Slashdot Mirror
New York City Moves To Create Accountability For Algorithms (propublica.org)
The algorithms that play increasingly central roles in our lives often emanate from Silicon Valley, but the effort to hold them accountable may have another epicenter: New York City. From a report: Last week, the New York City Council unanimously passed a bill to tackle algorithmic discrimination -- the first measure of its kind in the country. The algorithmic accountability bill, waiting to be signed into law by Mayor Bill de Blasio, establishes a task force that will study how city agencies use algorithms to make decisions that affect New Yorkers' lives, and whether any of the systems appear to discriminate against people based on age, race, religion, gender, sexual orientation or citizenship status. The task force's report will also explore how to make these decision-making processes understandable to the public. The bill's sponsor, Council Member James Vacca, said he was inspired by ProPublica's investigation into racially biased algorithms used to assess the criminal risk of defendants. "My ambition here is transparency, as well as accountability," Vacca said.
Every patriotic American should be working to oust Trump from office so he can't continue to aid Russia's attacks on our country.
It's often these same people who impose systematic discrimination based on these exact criteria.
More and more so-called 'AIs' are being used in place of algorithms (due mainly to magical thinking) but even the designers of these AIs can't tell you what they're really doing under the hood. That's where we're going to get in trouble with regards to 'accountability'.
Can people figure out how it discriminates against certain race or gender?
^(oo)^pig~
Algorithms don't discriminate if you remove the kind of data (race, age, etc.) that would allow them to make categorizations or judgments based on that data. But if you examine the results after the fact and reapply those labels and find some difference in outcomes, its because there is some difference in input, not a category identifier. If you find your algorithm thinks African Americans are a worse lending risk, it's likely because they're categorically less well off financially than other demographic groups, not because its racist against black people.
This kind of idiotic approach is just ignoring the actual underlying problems or differences in favor of trying to slap a band-aid on top of it to assuage guilty feelings. Worse yet, it prevents confronting the actual issues head on and is doomed to failure.
See subject & demand that your government BANS BUMP STOCKS IMMEDIATELY. Banning bump stocks is the ONLY way to prevent future mass shootings like the Las Vegas shooting.
The Vatican doesn't want bump stocks banned & is spending millions of dollars to LOBBY AGAINST banning bump stocks. The VAST FORTUNE of the Vatican was obtained through illicit means including funding wars in Europe. The tradition of using VIOLENCE to grow the wealth of the Vatican is alive and well. Collectively, the Vatican, George Soros, and the NRA, are fighting (figuratively, I hope) to keep them legal. Jesus Christ is ROLLING IN HIS GRAVE.
* Moderators will undoubtedly attempt to censor my post to -1. This censorship is wrong and is probably paid for by the Vatican. It's no better than Hillary Clinton and the DNC paying shills to support George Soros and spread conspiracy theories about Russia. Slashdot must ultimately ELIMINATE MODERATION.
The real purpose of the Russia investigation is to DEFLECT ATTENTION away from the Democratic COLLUSION with the Vatican. Any alleged Russian meddling in our election was dwarfed by the Vatican meddling, delivering tens of millions of dollars to support DEMOCRATS. Notice how Democrats paid lip service to banning bump stocks & then took no action. On the surface, they appear to support gun control, but in reality Democrats align their agenda closely with the Vatican on nearly all issues. The Vatican is by far the biggest manipulator of American elections, but does so behind the scenes. That is why only one Catholic has ever been elected President and he was assassinated by the Vatican for going against their plan to remain IN THE SHADOWS.
Good posts (like this one) keep ending up at -1, and it makes Slashdot IMPOSSIBLE to read. Moderation (censorship) is RUINING this site. The moderation system has become a means for enforcing groupthink and SILENCING users like Creimer who become unpopular with a small group of MORONS with mod points. Once their karma is depleted by the ABUSE, they can only post twice a day, and at a score of -1. I post anonymously so my imposters can't silence me in the same manner. Otherwise, I would no doubt be their next victim based on the amount of abuse I take.
Most of the abuse is because I won't OpenSORES my Hosts File Engine. JEALOUS LOSERS are angry they can't PIRATE my code and because I tell the TRUTH about them in my posts. Most are angry because they keep DESTROYING themselves against me. It's really quite sad and pathetic.
APK
P.S.=> Idiots like AssFux(lol) have no courage. They will keep censoring my posts and making false statements about me. I circumvent their feeble efforts with my Hosts File Engine. It's also FAR more secure than any antivirus or firewall can ever hope to be... apk
The AI book that everyone should get is available for pre-order. "Artificial Intelligence For Dummies" by John Paul Mueller and Luca Massaron.
It's the same way I.Q. tests discriminate against conservatives without input on their political beliefs.
Support Right To Repair Legislation.
Why not to cover these cases with a general law relating to accountability, matter-of-factual behaviour and equality regulating all legally bounding decisions and finalized government work products from the federal to local government level? You'd be covered at least until we bump into hostile aliens at the edge of the galaxy.
Can people figure out how it discriminates against certain race or gender?
The proposal here is to do a study to understand that, yes.
You did notice that this article was about studying the problem to see if there is algorithmic discrimination, right?
However, let me also point out that since the example discussed in the text was about DNA testing, I would point out that race and gender are encoded in DNA, so "does not have race/gender input" is not applicable here.
In other cases, however, yes, it turns out that there can be race and gender encoded into input data even if it is not explicitly listed as "enter race and gender here." You could look at the articles cited earlier, such as https://www.theatlantic.com/bu... or even look at the Yale Law review article about the book on the subject http://michiganlawreview.org/w...
When facts become painful, they must be discarded to prevent the truth from hurting. If an AI said that giving a loan to an African American is 15% more risky then say an Asian. It's racist right? But if that is what the metric data shows, a 15% increased change to default versus an Asian, it's just the fact of the matter. It could be because of racism that there is a 15% high chance of default, but it doesn't change the fact in this circumstance. You need to make decisions based on concrete facts, not warm fuzzy feelings. When facts don't align with perceptions and ideals, the facts are discarded and ignored now leaving a society blind to real problems. That is dangerous. It's why places like Chicago are a shit hole. There are clear metrics that show why, but because they are uncomfortable, they are ignored leaving spurious elements to 'address' without tackling the root causes. When math and science are discarded because of feelings.... fuck.... we're doomed.
A very good book that discusses the problems behind the blind implementation of algorithms is Weapons of Math Destruction by Cathy O’Neil.
So in your rush to appeal to authority you cite Yale and link to Michigan. Nice move
But why can't race be used?
Should we really treat a group of 90-year-old Japanese women the same as a group of 22-year-old African Americans?
After all, if black lives matter, it's those 22-year-old African American that are the greatest threat to black lives.
Go to the highest crime part of any city and you'll find one thing in common - it's wall to wall n1ggers.
People are afraid to tell you this. Computers are not.
Let California be the next to pass a bill for Accountability of Earthquakes. That way they can predict months into the future when the next one will hit and prevent quakes from disrupting neighborhoods of any particular ethnicity.
I read that as: statistically honest
Well, the issue I foresee in this effort is that while the algorithms will be perfectly fine, it's the policies created to make up for well functioning algorithms that will be the problem.
Because what policymakers will quickly find is that having equal algorithmic treatment or having equal standards for all does not lead to the outcomes they want, as people of different demographics, backgrounds, capabilities do not take up services or have success against different programs in the same way.
This is the problem with policy always -- a tendency to believe (at least in recent liberal democracy) that people are all drawn from the same starting set and have equal propensities for doing / being / acting / achieving / using certain things. And when policymakers find that to be the unavoidable truth, democratic pressure forces them to find ways around this truth and distort the outcomes.
No algorithm will get around that.
How about we put a pin in algorithms and focus instead on accountability for C-Suite executives, Boards of Directors, and Politicians? Once those people are held accountable, we can circle back to algorithms.
So what about proxy variables? What about variables that act like proxies in the aggregate?
How do you know whether it happens?
What about the difference between causality and correlation? What if a causal variable correlates with one of these that they are filtering?
This sounds to me like "same old same old", politicians looking for power by making promises they can't keep to people who are too dumb to realize they are being manipulated. It is a lose-lose proposition, because not only is there a lost of effectiveness of justice, it is done in the name of justice, and with a substantially higher taxpayer cost. Selling false hope for justice, when the only possible outcome is injustice, is a loss. A good leader should know better.
a previouslVy
If it was we wouldn't be having this conversation. Facebook can already guess your race, age and even sexuality based on the data they have about you even if you didn't tell them any of that. America is more segregated today then it was in the 50s, and that's not by choice, it's by design. This is what people mean by 'institutionalized racism'. It means racism is carefully built into the institutions rather than enshrined in law.
So Blacks can't get loans or buy decent houses. Gays can't take advantage of marriage tax advantages or health care advantages, and women face a large pay gap. But at no point are there any laws in place to make all this happen, it's all done organically. And if anyone complains you just point to your data set and say 'We don't collect race, gender or sexuality'.
Hi! I make Firefox Plug-ins. Check 'em out @ https://addons.mozilla.org/en-US/firefox/addon/youtube-mp3-podcaster/
It must have been something you assimilated. . . .
...fighting with a FFT-based algorithm that causes plenty of troubles. After reading TFA I started wondering if my algorithm does not work as expected just because it is just discriminating me. I will ask mr. Vacca about...
If a model of, say, the likelihood of recidivism or the probability of loan default results in disparate results for different races, yet can be shown to be accurate in terms of ability to predict, is that discriminatory? I can see that happening with A.I. systems where the datasets are fed in and the black box then spits out results that, while accurate, are completely opaque as to how the results are obtained. Is congruence with observed reality a defense against charges of racism?
If gender is "encoded in DNA" then why do all the SJWs scream bloody murder about bathroom equality when they can't pick and choose based upon how they "feel".
And if race is "encoded in DNA" then why does Rachel Dolezal insist she is black?
Controlling for race and culture, Let's compare average IQs of countries:
Korean Peoples
--
South Korea 106
North Korea 105
Han Peoples
--
Hong Kong 107
Singapore 105
Taiwan 104
China 101
Almost a 1.0 correlation between economic freedom and IQ.
Maxine Waters has gone a lot further than GP poster.
Which makes her a Republican trying to discredit Democrats?
If you say so....
When outcomes do not match preconceived notions of "fairness" the process must be skewed until it does. Consider advice from the Boston Fed to banks explaining how to comply with CRA:
Sources of Income: In addition to primary employment
income, Fannie Mae and Freddie
Mac will accept the following as valid income
sources: overtime and part–time work, second
jobs (including seasonal work), retirement and
Social Security income, alimony, child support,
Veterans Administration (VA) benefits, welfare
payments, and unemployment benefits.
Welfare payments and unemployment benefits should be considered as viable support for mortgage borrowers?
I'm a relatively enthusiastic supporter of my council member and council member elect and am quite sure neither could define "algorithm" much less explain what is being proposed and what exactly is meant by "Algorithms". It's just a buzzword to use so that they aren't "roadkill on the information superhighway" (for a trip back to 90s pol-speak).
[spoiler alert] "Bubble sort is sexist" ... actually there are better jokes here so get on it slashdot.
I don't have a problem with this. I have a problem with some of the Mayor's diversity initiatives - downgrading the school system for kids who do their homework isn't great - but they were going to waste some of that money somewhere so why not pay for some "studies". I know two number crunchers in midlevel city government and they'd both be happy to have a month of light duty investing "Algorithms".
Snark aside the paper on NYC school matching is interesting though an acquired taste (like all maths)
google: "Strategy-proofness versus Efficiency in Matching with Indifferences: Redesigning the NYC High School Match"
AI doesn't know to "look the other way." When people are in charge, they can take a subtle hint (give more to this group, don't mention this group if they commit a crime, etc.). We just need to inject a Social Justice Warrior loop into these AIs.
Gendered bathrooms are pointless and in the long run they're going to go the way of "whites only" bathrooms for similar reasons. In the sort term simply not respecting the gendering of bathrooms is the same kind of civil disobedience as Rosa Parks not giving up her seat on that bus.
Rachel Dolezal, being kind of an idiot doesn't really enter into the question of whether a company that figured out they can do a DNA screening using a system that gives results correlated with perfomance reviews, is or is not being racist, if the system has simply "learned" that perfomance reviews are correlated with the genes that control melanin production and therefore tends to reject job application from black candidates because they will probably get bad perfomance reviews according to their DNA profile. Additionally there is more to ethnicity than genetics most of it is also hereditary but not exclusively passed form parent to child. "Black" can mean more things than just was born to dark skinned parents. Consider "Jewish" as a more obvious example of how one might become part of a different ethic group that people are usually born into.