Exactly. The bias is not in the machine, it's in our interpretations of the results
No, the bias is that when you compare the predictions of the algorithm with the actual results, the algorithm predicted that blacks will offend much more than the data shows that they actually do, and predicted white will offend much less often than the data shows that they do.
This is what we mean by bias: the predictions vary from the actual data in a way that is not random, but is biased.
The problem is not that the data set reflects the reality. The problem is not that the AI makes mistakes, but that the particular mistakes the AI makes reflect the bias of the society that programmed it.
The link in the summary is to an article which is itself a summary. From the original (here: Machine Bias
There’s software used across the country to predict future criminals. And it’s biased against blacks.), the software attempted to predict the probability of future offenses of criminals on probation. It did not, of course, always get it right. But when the actual percentage of re-offenses was compared to the predictions, the AI got it wrong differently for blacks than for whites.
Here's what the article said.
We also turned up significant racial disparities, just as Holder feared. In forecasting who would re-offend, the algorithm made mistakes with black and white defendants at roughly the same rate but in very different ways.
The formula was particularly likely to falsely flag black defendants as future criminals, wrongly labeling them this way at almost twice the rate as white defendants.
White defendants were mislabeled as low risk more often than black defendants.
What are they calling "bias"? We read constantly about so-called racism based merely on the fact that one race objectively exhibits a particular trait over other races. That's called data, not bias.
It's a tricky question. Just because something is data, does not mean that it isn't biased: data can be biased-- in fact, 90% of what we do in experimental science is understanding the bias in data and figuring out how to get an unbiased measurement out of a biased data set. Almost all data is biased one way or another.
If, for example, white people caught shoplifting are usually given a warning and let off while black people caught shoplifting are arrested and prosecuted ("shopping while black"), the data will show a higher rate of shoplifting among blacks. You will need to go to the raw data to see the actuality. See: https://www.theguardian.com/la...
An AI with no correction for bias will reflect the bias of society.
Tell you what, why don't you learn something about the history of the development of integrated circuits, and get back to me when you actually know something.
The "shortness of summer" is bounded by temperature, not sunlight Wrong. This will happen earlier, and it will stay thawed later. How many days? 1or 2?
Before there is not enough sun, it won't thaw. Plain and simple. And that is depending on the length of the day, not on a magic temperature that comes from somewhere. (where should it come from? Hu? Polar night is polar night and everything is frozen... )
Tell you what, why don't you do some research here and get back to me when you've learned enough to form an opinion
Why should NASA get any credit for inventing something when they were merely a customer?
The answer is that they didn't get credit for inventing it. But "who gets credit for inventing the IC" wasn't the question posed. The question was about the return on investment of NASA funding. NASA (along with the Air Force) funded the research needed to turn the concept of an integrated circuit into an actual product. The answer to the question of what was the return on investment is that this particular investment, in developing the IC from a concept to a commercial reality, has a very large return.
The return on investment, however, did not return to NASA. It returned to everybody who uses computers, or cell phones, or any electronic device using integrated circuits.
p>NASA farted around with the idea of reusable rockets for over thirty years and got nowhere past an outrageously expensive and downright dangerous space shuttle and said "good enough" and called it a day -- for DECADES. No wonder companies like SpaceX are coming in and applying a little bit of modesty to their designs and in consequence are running circles around NASA,
You do have to keep in mind that SpaceX blew up their first three rockets in a row. At the time that NASA picked them to develop the Falcon-9, nobody else in the world had any faith that they would be anything other than a marginal company that would build a small capacity rocket to put a small payload in low orbit cheaply but with questionable unreliability.
For all practical purposes, the partnership between NASA and SpaceX is the very reason SpaceX even exists.
NASA's problem is not the bureaucracy, per se: it is the fact that NASA does everything in public, and the public does not allow NASA to fail. If you aren't allowed to fail, then it is very difficult to make large advances.
NASA would never have been allowed to continue a project that had three very public failures in a row. The first public failure would have produced congressional hearings, and the second would have cancelled the program with prejudice, along with sarcastic editorial cartoons in every newspaper in the world.
The reason that SpaceX can succeed is specifically because they are allowed to fail.
Considering NASA costs next to nothing (about 0.5% of the US govt's total budget), and the studies I've seen referenced show its return on investment to be about $10 for every $1 used (granted, it's a difficult figure to calculate, but even if assuming a huge error margin that's still great ROI), it's no wonder you chose to post that anonymously.
If that ROI were true, then NASA should be self-funded by now.
The integrated circuit was developed by two government programs: NASA's Apollo computer, and the U.S. Air Force's Minuteman guidance. While the IC had been invented by Noyce and Kilby, nobody in particularly had a use for it-- discrete parts already did fine, why put more than one component on a chip?-- except for NASA and the Air Force, who needed to develop lightweight computers, and funded the development of IC chips specifically for lightweight computers.
So, yes, if NASA had been able to take a cut from half the profits of the IC industry, and thus the computer technology that they (co-) developed, yes, the space program would be self-funded.
An AI will be terrible at playing soccer, because you can put 11 computers on a soccer field running whatever program you want, and no matter how good the AI they are running, they will just sit there.
WTF is "an" AI?
An AI is a program, running on a computer, that simulates intelligence and/or solves problems in the way an intelligence solves them.
.... I don't know what "an" AI is capable of because I don't know what "an" AI is....
Vegetation, harvest etc. ist mostly bound by the shortness of the summer, not by temperature.
No. The "shortness of summer" is bounded by temperature, not sunlight-- basically, when the thaws occurs, and temperatures are above freezing overnight. This will happen earlier, and it will stay thawed later.
As for sunlight, after the spring equinox Russia gets more sunlight than more equatorial places, not less.
Best models suggest that the forcing due to greenhouse gasses that we have already injected into the atmosphere is larger than the forcing due to the Milankovitch variations that cause the ice age glaciations.
But the periodicity for glacial advance and retreat is 100,000 years-- the hundred year time scale you mention is a bit fast. A new glaciation isn't coming that soon.
As the glaciers melt and sea levels rise, coastal cities around the world will be flooded, and arctic wastelands will spring to life.Russia is an arctic nation, which will benefit greatly from global sea levels. Most of America's business and culture is near the sea, and will be devastated by sea level rise.
Not sure why this is rated "troll". Actually, global warming probably will be good for Russia. At least, will be good for large parts of Russia.
Global warming isn't bad for everybody. It will be bad for places close to the equator, and places with cities on or near the ocean. But that's not true of most of Russia.
The comment seem rather accurate.
(the second part, sarcasm about Putin and Trump, is not, however.)
An AI will be terrible at playing soccer, because you can put 11 computers on a soccer field running whatever program you want, and no matter how good the AI they are running, they will just sit there. Soccer is a physical game, played on a physical playing field. The human players will dribble around the computers sitting on the field and kick the ball in the net.
Now, turn those AIs into robots and you may have something.
But what the article is talking about is not AIs that can play soccer, it is AIs that can play a soccer-themed online video game.
Your calling them "suspects" means that they are suspected of something.
My point is that this is done to whomever they like, they don't have to be even a suspect.
No, your point was
"Who said anything about criminals?"
An interesting article, but I draw away from it a far different conclusion than what the superficial/. summary suggests. The key is, as the examples in the article explicitly show, how the story is told (and in what forum.)
Apparently lots of people knew about the toxic environment at Uber. Why, then, did Susan Fowler's blog post end up being the one that set off the dynamite?
"It almost sounds naive, but it's clear she knows what's happening, and how in this high stakes game of poker, she has to be the coolest player at the table, lest she, like so many women before her, be labeled some hysteric. Her post is a masterpiece of tone and rhetorical control, and it had to be. No resorting to snark or irony or any number of tricks of the clever; she bore her own witness, and no better witness could an attorney have imagined."
This is for tracking criminals' location without using GPS, which is information that isn't already stored on a target computer.
Who said anything about criminals?
The post you are replying to.
As the post prior to yours attempted to point out using sarcasm, the use of the word "criminals" has already rendered judgement on the people being tracked: they're not suspects, they're "criminals".
The NRA. It pooped its pants right there in the public square. And rather than trying to clean up, it just stands there yelling "MY SHIT DON'T STINK!" while continuing to make squeaky farts..
This is probably go to a new school next year level public humiliation, but they apparently have no shame.
If you should see someone who works for the NRA, hand them a roll of toilet paper.
Exactly. The problem here is that people are trying to apply pre- information age thinking to post- information age constructs. This idea that you can build a cyber "weapon" that can only attack "bad" people and cannot be trivially altered to ignore whatever protections you put into place to keep it from being used against "good" people, is ludicrous.
Yes, exactly like guns. It's ludicrous to think you can proliferate millions of guns to "good" people, and they won't be also used by "bad" people.
Slashdot Mirror
Watch out guys, working pattern recognition is recognizing patterns!
But the patterns it is recognizing turned out to be ones inserted by the programmers, not ones that actually existed in the data.
Exactly. The bias is not in the machine, it's in our interpretations of the results
No, the bias is that when you compare the predictions of the algorithm with the actual results, the algorithm predicted that blacks will offend much more than the data shows that they actually do, and predicted white will offend much less often than the data shows that they do.
This is what we mean by bias: the predictions vary from the actual data in a way that is not random, but is biased.
The article being discussed is here, by the way: https://www.propublica.org/art...
The problem is not that the data set reflects the reality. The problem is not that the AI makes mistakes, but that the particular mistakes the AI makes reflect the bias of the society that programmed it.
The link in the summary is to an article which is itself a summary. From the original (here: Machine Bias There’s software used across the country to predict future criminals. And it’s biased against blacks.), the software attempted to predict the probability of future offenses of criminals on probation. It did not, of course, always get it right. But when the actual percentage of re-offenses was compared to the predictions, the AI got it wrong differently for blacks than for whites. Here's what the article said.
We also turned up significant racial disparities, just as Holder feared. In forecasting who would re-offend, the algorithm made mistakes with black and white defendants at roughly the same rate but in very different ways.
The formula was particularly likely to falsely flag black defendants as future criminals, wrongly labeling them this way at almost twice the rate as white defendants. White defendants were mislabeled as low risk more often than black defendants.
What are they calling "bias"? We read constantly about so-called racism based merely on the fact that one race objectively exhibits a particular trait over other races. That's called data, not bias.
It's a tricky question. Just because something is data, does not mean that it isn't biased: data can be biased-- in fact, 90% of what we do in experimental science is understanding the bias in data and figuring out how to get an unbiased measurement out of a biased data set. Almost all data is biased one way or another.
If, for example, white people caught shoplifting are usually given a warning and let off while black people caught shoplifting are arrested and prosecuted ("shopping while black"), the data will show a higher rate of shoplifting among blacks. You will need to go to the raw data to see the actuality. See: https://www.theguardian.com/la...
An AI with no correction for bias will reflect the bias of society.
The article linked is merely a summery of the propublica article, which is has more detail, here: https://www.propublica.org/art...
bye.
The "shortness of summer" is bounded by temperature, not sunlight
Wrong.
This will happen earlier, and it will stay thawed later.
How many days? 1or 2?
The northern hemisphere spring thaw has been advancing by about one day per year since 1988:
- https://www.nasa.gov/home/hqne...
- http://www.foodnutritionscienc...
- http://climatechange.lta.org/c...
- http://flatheadcore.org/featur...
- https://earthobservatory.nasa....
- http://www.sciencedirect.com/s...
- http://onlinelibrary.wiley.com...
Before there is not enough sun, it won't thaw. Plain and simple. And that is depending on the length of the day, not on a magic temperature that comes from somewhere. (where should it come from? Hu? Polar night is polar night and everything is frozen ... )
Tell you what, why don't you do some research here and get back to me when you've learned enough to form an opinion
While the IC had been invented by Noyce and Kilby
Why should NASA get any credit for inventing something when they were merely a customer?
The answer is that they didn't get credit for inventing it. But "who gets credit for inventing the IC" wasn't the question posed. The question was about the return on investment of NASA funding. NASA (along with the Air Force) funded the research needed to turn the concept of an integrated circuit into an actual product. The answer to the question of what was the return on investment is that this particular investment, in developing the IC from a concept to a commercial reality, has a very large return.
The return on investment, however, did not return to NASA. It returned to everybody who uses computers, or cell phones, or any electronic device using integrated circuits.
Uh, the government didn't fund them, the companies Texas Instruments and Fairchild Semiconductor did.
And NASA funded TI and Fairchild. (Also Raytheon and MIT).
They didn't develop it on their own money.
Do you happen to be aware of the issues regarding permafrost in the Arctic?
There are so many issues regarding permafrost that you'll have to be more specific here. Which of the many, many issues are you referring to?
p>NASA farted around with the idea of reusable rockets for over thirty years and got nowhere past an outrageously expensive and downright dangerous space shuttle and said "good enough" and called it a day -- for DECADES. No wonder companies like SpaceX are coming in and applying a little bit of modesty to their designs and in consequence are running circles around NASA,
You do have to keep in mind that SpaceX blew up their first three rockets in a row. At the time that NASA picked them to develop the Falcon-9, nobody else in the world had any faith that they would be anything other than a marginal company that would build a small capacity rocket to put a small payload in low orbit cheaply but with questionable unreliability.
For all practical purposes, the partnership between NASA and SpaceX is the very reason SpaceX even exists.
NASA's problem is not the bureaucracy, per se: it is the fact that NASA does everything in public, and the public does not allow NASA to fail. If you aren't allowed to fail, then it is very difficult to make large advances.
NASA would never have been allowed to continue a project that had three very public failures in a row. The first public failure would have produced congressional hearings, and the second would have cancelled the program with prejudice, along with sarcastic editorial cartoons in every newspaper in the world.
The reason that SpaceX can succeed is specifically because they are allowed to fail.
Considering NASA costs next to nothing (about 0.5% of the US govt's total budget), and the studies I've seen referenced show its return on investment to be about $10 for every $1 used (granted, it's a difficult figure to calculate, but even if assuming a huge error margin that's still great ROI), it's no wonder you chose to post that anonymously.
If that ROI were true, then NASA should be self-funded by now.
The integrated circuit was developed by two government programs: NASA's Apollo computer, and the U.S. Air Force's Minuteman guidance. While the IC had been invented by Noyce and Kilby, nobody in particularly had a use for it-- discrete parts already did fine, why put more than one component on a chip?-- except for NASA and the Air Force, who needed to develop lightweight computers, and funded the development of IC chips specifically for lightweight computers.
So, yes, if NASA had been able to take a cut from half the profits of the IC industry, and thus the computer technology that they (co-) developed, yes, the space program would be self-funded.
Do they have a GoFundMe page for their next picture-taking trip?
I'd contribute to that one!
"LED diode" ? What's that ?
It's like a Led Zeppelin, but with a diode.
An AI will be terrible at playing soccer, because you can put 11 computers on a soccer field running whatever program you want, and no matter how good the AI they are running, they will just sit there.
WTF is "an" AI?
An AI is a program, running on a computer, that simulates intelligence and/or solves problems in the way an intelligence solves them.
.... I don't know what "an" AI is capable of because I don't know what "an" AI is....
Now you do.
You're welcome.
Vegetation, harvest etc. ist mostly bound by the shortness of the summer, not by temperature.
No. The "shortness of summer" is bounded by temperature, not sunlight-- basically, when the thaws occurs, and temperatures are above freezing overnight. This will happen earlier, and it will stay thawed later.
As for sunlight, after the spring equinox Russia gets more sunlight than more equatorial places, not less.
Don't forget about Antarctica - it will become an earthly paradise as global warming takes effect.
Probably not. Average temperature of Antarctica is -70F. Average high is -49.
It will take more than the few degrees of warming we can produce with greenhouse effect gasses to make that "an earthly paradise."
But the periodicity for glacial advance and retreat is 100,000 years-- the hundred year time scale you mention is a bit fast. A new glaciation isn't coming that soon.
As the glaciers melt and sea levels rise, coastal cities around the world will be flooded, and arctic wastelands will spring to life.Russia is an arctic nation, which will benefit greatly from global sea levels. Most of America's business and culture is near the sea, and will be devastated by sea level rise.
Not sure why this is rated "troll". Actually, global warming probably will be good for Russia. At least, will be good for large parts of Russia.
Global warming isn't bad for everybody. It will be bad for places close to the equator, and places with cities on or near the ocean. But that's not true of most of Russia.
The comment seem rather accurate.
(the second part, sarcasm about Putin and Trump, is not, however.)
Now, turn those AIs into robots and you may have something.
But what the article is talking about is not AIs that can play soccer, it is AIs that can play a soccer-themed online video game.
Your calling them "suspects" means that they are suspected of something. My point is that this is done to whomever they like, they don't have to be even a suspect.
No, your point was "Who said anything about criminals?"
Your point did not mention the word "suspect".
Do you really want to know how deep the rabbit hole goes?
Yes, in fact I do.
I don't want claims and allegations. I want verifiable facts and details.
Apparently lots of people knew about the toxic environment at Uber. Why, then, did Susan Fowler's blog post end up being the one that set off the dynamite?
"It almost sounds naive, but it's clear she knows what's happening, and how in this high stakes game of poker, she has to be the coolest player at the table, lest she, like so many women before her, be labeled some hysteric. Her post is a masterpiece of tone and rhetorical control, and it had to be. No resorting to snark or irony or any number of tricks of the clever; she bore her own witness, and no better witness could an attorney have imagined."
This is for tracking criminals' location without using GPS, which is information that isn't already stored on a target computer.
Who said anything about criminals?
The post you are replying to.
As the post prior to yours attempted to point out using sarcasm, the use of the word "criminals" has already rendered judgement on the people being tracked: they're not suspects, they're "criminals".
The NRA. It pooped its pants right there in the public square. And rather than trying to clean up, it just stands there yelling "MY SHIT DON'T STINK!" while continuing to make squeaky farts..
This is probably go to a new school next year level public humiliation, but they apparently have no shame.
If you should see someone who works for the NRA, hand them a roll of toilet paper.
Exactly. The problem here is that people are trying to apply pre- information age thinking to post- information age constructs. This idea that you can build a cyber "weapon" that can only attack "bad" people and cannot be trivially altered to ignore whatever protections you put into place to keep it from being used against "good" people, is ludicrous.
Yes, exactly like guns. It's ludicrous to think you can proliferate millions of guns to "good" people, and they won't be also used by "bad" people.