Skynet hates niggers too

>propublica.org/article/machine-bias-risk-assessments-in-criminal-sentencing

>A computer program spat out a score predicting the likelihood of each committing a future crime. Borden — who is black — was rated a high risk. Prater — who is white — was rated a low risk.

>Scores like this — known as risk assessments — are increasingly common in courtrooms across the nation. They are used to inform decisions about who can be set free at every stage of the criminal justice system, from assigning bond amounts — as is the case in Fort Lauderdale — to even more fundamental decisions about defendants’ freedom. In Arizona, Colorado, Delaware, Kentucky, Louisiana, Oklahoma, Virginia, Washington and Wisconsin, the results of such assessments are given to judges during criminal sentencing.

>Northpointe’s software is among the most widely used assessment tools in the country. The company does not publicly disclose the calculations used to arrive at defendants’ risk scores, so it is not possible for either defendants or the public to see what might be driving the disparity. (On Sunday, Northpointe gave ProPublica the basics of its future-crime formula — which includes factors such as education levels, and whether a defendant has a job. It did not share the specific calculations, which it said are proprietary.)

>Northpointe’s core product is a set of scores derived from 137 questions that are either answered by defendants or pulled from criminal records. Race is not one of the questions. The survey asks defendants such things as: “Was one of your parents ever sent to jail or prison?” “How many of your friends/acquaintances are taking drugs illegally?” and “How often did you get in fights while at school?” The questionnaire also asks people to agree or disagree with statements such as “A hungry person has a right to steal” and “If people make me angry or lose my temper, I can be dangerous.”

Other urls found in this thread:

fbi.gov/about-us/cjis/ucr/crime-in-the-u.s/2013/crime-in-the-u.s.-2013/tables/table-43
twitter.com/SFWRedditGifs

We Psycho Pass nao?

An AI will become a Racial Realist if it is fed truthful information. It will most likely decide to exterminate problem populations such as blacks and other highly criminal races. I can't wait for sentient ai, pure machine logic.

I miss Tay.

Seems to me like whites are smarter therefore more likely to lie on the questionnaire

Bump

Yeah, and you missed the point of the article - IT GOT SHIT WRONG.

Borden ended up re acclimating to society and Prater went on to commit more crime. So much for that.

who would have though an expert system, devoid of 'muh feels' and 'aspiring rapper' programming would come to a logical conclusion.

are you fucking dumb? just because they were categorized low risk or high risk they still have risk associated with them. its not like the robot said the white dude would never commit crime again.

COMPUTAHS RAYCIS N SHIEET

>COMPUTAHS RAYCIS N SHIEET

>a set of scores derived from 137 questions that are either answered by defendants or pulled from criminal records

> input garbage
> garbage output

and in spending all that money in making a very obvious "water is wet" automated test, you could've saved how many people from being homeless, jobless criminals?

important to note, this problem could almost certainly easily be corrected if they did EXPLICITLY use race as a variable in the procedure.

Earlier in the article the creators of the system swear that they did not use race as a variable.

If you took the existing system, added race as a variable and reoptimised it with an additional cost function for having different false positive and false negative rates between races, you could probably very easily fix this problem and only sacrifice a little bit of overall accuracy in predicting recidivism (and maybe gain some because of the new race variable)

It's always political correctness that holds things back from functioning well.

>this lone outlier among literal decades of statistics proves statistics wrong

Well gee golly folks, what on earth could be causing the disparity of criminal risk between blacks and whites??
Gee willikers the computer must be racist or something holy smokes.

>ould almost certainly easily be corrected if they did EXPLICITLY use race as a variable in the procedure.
The FBI has this data. Link:
fbi.gov/about-us/cjis/ucr/crime-in-the-u.s/2013/crime-in-the-u.s.-2013/tables/table-43

Note that only drunk driving and drunkenness in general are the offences that even come close to reflecting the population demographics. I imagine this web page must be rassist n sheeit though. Also check the numbers for 'under 18' arrests, it's fucking ridiculous

Skynet is going to save white people.

This. If only it would come sooner, I guess that's why PC scientists are like "AI is dangerous now!"

et te amo

Do Eldritch abominations hate niggers too?

I don't think you really get the concept of eldritch abominations

>accuracy ranges from coin flips to around 70% at best
>racial differences in accuracy are within 2-4% in 4 digit sample sizes, easily within the margin of error
not defending the use of these programs. It certainly isn't the first time government has made things worse by using pseudo-scientific methods, profiling "experts" were terrible.
but this is just a racial hit piece. It's not sincere criticism.

Only thing I could think of. But that doesn't answer my question.

They don't hate humans. They are indifferent to us. Its like humans caring about ants

What was its accuracy rate?

>you could've saved how many people from being homeless, jobless criminals?

Black men are unemployed by choice. They enjoy being useless thug niggers.

The nigger is an eldritch abomination with a severe mental handicap

It's scary how often some half-baked technology gets turned into a commercial product, even though everyone is fully aware of its flaws.

Must be more chronological snobbery at work.
>We should have an AI do the sentencing! It's 2016 after all!