concerned with performance will explicitly specify quicksort/bubblesort/whatever rather than just calling "sort" and letting the compiler choose.
I agree that some people take this too far, and basically "micromanage" their code, but then people also go too far the other way at times.
The thing is, there's a whole spectrum that you're dismissing with "reinventing the wheel". Sometimes you're not actually "reinventing", just building an appropriatel sized one for the task at hand, rather that grafting on an oversized wheel with unneccesary chrome alloy rims etc, just because it's already there. The line between these things is fuzzy, and people often get it wrong, but no system will ever be fool-proof.
There's a huge gap between "not investigating someone" and "going to the judge for a warrant before even talking to a guy who has a non-zero statistical possibility of being potentially involved." Seek the middle ground.
For instance if you find the murder weapon and it has the suspect's DNA on it, you only have direct evidence than his DNA was in fact on the knife. It is circumstantial evidence of the suspect actually committing the murder, since you don't know for a fact how the DNA got on the knife.
But that's a world apart from "someone probably (but not definitely) related to him did it, and he happens to know a few people who lived somewhere in the sort of vicinity, even though he didn't".
This is semen. On the body. You pretty much had to be there.
They shouldn't have been granted a warrant! There was no friggin evidence!
The problem is that it was ridiculously heavy-handed. Think about it - if they lied to him, if they got a warrant to take his DNA before he even knew the truth of what he was being investigated for, then they were clearly convinced he did it... with only tiny slivers of circumstancial evidence. It was disproportionate action. Yes, if he had been guilty, he would have been a flight risk, but that's the case with all serious crimes. No-one treats a potential suspect as a flight risk unless they are convinced its the right man, and once the investigators are convinced they've found the right man, there's a lot of history that shows they're not liable to let go quickly, even when the evidence is against them.
Detectives traveled to New Orleans in December and persuaded a magistrate judge to sign a search warrant ordering Usry to provide his DNA for comparison.
The caller identified himself as a law enforcement official in New Orleans and said he was investigating a hit-and-run in City Park.
They got a court order without ever asking the guy to help with their enquiries. They brought him in with an outright lie and forced him to give a DNA sample. That's treatment unwarranted by the evidence, whatever the judge thinks.
This isn't an abuse. This is the system working (as well as it can, though getting him in to the office on a lie is kind of skeevy). And since even the cleared suspect is considering the branches of his family tree even he's buying into the validity of the technique.
There was no concrete evidence, only circumstancial. I find it extremely worrying that he was brought in under false pretenses and that they granted a court order for the DNA sample before any request had been made to him for a voluntary submission.
Even if someone is your only suspect, he shouldn't be treated as a "prime suspect" without evidence. Usry says in the article that he was worried that they were going to push for charges even when the sample came back negative, and it was the police and FBI's treatment of him that made him worried. It wouldn't have been the first time that an overzealous law enforcement officer had got so fixated on "getting a result" that they would have pushed on, ignoring the evidence.
The biggest warning light in the article is actually an innocent-looking remark from the Sgt in the case: " “I wish it wasn’t a dead end, but it was.”" I know it's important to solve crimes, but no police officer should ever be disappointed to find out a suspect is innocent. Do you think that one-month wait was lab time? I doubt it. During all that time, some of the officers and agents involved may well have been pushing to continue, on the grounds that he probably passed through the town once, and they think it was a cousin of his, so they would be arguing that he "probably" drove his cousin up and was an accessory.
There is always the danger with these speculative searches that you end up identifying a suspect then convincing yourself he's the guy, then confirmation bias leads the officers to build a case around lots of selective evidence. In a cold case, you'd be essentially building a "loser edit" for the poor sod you've picked.
Would poison be considered assault or just murder?
Why are people still wasting their money on this? there are plenty, more advanced embedded boards at about the same price....
The Pi has the advantage of near uiquity -- you don't have to write anything new or faff about too much with configuration, as someone has already done it for you.
The Pi should have become a reference platform for low-processor-power computing, but then they decided to go for a non-commodity part and locked out interoperability at the hardware level....
All processes are based on cause and effect.
And all calculus is based on arithmetic. This doesn't mean calculus is arithmetic.
Natural cause-and-effect is kind of fuzzy and can lead to incorrect generalisations, which we call "superstitions"; but process thinking has to be rigid, complete and unambiguous.
It is not just about frame rate. It's about the lag between moving the head and the image changing.
The problem with many laws and corporate policies is that the letter and spirit are not properly aligned. People write rules that can be gamed. "You know what I mean" doesn't work for computers, and it doesn't work for big organisations.
The reason the Occulus Rift still isn't available as a consumer device is that they still have problems with lag. If you know the solution, maybe you should offer them your services as a well-paid consultant.
It's ideal for highlighting the often-missed initial concepts, such as the facts that statements are executed in order, variables can store information and change, and that certain statements can change the program flow.
Or you could use a traditional shell-scripting language (eg BASH), and then start introducing control-flow based on the output of other programs. "But wouldn't it be cool if we could have multiple programs in a single file...?" -- et voilà, everyone wants to know about procedural programming, and the notion of a master procedure (main) is a natural consequence.