Slashdot Mirror
Your Next Job Interview Could Be With a Racist Bot (thedailybeast.com)
An anonymous reader quotes a report from The Daily Beast: Companies across the nation are now using some rudimentary artificial intelligence, or AI, systems to screen out applicants before interviews commence and for the interviews themselves. As a Guardian article from March explained, many of these companies are having people interview in front of a camera that is connected to AI that analyzes their facial expressions, their voice and more. One of the top recruiting companies doing this, Hirevue, has large customers like Hilton and Unilever. Their AI scores people using thousands of data points and compares it to the scores of the best current employees. But that can be unintentionally problematic. As Recode pointed out, because most programmers are white men, these AI are actually often trained using white male faces and male voices. That can lead to misperceptions of black faces or female voices, which can lead to the AI making negative judgments about those people. The results could trend sexist or racist, but the employer who is using this AI would be able to shift the blame to a supposedly neutral technology. Companies are also having people do their first interview with an AI chatbot. "One popular AI that does this is called Mya, which promises a 70 percent decrease in hiring time," reports The Daily Beast. "Any number of questions these chatbots could ask could be proxies for race, gender or other factors."
We have sexist hiring now. What about all those blind hiring trials that ended up hiring more men... and then got cancelled and the result buried ASAP.
So we do have sexist hiring now... just not the kind feminists want to talk about.
... but wasn't a 21 year old recent college graduate from India. So yea, these already exist.
Translation: The sick freaks of the left are concerned that AI's will not be front-loaded with the politically-correct amount of anti-white bias, as defined by shrieking fascist moron SJW's.
I know somebody looking for a retail job. She would walk into stores and ask if they're hiring, only to be directed to a website where you have to fill out some massive 100 question test and hope that your name is picked by some algorithm for the manager to call and arrange an interview.
She must not fit the computerized profile that the tests are looking for, because she rarely got called back. It's been a frustrating and dehumanizing experience.
My mother said that when she was looking for retail jobs in the 60s and 70s, it was easy as hell. See a "Help Wanted" sign, walk inside, talk to the manager, have a quick interview, and if they liked you, you were hired. You didn't even need a freaking resume. It was a much more sensible experience.
I work in IT and I think computers are neat and have changed the world in many ways for the better. But holy shit have they totally fucked up other things.
Fuck you and that bot. I will work elsewhere. If employers find that they are getting the top 10% of the bottom 5% who will tolerate such nonsense they will probably rethink it.
Really it's the people that assign the parameters of the bot, not the bot itself.
I'm a good cook. I'm a fantastic eater. - Steven Brust
Am imagining a humourously un-PC bot interviewing me, please let this happen!
White men program it, so it can't be fair ... god fucking damn hypocritical sexist, racist twat.
PS. I have no problem with the sexist and racist part, I think everyone should have complete freedom of association, I only take exception with the hypocrisy.
So, I'll be interviewing against the same HR manager for the rest of my life, and if it decides that I am not fit for a job, it means I'm not fit for it for the rest of my life.
But Hitler-loving sex fiend bots are worse https://www.telegraph.co.uk/te...
https://youtu.be/U9uWSHWcHag
Wrong. When the Liberals get done with this, the bot will start by assigning your a fairly computed score. This will then be adjusted by Social Privilege factors, so for example you'll automatically be deducted points for being Male, white, or Asian, and given points for being female or black. Double points added for being transgender.
I got tired of always being interviewed by racist white men.
Unfortunately, it seems the term "Racist" doesn't actually mean anything anymore except a generic bad-word mud-slinging.
Doesn't matter what you do, you are racist these days.
Take for example, this test I had in public school on diversity:
Question: You are a hiring manager and you have two candidates for a job. One black, one white. Whom do you hire?
My answer: Interview both candidates and choose the candidate most qualified for the job.
My answer was marked incorrect. The correct answer is: "You hire the black man"
The whole test was like that. I scored 100% racist. Take that into perspective - I'm a racist because I REFUSE to treat anyone differently by their race. One of the great misinterpretations of Martian Luther King comes from the quote, "...[african americans] are not judged by the color of their skin but by the content of their character" gets warped into "...[african americans] are not judged."
Don't forget that in the last Olympics a coin toss was considered racist because it didn't favor the right race.
So is the chat-bot racist? 100% yes, because it can't see skin color and therefore can't give preferential treatment.
Not to worry, the SJW's will make sure it ends up racist against white and males before long.
https://www.usatoday.com/story...
https://www.cnbc.com/2018/03/0...
https://www.americanthinker.co...
https://www.theaustralian.com....
crap source
PROTIP: If you are against this sort of thing, and want to actually change the world for the better, NOW would be the time to refuse to do any sort of interview in front of a robot/AI.
You're welcome.
the robots in many cases aren't even aware of your race or gender so how are they going to select against your race and gender?
They're clearly NOT deciding on that factor as they literally can't because they're literally not given that variable most of the time.
What they decide upon are your qualifications. Now if the most qulaified people tend to be from group X or Y then that isn't racism to predominantly hire people from those groups. Statistically if you limit the population being examined to those with the qualifications there is no statistical variance in hiring patterns. You only see a statistical variance if you IGNORE qualifications. Which is idiotic because the entire point of setting an AI on hiring people or hiring someone to hire people is to have them filter the people hired based on relevant criteria.
What these "robot HR is racist" arguments ultimately are requesting is lottery based hiring. Where in random people in society are randomly hired for given jobs indifferent to qualifications.
Harrison Bergeron by Kurt Vonnegut is a dystopian society where this concept was applied to its logical conclusion. Everyone is forced to be equal. The clever are made to be stupid so they enjoy no advantage over the stupid. The strong are made to be weak. The graceful are made to be clumsy. The beautiful are made to be ugly. Employment in everything is determined by literal lottery. Total chance. Everything from the police to the president to your doctors to whatever.
It is a nightmare society.
The robots are not racists. The plaintiffs are equalitarian intersectional communists in most cases. The sort of people that advocate bad ideas that if applied lead to the society starving to death.
Any group that votes for that deserves the consequences without mercy. And to be very clear... that happens properly anywhere and those able to do better will leave. You'll be left with an incompetent rabble that simply couldn't do better anywhere else. Poverty and failure is the best you can expect. Literally starving to death is quite likely. Cannibalism is not off the table.
I've decided to stop wasting my time responding to AC trolls/sockpuppets... so if you want a response from me... login.
It's hard for most humans to ignore "not like me" mannerisms that lead to unintentional race/ethnicity, gender, age, or class bias. For many, it's also hard to ignore when a person's behavior is "wildly non-conforming" to subconscious expectations of behavior that are based on perceived race/ethnicity, gender, age, or class.
A computer can be programmed to ignore things like accents, dialects, slow or fast speech, choice of dress, or behavior that doesn't conform to most people's expectations based on gender, etc. beyond a broad "crearly acted unprofessionally" yes/no mark. If you cuss in the interview without a good reason, that's probably a "no."
A computer science Ph.D. is a "proxy for race, gender and other factors". Exceptionally test scores are a "proxy for race, gender, and other factors". Are you going to eliminate all objective measures of performance because it correlates with "race, gender, and other factors" in ways that you disapprove?
Stop posting these Reddit-level stories that are designed to get people riled up
It has become a worthless troll site for SJWs to push their grievances.
There's plenty of interesting science and technology news, but Slasdhot posts endless streams of shit.
You missed the point.
Having problems at point of employment isn't a reason to hire you anyway.
Do you want a doctor that didn't get decent grades in medical school? Most people never go to medical school at all.. would you like a doctor that spent ZERO time studying medicine?
If you would prefer competent doctors then you desire merit based employment.
I'm not saying anywhere that I think we shouldn't try to help communities with problems. However, I think it is also racist to look at these communities racially. Rather you should focus on other more important factors.
Did they come from a single parent house hold? Did they come from a troubled neighborhood? Was there drug abuse in the house hold? Was there positive reinforcement of good study habits etc in the child's household.
These things ultimately determine if there will be problems. Not the race of the individual. To suggest otherwise is to presume that given races are inferior.
I made the point above that your position is actually inherently racist. You wish to classify people on the basis of race.
Are there f'ups from every race? Yep. Are there high achievers from every race? Yep.
Focus on what separates the one from the other and you might actually help people.
Focusing on race will help no one. You will doom those you presume to help to continuing poverty by not addressing the underlying problems in given communities that lead to failure. What is more, your entire concept requires that we hire empirically incompetent people to do jobs. Which you will hypocritically assign to other people or other jobs you don't care about whilst betraying your supposed values by requiring merit based employment when it might actually affect you.
I like that you tried to start your argument with an ad hominem.
Let me try one on you which would be only fair.
I think you're also white, spoiled, feel inferior to your peers, and are attempting to play these pathetically constructed moralistic games as some sort of ploy to claw your way up a social hierarchy.
People like you hurt the people you presume to defend. You're a parasite. You feed upon the suffering of others and use it to humble your opposition and aggrandize yourself. You have nothing positive to offer this situation. And your insecurity is likely very well deserved.
I want everyone to succeed and for society to work. You didn't even read my argument before you dogmatically responded to it. You are no better than the foaming fundamentalists of bygone times that would thump the cover of a book they couldn't even read.
See? I can play this game too... and I'm better at it.
I've decided to stop wasting my time responding to AC trolls/sockpuppets... so if you want a response from me... login.
It is unfair to everyone, rotten to some, and you never even know why. Just like human resources wanted.
Not likely. I really wish tech reporting would get away from the 'coulds' and 'mights'. It's not just bad journalism, it's bad science. Until something is proven 100%, and found to be accurate 100% of the time, it's just a hypothesis. There are a great many reasons this will never become common place. I suppose it might be difficult to see if one's own brain works algorithmically. Fun fact: most human brains do not function strictly, or even mostly, algorithmically.
If an AI learns through experience that green people tend to lie more, and purple people tend to be more honest, then the AI will come to its own conclusion that it will be less likely to trust or work with green people.
I imagine the AI's creator (aka SJWs) will get all butt-hurt and override what the AI learned, and instruct it to ignore the facts that are non-PC.
As Recode pointed out, because most programmers are white men, these AI are actually often trained using white male faces and male voices. That can lead to misperceptions of black faces or female voices, which can lead to the AI making negative judgments about those people. The results could trend sexist or racist, but the employer who is using this AI would be able to shift the blame to a supposedly neutral technology. Companies are also having people do their first interview with an AI chatbot. "One popular AI that does this is called Mya, which promises a 70 percent decrease in hiring time," reports The Daily Beast. "Any number of questions these chatbots could~/b> ask could be proxies for race, gender or other factors."
What an elaborate argument for a possibly Racist interview. I'm curious, what is the alternative - rely on 'gut feelings' from experienced managers who've only previously worked with programmers of their own ethnicity?
As I see it, the AI software likely looks for 'tells' or other indicators an interviewee is lying/being evasive (eye movement, shifting in seat, nervous behavior, etc), and NOT making blanket decisions based on, say, surname, gender, or skin color.
Might there be a bias? Sure, but it will likely be much less than a human confronted with a mountain of resumes, trying to whittle it down to a manageable pile.
We'd like to think that a manager of a programming Dept staffed nearly completely with white, Indian, or some other ethnicity would be open to considering an applicant of a different ethnicity, she may fear a culture clash and focus on candidates that might fit better with her current team... is it wrong, sure. But it's understandable, to a certain extent.
Ken
In 1845, Frederic Bastiat penned the satirical Petition of the Candlemakers:
http://bastiat.org/en/petition.html
He presents a scenario where the professional candlemakers try to have regulations placed on the sun so that more people will buy their candles for light. Obviously this is ridiculous, but it is THE SAME THING that happens when people call for protectionism, either of industries or types of job.
Retail work is dying. That means the supply of jobs in that field is contracting, leaving more people to compete for the same jobs. Rather than complain about it, your friend should invest time in learning new skills, differentiating, and adapting to the changing market.
Unless you want to buy some candles.
What we really need is an AI that can identify the lizard people living among us. They're the ones responsible for everything that's going wrong these days and they're the ones blaming the world's problems on women, muslims, jews, immigrants, liberals, conservatives, etc.. Let's stop giving the lizard people our jobs!
Debate is a form of harassment. Do not question my truth.
Take for example, this test I had in public school on diversity:
Question: You are a hiring manager and you have two candidates for a job. One black, one white. Whom do you hire?
My answer: Interview both candidates and choose the candidate most qualified for the job.
My answer was marked incorrect. The correct answer is: "You hire the black man"
Cool story, bro.
Comment removed based on user account deletion
I'm sad to say that diversity, and encouraging diversity, can cost a great deal of money. Older employees need more health care, and refuse to do hours that younger, more eager starting employees accept as part of the work environment. Racism helps keep the work environment consistent in social needs, such as weekly work schedules and vacation time. And sexism helps prevent one of the greatest burdens of a modern workplace: transgender transformations, whether surgery is involved or not. The workplace cost of an employee changing gender far exceeds that of a pregnancy. Been there, done that, and the costs are *outrageous*, especially if work medical insurance needs to pick up the costs of surgery. And, *my god*, the costs of dealing with the mood swings far exceeds those of pregnancy, menopause, or midlife crises. I've had to sign the checks for all of those, and know which I can possibly afford to fund.
Ever since then I've been receiving unsolicited offers from companies ranging from game devs to private equity.
Anyone sweating robo-interviews should seriously question if they're producing any real value.
Once you can produce value, you will have more offers than time.
Everything else is excuses.
What youre saying is the Chinks are taking over? Quell suprise.
This may be true in Western countries, however I doubt most of the programmers in China or India are white.
Unlike the previous AC I did read your post. However, I disagree. AI can be racist, but it is an easy thing to avoid. AI is based off the human brain, and guess what it can be trained the same way. However, this is not its downfall it is its strength. If your company uses an AI to find qualified applicants and your hiring and promotion practices are racist, well guess what the training set will be as well. Don't forget you will be getting pretty much a carbon copy set of what you already have. If you want inovation you have to take risks, and guess what AI can't be trained that way. At least not yet, as far as I know we have not yet found a way to train AI to recognize something we have not seen before.
So you get out what you put in. But if your company is racist or sexist, caugh, (google) caugh. Then you can't expect any improvement from an AI. No matter how much of the human element you remove.
After reading your post (a couple times to try to detect hidden sarcasm) I have to say that I am 95% sure it is the most racist/sexist/stupid thing I have read today, although the day is young. I always love posts that start with the whole, your life experience is worthless because it doesn't conform to my perceived victimhood bit. Thank you for keeping my non-faith in humanity alive and well.
Your 'pick your parents' bit is the same sudo-religious clap trap I have come to expect from the poorly educated/highly indoctrinated.
Your 'born with the abilities' bit is extremely humorous when you use it in your screed against the US having a Meritocracy. You might want to look up what that word means. Here is a hint, if you don't have abilities you shouldn't make it far in a Meritocracy by definition.
I disagree with your 'just lucky' bit but as it has no real definitions its hard to argue. In my experience consistent hard work over a more than 5 years has always had a good payoff. On the other hand slacking and playing the blame game seems to result in a high rate of failure. Compounded over a lifetime lots of good choices seem to become significant while lots of bad choices seem to just reset you to a default poor state.
Your employment bit seems to be complete bull. Google specifically has a higher percentage of women devs than are available in the general workforce. This implies, but doesn't prove, that all things being equal they go with women. As of their last report, I can find, they are well under half white males. In fact the most obvious racial disparity is that they have over 3x the number of Asians as in the general population.
Further, in all employment situations that I have been involved with in the last 20 years there isn't some implicit bias towards sameness. Instead there is an explicit bias towards whatever is the pity group of the moment. I have seen the obviously worse candidate chosen time and time again, this is real racism.
Finally, all put together I think your post proves the reality of the grandparents post as a best case situation. At best your concept of a Meritocracy seems to be that everyone is dumbed down to the same level (remove the luck of birth) so there is some form of 'fair' results (not fair competition).
Posting annon to protect my family in the age of #metoo.
and the college graduate may more likely to stealing cash to cover there big loans as min wage part time at mc does not pay them off.
The companies using these kinds of systems should worry more about the impact on the organization, their measurements of the best employees and their competitive position in the future as the ideas for new, unexpected products will pass them by. This is anti-diversification technology on the deeper level where it actually matters, not on the superficial levels of racial or gender bias.
With AI, we can examine the source code to see if it really is discrimimating against people.
Most likely, blacks are not going to has the same representation per capita as whites because less of them learn software development.
This isn't because of racist hiring, tech is the least racist sector I know. Go find some rural farmers if you're looking for racists.
Many in tech however, oppose affirmative action because they believe in equal opportunity and not equal outcome. I suspect these tech employers will not hire blacks in significant numbers since affirmative action employers will get all the half-decent ones.
We can't have a discussion about ending racism until the left acknowledges that affirmative action is racism. I'd be all for ending racism, but the left only wants to stop racism against minorities. In thier view, racism against white's is okay.
I've made it a personal point to oppose and avoid supporting any "diversity" or "anti-racist" platforms for this reason. The reality is that they too are racists. And if everyone is going to be racist against me (a white) I might as well be racist against non-whites in retaliation.
Retaliatory racism is the reason minorities would be better off without affirmative action.
More evidence that diversity CAUSES problems. It's not a strength.
Tired of the constant clickbait stories and titles, /. is no longer going to be a tech site eventually.
Most programmers are Caucasian because in general, caucasian programmers just happen to be the most qualified. This also tends to mean those programmers test what they build on themselves, which isn't a racial issue, it's a practical issue. As a "white" developer, it would be very hard to test my AI programs on a "black" face, as I'm so white marshmallows make gangster jokes.
That's NOT what companies want. Have you ever been a hiring manager? The vast majority of people coming out of colleges these days are totally useless in the work force; that's why so many of them (even with PhDs) are unemployed.
I have a B.A. and M.A., both in liberal arts. Both degrees are completely useless on the job market. As a hiring manager, I couldn't care less what degrees someone has. A person trained by the academic guild (and often very poorly, with little critical thinking) doesn't help me sell product, cut costs, or innovate against the competition.
College is a waste for 95% of people, and it has been propped up by student loans. The student loan debt bubble in the U.S. exceeds consumer credit card debt. It's going to implode spectacularly.
The people who are actually intelligent know enough to build skills rather than relying on college to get them a job or make a living.
I wonder what your thoughts on Timecube theory is.
The would "could" is used a lot in that post. It doesn't mean the they "have" use biased training sets. Debiasing is a thing and it's not that hard. "If" they did not used debiased encoding vectors then, yes, they should be ashamed of themselves.
The people who are making the argument don't want society to work. They want it to fail. Facts and out-arguing will just get you shouted down.
Interviewee: "Whaaa..."
Interviewer: "Who's next?"
we need to eliminate the concept of the corporation, and strip all business entities of all rights, period. you should be persoally liable for any business decision you make. if we need a large project, we the people can fund it through taxes. no corporate personhood, no AI personhood, no allowing AI to make human decisions. we need to think seriously about what frank herbert wrote about, limiting computing power. we dont NEED it controlling anything that humans can do intellectually. all we need robots for is rote labor. we are gonna regret giving corporations all this power. it wont be worth it.
Even if majority of the programmers are white men, why would they use themselves as the training data for the bots? Wouldn't they use a demographic that is more likely to be interviewed for the jobs for which the interviewing bots are designated?
Any guest worker system is indistinguishable from indentured servitude.
I can throw rhetorical rotten fruit at mental cripples if I want. Its a free country.
I've decided to stop wasting my time responding to AC trolls/sockpuppets... so if you want a response from me... login.
First, I don't argue that AI "cannot" be programmed to have a racial bias. My argument rather is that there is no evidence that anyone has actually implemented such a hiring system anywhere... certainly not amoungst the major multi national corporations. And those are generally the targets for legal and political agitation probably because slimy people think they can get a fat settlement simply by being annoying.
Second, racial diversity is not actually hugely helpful for innovation. This is one of the odd logical contradictions you get from PC politics. This notion that "we're all human" but "if you have different pigment then you have different ideas"... its nonsense. You can have a more monolithic set of ideas with a very racially diverse group or a very diverse set of ideas with a single race or even a single race and gender. To say otherwise is to suggest that races have different mental processes biologically which if you don't contradict the position opens the door for saying given races will be better or worse at given tasks. Since we all agree that all races are generally equally able to do the same things at the same level... you can't contradict that position and then say that mixing the races offers some kind of intellectual advantage. This is a "x=2, x=3" argument. X cannot equal 2 and 3 at the same time. How this sort of patently absurd logic gets through our cultural filters is pretty terrifying really.
Third, your point about Google is well taken and reveals the double standards that are at play here. Google is naturally openly discouraging certain races from applying to jobs whilst these people here seem convinced that phantom discrimination exists despite their inability to actually present any evidence.
I've decided to stop wasting my time responding to AC trolls/sockpuppets... so if you want a response from me... login.
AI trained on white men will be more accurate with white men, but it doesn't mean that they will favor white men. In fact, if the black women it was trained with were particularly suitable, it will be more likely to be biased towards them.
Having spent decades looking at tools that allegedly do not utilise race, gender, ethnicity, etc to determine the best candidate, it increasingly looks like claims of protected-class-bias-free tools are a pipe-dream.
There won't be a next job interview for me. If I get laid off then I'm done. Fuck looking for another job.
AI or human, if a company actually hired by racism, they'd be handing their competitors a huge competitive advantage.
After all, they'd be using a meaningless criterion, while their competitors would be focusing of course on skills and experience alone. Right?
(You'd think an AI might catch this after awhile, but whatever. if it didn't, the market would.)
Tay is doing interviews?
Hey, That's a new way big companies can throw money at MS: Use Tay for their job interviews.
What. could. possibly. go. wrong?
Tay is cool
aaaaaaa
If you're a straight white male then it's time you realized this meritocracy stuff we believe in? We're the only ones who actually believe it, everyone else uses it against us. Every group we've extended freedom to historically is at the present time using that freedom in direct opposition to our interests, the interests of the founders of our nations and the men who came before us. I don't know why shit like this winds up on /.
There was a company around here that asked people to send a video to even be considered to a job. I hate speaking to a computer, so I just ignored them. I won't do anything like that as long as there are any alternatives. I more people refused bullshit like this it wouldn't spread.
As long as it's not a black robot. Red or Green is fine, but I don't like black robots.
If the majority of white male applicants are good, then why not skew that way? This machine learning will put racial realities that people would rather dismiss to the forefront.
This can be overridden of course by humans. The AI could have scored the black applicant at a 60%, but then thereâ(TM)s a racial override rule to satisfy diversity quotas. This would be exactly how modern hiring works at the big companies.
I work on the algos that are used to screen and interview employees, and know hirevue well. The algos do not take race into account and the relatively basic âmoodâ(TM)/âexpressionâ(TM) logic that video interviewing tools provide is not useful - or used - its more of a marketing gimmick.
State of the art at the moment is to be able to find holes in resumes and ask relevant questions before passing the candidate to a human interviewer.
Btw, most companies doing development in this space are so concerned about racial bias as some of the conspiracy theorists here believe. To the point of intentionally crippling the efficiency of the said algos so that racial bias does not show up in aggregate statistics. Nothing to see here, I think.
Fart fart poopy doop pee pee poo poo farty poop
so mad at those racist starbucks coffee bots
Seriously, work on your grammar before you ever post again. "May more likely to stealing"? "there big loans" "does not pay them off"? You barely have a point and you have no idea how to express it.
The St. Louis Symphony was one of the pioneers of auditions where the performers being considered for a position played behind a barrier where they could not be seen and with their identities unknown to those making the hiring decision.
I don't know if the results of this process resulted in statistics that made some people get irrationally upset. But it wouldn't surprise me.
There's no time like the present. Well, the past used to be.