A specific example of government incompetence literally killing people is the FDA and the AIDS epidemic of the 80s and 90s. The government 'helping' patient safety resulted in the early deaths of many thousands of patients by denying them access to life saving drugs that were yet to be approved.
By comparison, how many lives have the government saved by stopping companies from selling ineffective or flat out toxic medications?
Look, the only thing governments have consistently done well is kill and destroy.
Looks like someone needs to live for a year in the Libertarian Paradise of the Congo to see what life without a functioning government it like.
This is all fear mongering. There are reefs in the world with water FAR hotter than the great barrier reef. Despite what researchers who benefit from government funding would like you to believe, there is far from agreement that the reef is in any way in trouble.
If that's fearmongering then your post is stupidmongering.
Yes there are warmer reefs elsewhere. That doesn't mean you can magically transform the great barrier reef into a hot water reef without it first crashing badly and for a long time, and losing the huge diversity that's present there now.
I like how you layer it on with a nice chunk of anti-research and anti-government paranoia. Though it's kind of entertaining that you thing researchers do it for the money.
Simply look at the UK. Now they're talking about knife bans and one judge in the UK even suggesting people deliberately dull their kitchen knives(!). That's stupid in the extreme.
and who is "they"? A retired judge and an activist organisation. Yeah the UK is totally going to hell in a handbasket because out of the 60 million inhabitants you managed to find a few who publicly said something stupid. Good job no one ever says anything stupid in your country.
Flowers bought from the quikie mart, have a different 'sacrifice' value to flowers bought from a florist.
Supermarket flowers are genreally fine (same quality) but the range is usually quite restricted and they're only available in pre-selected bunches. If you want to customise things or want things a bit more specialist, then the florist is the way to go.
Yes, damn that Gamergate crowd for demanding that games be reviewed on their merit and not on the basis of what sexual favours the developer has offered the reviewer.
Wow the gamergate crowd is still sore. I guess that's why you have nothing to od but sit around in your mother's basements stewing over threads and spending all your modpoints shouting people down.
so tell me, Mr Ac and the cowardly mods who omdded this up where is the slightest shred of evidence that Anita Sareesian (who I singled out in my post) gave any sexual favours to anyone?
And yes, you should be moderated to -1 for your comment that misrepresents Gamergate.
Ah yes the old -1 truth mod. I know it well.
You apparently don't even know what it was about and seem to believe it was about Anita Sarkeesian. It's clear that you have preconceived ideas about Gamergate and have decided that gamers are a bunch of misogynists without even investigating what happened.
I watched the whole thing unfold at the time. they are a bunch of rampant myisogynists.
They aren't the ones disposing of their products, so it seems reasonable that they aren't responsible for consumers who dispose of them inappropriately.
If you are drowning in plastic garbage, talk to your neighbors. They're the ones disposing of things incorrectly.
Not really, no. The manufacturers are selling this stuff to humans. We know how humans behave. Wishing that humans behaved in some sort of ideal way isn't going to fix anything. Wishing that people simply human better than they do is the ultimate failing of both libertarianism and communism.
Given your moderation score, it looks like the gamergate crowd is still going strong here. I, too, fully expect to hit -1 for trolling, becaue there's nothing quite so trollish as demanding evidence that someone accused of doing all sorts of bad things (e.g. Anita Sarkeesian) actually did them!
When you have two wildly different approaches (human jury and SVM) produce nearly the same results and the same "unfairness" I feel rather safe taking as a working hypothesis that it is perceptional and actually a result of the underlying statistics when you purposely try to ignore race.
The link you posted demonstrates that COMPASS is a complete shitshow. It's no more accurate than lay people with no expertise in criminal justice.
For example: I need to fly RTF now, but all the flights are full. No problem, they book me to London, then cancel the international leg. Warn me to get to the gate early, so I'm not the one to get bumped.
Ha! Nice.
The rise of these systems is ultimately because of poor accounting practices. Someone (presumably) does the math and sees how much $ is spent on the travel agent and how much they can save in their department by switching to an automatic system.
The thing is becaue of poor practices, their department is never charged for the extra time that everyone else starts to incur. You know an hour or two here and there per trip wragling the travel system (as opposed to 5 minutes emailng the agent), then the next three hours whining about it because it's so fucking annoying, and the lost productivity ot the extra stress etc. And of course the extra employee time spent tooling around at airports because they can't use the travel agent's neat tricks any more and so on and so forth.
but you refuse to mention what an unbiased algorithm and it's result would look like.
Right, so because I, like the entire rest of the ML community don't know how to go beyond the current state of the art we should just not bother trying to correct flaws.
. Compass is close to the best you are going to get
You don't know that, because you don't know what algorithm it uses.
without affirmative action
Thi is the first time I've heard that not cracking down on black people merely because they're black called "affermative action".
f the algorithm is unfair, it's because life is unfair,
You know what seems remarkable to me: most times there's an article on software, people are quick to jump all over the flaws of it ESPECIALLY if it's AI/ML because a lot of us are software people and know how crap the average piece of software is, how GIGO works and so on and so forth.
But when someone points out that some software is a bit crap (and disfavours black people) people are juping all over it to say how perfect it is.
You don't have to be rich to get married, nearly three fucking quarters of black kids are born to an unmarried mother. If you think that won't have impact on criminal behavior you're dreaming.
I think you've just demonstrated the point of the article: that's a non causitive correlation. The underlying cause is the lack of a stable family. That commonly manifests as not being married, but not being married is the symptom not the cause. It's perfectly possible to have a stable family without marriage and more and more couples are choosing to not marry.
Blaming it all on systemic racism and poverty is silly.
You're trying to refute my argument by reading a more extreme one than the one I wrote, then refuting that instead. I didn't blame it *all*, just a large amount, and that isn't silly.
Regardless, any difference in recidivism rate will cause the imbalances seen in the Compas result.
Yes. No one's denying that.
ick your metric (false negative rate for instance) and pretty much the only way to get equal outcomes
Again you're inventing a pint of view of mine and refuting that. Basically whenever I see someone banging on about "equal outcomes" I know you didn't read what I wrote, you read what you belive I would have written.
Now I know you're not reading what I wrote, merely arguing against what you believe to be some sort of generic liberal position there's little point in continuing to discuss this further.
It makes no sense to single out 'race' as a problem, when there are hundreds of other non-causative correlations that are equally problematic.
Sure it makes sense. That's not to say the other non causitive correlations are not equally problematic---they are---but that doesn't mean that it makes no sense to single out race as one.
The reason for that is that the race one is simple, easy to understand and people are hopefully goig to think twice before trying to argue "oh well maybe black people are more criminal" or some equivalent.
Statistics and machine learning is hard, techincal and deeply mathematical. It's not easy to explain t opeople in the general case. But if you can show people an answer that they more intuitively understand and know that it's giving the wrong answer, it's much less easy to fall back on the sort of "algorithms are magic" kind of thinking that pervades opinions on computers in general and machine learning in particular.
So by singling out race rather than talking about general cases of bias variance tradeoffs, non causitive correlations and so on you might actually be able to get lay people to understand enough about the problem to effect some sort of change.
The end goal would be perfect algorithms. Don't discount a decent step towards that just because it doesn't go all the way in one go. If you do, not only will you never get there, you'll never get closer than you are now.
[1] This post is not a vote of confidence in SAP but a vote of no confidence in the IBM solution we used previously.
Aaah IBM: the living embodyment of "no matter how bad you think it is, it can always be worse".
Way back we used to have this system called "agent" or "travel agent" or something. It was amazing, you sent it an email and it must have had some mad-ass NLP or something since it parsed out your requirements and emailed back you an itenerary.
Remember, Citizen: Equality means including an equal
No, citizen, equality means not giving you a harsher conviction simply because people who look like you have been convicted in the past. What I don't really get is why you'e against true equality.
I'd want a justice system that doesn't consider the race or skin color in the verdict. That doesn't mean there won't be any correlations though.
Well then it's kind of a shame that machine learning algorithms are good at picking out non causitive correlations! If only some researchers made a tool to help find those...
. In the end comparing blacks and whites is apples and oranges. Blacks recidivism rates is fundamentally higher than whites
It's not fundamantally higher. It's higher for two reasons, one is socioeconomic (poverty is higher on average) and the other is simple racism (the justice system is harsher on black people than white).
. You could arbitrarily force the false positive or negative rate to be equal by making race an input and using affirmative action, but that would degrade fairness in other ways.
t's not in any way fair to bake existing structural racism into the algorithm because that's the way things currently are.
Slashdot Mirror
We just did that.
I guess you could say that. I mean sure he's not the same one, he's an even bigger one. But at least it's a change...?
A specific example of government incompetence literally killing people is the FDA and the AIDS epidemic of the 80s and 90s. The government 'helping' patient safety resulted in the early deaths of many thousands of patients by denying them access to life saving drugs that were yet to be approved.
By comparison, how many lives have the government saved by stopping companies from selling ineffective or flat out toxic medications?
Look, the only thing governments have consistently done well is kill and destroy.
Looks like someone needs to live for a year in the Libertarian Paradise of the Congo to see what life without a functioning government it like.
Open your eyes and question everything you mindness robot.
And this is why we still have flat-earthers.
This is all fear mongering. There are reefs in the world with water FAR hotter than the great barrier reef. Despite what researchers who benefit from government funding would like you to believe, there is far from agreement that the reef is in any way in trouble.
If that's fearmongering then your post is stupidmongering.
Yes there are warmer reefs elsewhere. That doesn't mean you can magically transform the great barrier reef into a hot water reef without it first crashing badly and for a long time, and losing the huge diversity that's present there now.
I like how you layer it on with a nice chunk of anti-research and anti-government paranoia. Though it's kind of entertaining that you thing researchers do it for the money.
Simply look at the UK. Now they're talking about knife bans and one judge in the UK even suggesting people deliberately dull their kitchen knives(!). That's stupid in the extreme.
and who is "they"? A retired judge and an activist organisation. Yeah the UK is totally going to hell in a handbasket because out of the 60 million inhabitants you managed to find a few who publicly said something stupid. Good job no one ever says anything stupid in your country.
You're an ex-pat, right?
I think the point is that you can just choose not to buy them without any loss of quality of life.
Then it's an incredibly stupid point. Hey diamond mining slaves? Think deBeers is evil? Don't buy their diamonds! Problem solved!
Flowers bought from the quikie mart, have a different 'sacrifice' value to flowers bought from a florist.
Supermarket flowers are genreally fine (same quality) but the range is usually quite restricted and they're only available in pre-selected bunches. If you want to customise things or want things a bit more specialist, then the florist is the way to go.
Well, to be fair, it's not like people are starving because they can't afford diamonds.
Are you being contrarian for the sake of it or are you really just unaware of what deBeers have done in pursuit of their profits?
And in how many hotels could you use it yet ?
Hotels? idon't know but it's not an Apple innovation. My local hackspace has had NFC based door opening for ages. Works with phone NFC.
Yes, damn that Gamergate crowd for demanding that games be reviewed on their merit and not on the basis of what sexual favours the developer has offered the reviewer.
Wow the gamergate crowd is still sore. I guess that's why you have nothing to od but sit around in your mother's basements stewing over threads and spending all your modpoints shouting people down.
so tell me, Mr Ac and the cowardly mods who omdded this up where is the slightest shred of evidence that Anita Sareesian (who I singled out in my post) gave any sexual favours to anyone?
And yes, you should be moderated to -1 for your comment that misrepresents Gamergate.
Ah yes the old -1 truth mod. I know it well.
You apparently don't even know what it was about and seem to believe it was about Anita Sarkeesian. It's clear that you have preconceived ideas about Gamergate and have decided that gamers are a bunch of misogynists without even investigating what happened.
I watched the whole thing unfold at the time. they are a bunch of rampant myisogynists.
So, you are saying that anyone who supplies something to someone else should be responsible for their use of it 'because'?
So you're saying we should adopt your model where no one is responsible for anything especially if they're turning a profit?
Did you say that? Who cares, if you're going to take extreme misreadings of my posts, why shouldn't I do the same?
4) Be very suspicious of anyone pushing these bans. They are likely virtue signalling
Be very suspicious f anyone using the phrase "virtue signalling", they are likely a fuckwit with an axe to grind.
They aren't the ones disposing of their products, so it seems reasonable that they aren't responsible for consumers who dispose of them inappropriately.
If you are drowning in plastic garbage, talk to your neighbors. They're the ones disposing of things incorrectly.
Not really, no. The manufacturers are selling this stuff to humans. We know how humans behave. Wishing that humans behaved in some sort of ideal way isn't going to fix anything. Wishing that people simply human better than they do is the ultimate failing of both libertarianism and communism.
Sure it's for safety. They just trust passengers in business more, so no safety issue with giving them metal utensils
Well of course. Nobody with money ever did anything bad.
Given your moderation score, it looks like the gamergate crowd is still going strong here. I, too, fully expect to hit -1 for trolling, becaue there's nothing quite so trollish as demanding evidence that someone accused of doing all sorts of bad things (e.g. Anita Sarkeesian) actually did them!
When you have two wildly different approaches (human jury and SVM) produce nearly the same results and the same "unfairness" I feel rather safe taking as a working hypothesis that it is perceptional and actually a result of the underlying statistics when you purposely try to ignore race.
The link you posted demonstrates that COMPASS is a complete shitshow. It's no more accurate than lay people with no expertise in criminal justice.
For example: I need to fly RTF now, but all the flights are full. No problem, they book me to London, then cancel the international leg. Warn me to get to the gate early, so I'm not the one to get bumped.
Ha! Nice.
The rise of these systems is ultimately because of poor accounting practices. Someone (presumably) does the math and sees how much $ is spent on the travel agent and how much they can save in their department by switching to an automatic system.
The thing is becaue of poor practices, their department is never charged for the extra time that everyone else starts to incur. You know an hour or two here and there per trip wragling the travel system (as opposed to 5 minutes emailng the agent), then the next three hours whining about it because it's so fucking annoying, and the lost productivity ot the extra stress etc. And of course the extra employee time spent tooling around at airports because they can't use the travel agent's neat tricks any more and so on and so forth.
but you refuse to mention what an unbiased algorithm and it's result would look like.
Right, so because I, like the entire rest of the ML community don't know how to go beyond the current state of the art we should just not bother trying to correct flaws.
. Compass is close to the best you are going to get
You don't know that, because you don't know what algorithm it uses.
without affirmative action
Thi is the first time I've heard that not cracking down on black people merely because they're black called "affermative action".
f the algorithm is unfair, it's because life is unfair,
You know what seems remarkable to me: most times there's an article on software, people are quick to jump all over the flaws of it ESPECIALLY if it's AI/ML because a lot of us are software people and know how crap the average piece of software is, how GIGO works and so on and so forth.
But when someone points out that some software is a bit crap (and disfavours black people) people are juping all over it to say how perfect it is.
Wow, just wow.
You don't have to be rich to get married, nearly three fucking quarters of black kids are born to an unmarried mother. If you think that won't have impact on criminal behavior you're dreaming.
I think you've just demonstrated the point of the article: that's a non causitive correlation. The underlying cause is the lack of a stable family. That commonly manifests as not being married, but not being married is the symptom not the cause. It's perfectly possible to have a stable family without marriage and more and more couples are choosing to not marry.
Blaming it all on systemic racism and poverty is silly.
You're trying to refute my argument by reading a more extreme one than the one I wrote, then refuting that instead. I didn't blame it *all*, just a large amount, and that isn't silly.
Regardless, any difference in recidivism rate will cause the imbalances seen in the Compas result.
Yes. No one's denying that.
ick your metric (false negative rate for instance) and pretty much the only way to get equal outcomes
Again you're inventing a pint of view of mine and refuting that. Basically whenever I see someone banging on about "equal outcomes" I know you didn't read what I wrote, you read what you belive I would have written.
Now I know you're not reading what I wrote, merely arguing against what you believe to be some sort of generic liberal position there's little point in continuing to discuss this further.
It makes no sense to single out 'race' as a problem, when there are hundreds of other non-causative correlations that are equally problematic.
Sure it makes sense. That's not to say the other non causitive correlations are not equally problematic---they are---but that doesn't mean that it makes no sense to single out race as one.
The reason for that is that the race one is simple, easy to understand and people are hopefully goig to think twice before trying to argue "oh well maybe black people are more criminal" or some equivalent.
Statistics and machine learning is hard, techincal and deeply mathematical. It's not easy to explain t opeople in the general case. But if you can show people an answer that they more intuitively understand and know that it's giving the wrong answer, it's much less easy to fall back on the sort of "algorithms are magic" kind of thinking that pervades opinions on computers in general and machine learning in particular.
So by singling out race rather than talking about general cases of bias variance tradeoffs, non causitive correlations and so on you might actually be able to get lay people to understand enough about the problem to effect some sort of change.
The end goal would be perfect algorithms. Don't discount a decent step towards that just because it doesn't go all the way in one go. If you do, not only will you never get there, you'll never get closer than you are now.
[1] This post is not a vote of confidence in SAP but a vote of no confidence in the IBM solution we used previously.
Aaah IBM: the living embodyment of "no matter how bad you think it is, it can always be worse".
Way back we used to have this system called "agent" or "travel agent" or something. It was amazing, you sent it an email and it must have had some mad-ass NLP or something since it parsed out your requirements and emailed back you an itenerary.
If there's structural racism, that needs to be fixed, and then the algorithm will follow automatically.
It will only follow if the algorithm is re-trained.
At the oment, the algorithm trained with biased data is part of the problem.
Remember, Citizen: Equality means including an equal
No, citizen, equality means not giving you a harsher conviction simply because people who look like you have been convicted in the past. What I don't really get is why you'e against true equality.
I'd want a justice system that doesn't consider the race or skin color in the verdict. That doesn't mean there won't be any correlations though.
Well then it's kind of a shame that machine learning algorithms are good at picking out non causitive correlations! If only some researchers made a tool to help find those...
. In the end comparing blacks and whites is apples and oranges. Blacks recidivism rates is fundamentally higher than whites
It's not fundamantally higher. It's higher for two reasons, one is socioeconomic (poverty is higher on average) and the other is simple racism (the justice system is harsher on black people than white).
. You could arbitrarily force the false positive or negative rate to be equal by making race an input and using affirmative action, but that would degrade fairness in other ways.
t's not in any way fair to bake existing structural racism into the algorithm because that's the way things currently are.