SoylentNews is people
SoylentNews
Submitted via IRC for TheMightyBuzzard about a story that appeared on CNN on September 9, 2016.
From targeted advertising and insurance to education and policing, Cathy O'Neil's new book 'Weapons of Math Destruction' [WMD] looks at how algorithms and big data are targeting the poor, reinforcing racism and amplifying inequality.
[...] In a vacuum, these models are bad enough, but O'Neil emphasizes, "they're feeding on each other." Education, job prospects, debt and incarceration are all connected, and the way big data is used makes them more inclined to stay that way.
"Poor people are more likely to have bad credit and live in high-crime neighborhoods, surrounded by other poor people," she writes. "Once ... WMDs digest that data, it showers them with subprime loans or for-profit schools. It sends more police to arrest them and when they're convicted it sentences them to longer terms."
In turn, a new set of WMDs uses this data to charge higher rates for mortgages, loans and insurance.
[...] "Big Data processes codify the past," O'Neil writes. "They do not invent the future. Doing that requires moral imagination, and that's something only humans can provide."
I'm not interested in the story. I'm interested in what it says about once proud CNN's current quality of journalism. Fox News: Left Division?
Source: http://money.cnn.com/2016/09/06/technology/weapons-of-math-destruction/
(Score: 1, Insightful) by Anonymous Coward on Monday October 03 2016, @02:13AM
You see this kind of shit all the time these days. Lots of whining, very few solutions.
What do you want - for banks to give money to people with bad credit? For colleges to accept more idiots with bad grades? For businesses to hire people without any qualifications?
Math isn't racist. Math is math. What you do with the answers you get is up to you. Obviously most decision-makers aren't going to sabotage themselves by hiring based on political correctness instead of merit. Smart people understand that. The exception is our government where simply hanging around the longest gets you promoted and it's not your own money you're spending.
(Score: 3, Insightful) by SanityCheck on Monday October 03 2016, @02:30AM
Colleges do accept more people with shit grades because of their color, including medical colleges. Ask me what color my doctor is. And it's not because I'm racist, it's because I don't want to die.
(Score: 2) by butthurt on Monday October 03 2016, @03:33AM
Just a thought--what if medical colleges were to implement some sort of post-acceptance evaluation system, such that not all their students would automatically become medical doctors?
(Score: 4, Informative) by Zz9zZ on Monday October 03 2016, @04:01AM
There are a LOT of barriers in front of anyone who wants to become a doctor. I don't think playing the race card due to affirmative action is an accurate portrayal of reality. Even if a person gets into a school because they are a minority doesn't mean they are a worse doctor. They still have to pass med-school and there is nothing in affirmative action that makes a professor be lenient on a minority student.
~Tilting at windmills~
(Score: 2) by butthurt on Monday October 03 2016, @06:36AM
They still have to pass med-school [...]
That is what I was hinting at. Too oblique?
(Score: 0) by Anonymous Coward on Monday October 03 2016, @04:09PM
Ya, for those of us not knowing enough info off the cuff during an Interneconversation, yes always include a direct reference. Too easy to write off comments.
(Score: 3, Funny) by jmorris on Monday October 03 2016, @08:04AM
Nice idea. Too bad it is illegal. Ever heard of "Disparate Impact" as a legal concept? It says that if you are handing out licenses to practice medicine and your process doesn't have the same percentage of any 'Protected Minority" (and the list grows, MENA in currently being split from White for the next Census and will certainly get included in all of the quotas) on the output (licenses) as input (applications) that any judge or Diversity Czar at the DOJ, EEOC, HHS, etc., can declare "Disparate Impact" and order you to change your discriminatory process. Knowing this, you ensure you award licenses in carefully measured and documented quotas.
It is a fact that every graduating class has above average, average and below average doctors. It is a fact that unless you spend a fair amount of effort can't know which group YOUR doctor is in. It is a fact any minority (other than Jew or Asian) was awarded bonus points on admission and special consideration to ensure as many as possible graduated. In light of both facts (neither disputable) that the winning move if picking between two doctors you know nothing about but race/sex, you should pick the Jew/Asian/White and go for the dude. And probably a Jew/Asian/White female over a black/hispanic man but I'd have to look at the current college bonus point system to run that math, Hispanic man might beat white woman. What a colorblind society we have created.
Of course the real winning move is to learn more about something as important as your doctor and we should be making that easier. If we could see that the black female Dr. who studied hard has better measurables vs a white bro who partied too much in college and it impacts his patient outcomes, it would totally eliminate the need for crude stereotypes. If somebody has the hard numbers and still wants to be racist on something so important... well think of it as evolution in action. Which gets to the topic of today, big data. It shouldn't just be for the huge faceless megacorp.
Imagine if poor people, who now have the same Internet as everyone else, could run a search on a proposed bank loan and see what percentage of loans of that class (i.e. to people like themselves) go bad (remember, failing to pay back a loan hurts both sides, often unequally; hence predatory loans) and what the average interest in the industry is, etc. If they could quickly see what changes, of those within their means, to their credit score would have the biggest impact. Some of these things are actually beginning to appear and if they become widespread would do more to help the poor than any of this proposed whining about racist math.
But bottom line? Racist math? First we started seeing headlines in the eight years of Obama's "Post Racial" Presidency that looked ripped from The Onion. Now we are so far beyond, The Onion never would have tried to pass this crap off as satire.
(Score: 5, Insightful) by NotSanguine on Monday October 03 2016, @08:53AM
Nice idea. Too bad it is illegal. Ever heard of "Disparate Impact" as a legal concept? It says that if you are handing out licenses to practice medicine and your process doesn't have the same percentage of any 'Protected Minority" (and the list grows, MENA in currently being split from White for the next Census and will certainly get included in all of the quotas) on the output (licenses) as input (applications) that any judge or Diversity Czar at the DOJ, EEOC, HHS, etc., can declare "Disparate Impact" and order you to change your discriminatory process. Knowing this, you ensure you award licenses in carefully measured and documented quotas.
[lots more blathering deleted]
Two questions for you, jmorris:
1. Have you ever heard of Medical Board Exams [wikipedia.org]?
2. Where did you get your law degree?
I'll have to make sure never to employ an attorney who attended that school. What's that? No law degree? No wonder it smells so bad here -- because you're talking out of your ass, as usual.
No, no, you're not thinking; you're just being logical. --Niels Bohr
(Score: 4, Funny) by butthurt on Monday October 03 2016, @09:28AM
I had not written clearly. I attempted sarcasm, but neglected to identify it as such. Later I explained what my meaning was:
/comments.pl?sid=15807&cid=409284#commentwrap [soylentnews.org]
Apparently you didn't see the clarification, so my writing style caused you to waste nearly an hour and a half of your time. Please accept my apology.
(Score: 0) by Anonymous Coward on Monday October 03 2016, @01:59PM
It is a fact that every graduating class has above average, average and below average doctors. It is a fact that unless you spend a fair amount of effort can't know which group YOUR doctor is in. It is a fact any minority (other than Jew or Asian) was awarded bonus points on admission and special consideration
ha ha ha
[citation needed]
(Score: 2) by Wootery on Tuesday October 04 2016, @05:22PM
Stuff like this does happen, [slashdot.org] but I hardly think it happens everywhere.
(Score: 0) by Anonymous Coward on Monday October 03 2016, @03:24AM
I guess you missed the point and just want to espouse your own politics. The whole concept of credit and using statistics to make decisions on human lives is the problem. Lots of poor people need a break, but I'm sure your answer would be something along the lines of "bootstraps".
Smart people tend to write off other human beings, but we got here together and that is the only sane way forward.
(Score: 5, Insightful) by butthurt on Monday October 03 2016, @03:51AM
What do you want - for banks to give money to people with bad credit? For colleges to accept more idiots with bad grades? For businesses to hire people without any qualifications?
The article is about, among other things, criticism of banks for charging different interest rates to people based on where they live, and criticism of employers for making hiring decisions based on potential employees' credit records. It's not about the things you mentioned, which of course are more defensible. Why don't you have a look at the article, and, if you feel that the practices described in it are valid and desirable, explain to us why you feel that way?
(Score: 2) by jmorris on Monday October 03 2016, @08:16AM
The interest rate you are charged (vs other customers) is based on risk. Every factor which influences risk should be fair game, both when it gives some customers a better rate and when it charges some others more. Live in a bad neighborhood? It means your property value (often the object of the loan or the collateral) is lower AND it generally means it will be on a downward slope. Contrast to a similar valued property in a good neighborhood that generally goes up in value over the life of the loan. Life in a rough neighborhood tends more toward the brutish and short, also impacting ability to repay. You might be a standup guy (equal to the similar fellow in a good neighborhood) but your odds of being killed in random street violence / home invasion, etc. is greater. As it is for your spouse and children, any of which could disrupt your life and impact your earning power, again influencing your ability to repay. The word for these things is risk premium.
It isn't supposed to be 'fair' in the cosmic sense, only the economic and mathematical senses. The only other option is to charge everyone else more and 'redistribute' it to bad risks. There is also a word for that philosophy.
(Score: 2) by sjames on Monday October 03 2016, @09:01AM
Consider, idiot with bad credit inherits the house next to yours. Suddenly the interest rate on your credit card goes up and the limit goes down. Sound reasonable?
Of course, property values fall and people with worse credit move in. Within a year, you are a bad credit risk even though nothing about you changed.
(Score: 2) by maxwell demon on Monday October 03 2016, @09:57AM
Alternatively: Someone who's personally OK, but unfortunately falls into a high risk group and therefore has a bad credit rating, inherits the house next to yours.
The Tao of math: The numbers you can count are not the real numbers.
(Score: 2) by butthurt on Monday October 03 2016, @10:11AM
The only other option is to charge everyone else more and 'redistribute' it to bad risks. There is also a word for that philosophy.
The word, of course, is "insurance."
[...] your property value (often the object of the loan or the collateral) is lower AND it generally means it will be on a downward slope [...]
The sort of reasoning expressed in the article was: "people in your zip code tend to be riskier borrowers." ZIP codes are the postal codes used in the United States.
SoylentNews recently had a story about them, "How ZIP Codes Nearly Masked the Lead Problem in Flint" [soylentnews.org]. SunTzuWarmaster commented on it:
The problem comes when you use the Mail Delivery Code for segregating people into water zones, or "affluent neighborhoods", demographic characteristics, or other features which are _not_ mailing letters. Like every engineering tool, it has advantages, disadvantages, and limits.
What you're writing about is different. Like the original poster, you are proposing practices that are different, and perhaps more defensible that the ones the article is about. Certainly the value of something being offered as collateral is pertinent to a loan. This is commonly evaluated in an appraisal. I can see how the overhead in making and collecting on micro-loans could be greater, as a percentage, than that for mega-loans. If a bank were to charge higher rates for smaller loans, that could be justified (then again, a default on a large loan could render the bank insolvent)--but again, that's not what the article says. I'm not sure how future changes in value are assessed, or whether conventional bankers make the attempt. Someone who could reliably predict such trends could become very wealthy indeed.
(Score: 0) by Anonymous Coward on Monday October 03 2016, @04:20PM
Yes, because what humanity really needs is a sociopathic system that does not value human life. Or worse, that values human life with an actual numeric value.
The "system" exists to serve humanity but there are always those who want to game the system for their own selfish goals. Insurance is already a scam of sorts which could be replaced by a coop system. Trying to squeeze every last bit of change out of people, and specifically targeting those who have the hardest lives, is a sociopathic activity. If you defend this practice simply because you understand the logic of because A then B so we charge C, well then please get yourself to a psychologist asap. The only way to mitigate sociopathic tendencies is to know they are there and make a choice to behave differently.