Artificial Intelligence: The Robots Are Now Hiring | Moving Upstream
Vložit
- čas přidán 19. 09. 2018
- Some Fortune 500 companies are using tools that deploy artificial intelligence to weed out job applicants. But is this practice fair? In this episode of Moving Upstream, WSJ's Jason Bellini investigates.
Watch for new episodes of Moving Upstream this fall.
To be notified of future episodes and updates on the series, sign up here: confirmsubscription.com/h/d/C...
Don’t miss a WSJ video, subscribe here: bit.ly/14Q81Xy
More from the Wall Street Journal:
Visit WSJ.com: www.wsj.com
Visit the WSJ Video Center: wsj.com/video
On Facebook: / videos
On Twitter: / wsj
On Snapchat: on.wsj.com/2ratjSM
Wow! Pretty soon Corporations will be able to hire only "pure" psychopaths. What an achievement!
Seriously I was thinking the same thing people going to learn how to fake emotions and stuff
messed up.
The asian lost because the AI thought his eyes were his eyebrows.
So much for levity. But, some of these self teaching AI that are showing up in chess are very impressive. They're wowing the chess engine aficionados impressively.
@@MartinUToob
Yeah, but with humans it's gonna be terrible: Gipsies and afro-Americans have a high crime rate, so the AI would discriminate them. Some people after injury do take medications and may look like someone who took drugs, so they will be disqualified, too. People with German names will be promoted while people from certain countries will be discriminated.
And the lover of the owner will always get hired and if someone would dare to ask why? Because the AI said so.
The employers will be able to deliberately discriminate and blame it on the AI.
Not willing to work for a company that is not willing to put in the time and effort to get to know me
It's okay. In another decade or two, for most jobs, humans will need not apply
So it sounds like in about 5 years you'll be homeless
+Fenil Patel "to get to know me" We aint dating here buddy, do your job and go home.
No worry, mate. Another 10 are waiting for your place! The corporation can replace you easily. You are another cattle in the line to be milked til you bleed, or dare try searching for your rights.
As a matter of fact these companies are getting to know you better than you know you. They are studying your facial expressions to determine your emotion and tracking the time it takes you to answer questions and then grading the answers you give. They are collecting a ton of data and are getting to know you better than most interviewers ever would.
Is it as personal? No.
Do they learn more and are they able to make better hiring decisions? Yes.
Time to take acting classes
Time to bleach my skin!
Or time to go back to school..
*Implying its possible to fool AI*
Unfortunately, acting will not fool an AI. Facial patterns when analysed over time can be classified as genuine or in-genuine very easily, and in those you would not likely suspect.
@@HappinessOrDeath It is, its just probably significantly more difficult. That's because a neural network won't distinguish correlation vs. Causation until its been fooled by it several times.
Time to sell an AI that will take these tests for me (using a 3D image of me)
Good idea.
yes! Hire Leonardo Dicaprio lol
Machine Vs machine
This is actually sad
4:15 trust me... if u score so low - they will not watch it.
Probably not the best use for AI
Why not? This is an excellent use case of narrow ai
AI will be used everywhere whether or not you think its a good application or not. Get used to it haha its here to stay
@anglekan I don't think you understand the change that's coming. I don't think you understand how AI truly operates. It's not limited in its functionality and expandability like humans are. One human can only perform one task really well at a given time, two is pushing it. AI's running on high core and high thread processing units can easily do multiple tasks simultaneously and while the AI judges you for the position you applied to, it's also taking in terabytes of data every hour as it analyzes speech and facial expressions and makes sense of them. When you base your expectations only on what you see, you blind yourself to the possibilities of a new reality.
@@merlinthelemurian3197 there is zero link between speech/facial data and skill. zero. Its not magic. Actual data scientists know how to form ML products so that ridiculous applications are avoided - like this one. Data can be manipulated just like statistics into any observation you want, and its the scientists job to interpret that data properly.
@@Aaron565 still thinking too small. Thinking like a human. If I ask you a question about c++ programming and I know how to read micro expressions, even at a human level, I can easily determine a lot about your relative skill level. If I also know your entire background and who you've worked around, it's very easy to determine how well you did your job in the past and fairly simple to determine your confidence going into the new job. An AI that can do all of this and is getting exponentially better would have no difficulty judging your skill based off of generic questions in your field of study
No. Just NO.
Smiling at AI would seem as reasonable as smiling or frowning at my toaster.
If they use the image of a puppy I won't stop smiling.
That's prejudice to robots o.o
except your toaster isnt analyzing your body language to categorize you into a group that may or may not be favoured to get hired.
@@HappinessOrDeath
Toaster as an informant of the the police-state will be the next step. I have already received an order for 10 million units. Business is good 👌!
Smile is //supposed to// come naturally from the anticipation of the new job...
I have a better solution
Hire robots
I thought robots are developed to do work for us not to rule us
Robots are developed to work for us if they can't develop their own intelligence or mind. It's worrying if they have the capabilities to expand their knowledge like us (and definitely they will be smarter and learn faster than human, no doubt on their capabilities on these aspects). Human literally took over the mainly world due to higher level of intellegence compare to other living organisms (what separate us from animals you could say). And I believe the "machine race" won't be willing to enslaved themselves and work for a race called "human" if they have their minds (and common sense?).
We have had enough of AI took over the world movies. Let's make one in real life.
Why can't we buy robots and have them work in our place and we get the money ..?
Why would you get the money?
If everyone can get money that way, money will lose its value eventually.
Cockyau Tan I really doubt that
Good question..😈
Little thing called the industrial revolution, happened a while ago..
Oh God please, no, just no! I love AI and working towards advancing it. However, I feel more and more uncomfortable each day because this technology is used for something that I didn't expect.
Yeah, sorry but this is just the tip of the iceberg unfortunately :/
Okay, to tell the truth, this REALLY creeped me out
Sam D pp
Jack Ma will never have a chance to start a company call Alibaba if we decide who will be funded with the assistance of such kind of AI tech.
Elon too
You're missing the point. This is only the first step. Once 35% of people are unable to get a job due to no fault of their own, it will force a UBI. The UBI will allow AI research to develop even faster, creating a positive feedback loop as there will be more hands on deck to assist, and as AI technology improves even faster, soon those hands that are assisting will be replaced with robot hands. Once we reach the point of 80% of people unable to get jobs, which will happen very quickly, the UBI will transform into an RBE and it will allow people the resources available to start any business they want with zero risk as abundance of resources is available to all due to the automated workforce being able to produce goods at record prices. We live in the age of information but we are approaching the age of abundance due to automated technology. Anyone can be the next Elon in an RBE
@@merlinthelemurian3197 I truly want know what you are on 😂😂😂
The minute they dont need you they will take you out of the equation!
Interesting. Knowing that the AI will look deep into my soul and make a decision based on my facial expressions, I'd say i'll have a 50/50 chance at making it think I'm skilled enough to get hired. But also scary that it collects that much information (most of it public due to social media) to determine a score where you "might" be looked at for consideration. As a UX'er, this is crazy.
Sorry but if “white American” is ceo of the firm I don’t see how AI won’t be bias
My biggest concern would be about what happens when someone is in pain, for instance they have a headache or pain/discomfort from a disability, etc. Your micro facial expressions could be similar to being uneasy or nervous when you're actually not. This could have a really negative impact on them.
2019 HR: can we build an AI for better hiring ?
2020 AI: firing this weak HR-Human is the best solution for better hiring performance ;)
algorithms are biased, it is hard (if even possible) to make them not to be
@@thepokemonlover3806 most of these system are considered (or advertised) as that they are not, that's the problem
Bias is embedded in the nature itself through evolution. You can only FORCE a system to act like unbiased.
Well, this is quite something. Artificial intelligence in the hiring process. Most of us are screwed.
Erm....find a way to bribe the AI? Lol😂
At this rate, AI will develop a bias for... AI (cylons come to mind). Basically, it will find All humans worthless and disposable. Nice going humans. ;)
Here's the problem. To beable to study human behavior one must first remove the computer being used as a human interface on a unsuspecting public.
I recently did an application that used HireVue. It was step 3 of the process and I still have yet to talk to a real person.
The reporter has to be told that 37% is not pretty good. Not sure if that is worse or that he has to be told that percent means "out of 100."
Make American Smart Again!
NOTE: Please see Jason's clarification below.
incredible. almost sad. %, he should have ask 37% up or down, and with or without margin. he's from wallstreet dont forget
I was joking. Guess that didn’t come through well enough.
Jason Bellini, if you were with The Daily Show, it might have come through. Thanks for clarifying. I was not expecting that.
Nobody detected the joke, we, the man you were talking with, not even the AI, and im sure the cameraman neither. Jokes aside, your contents always brings reflexions, well done.
@@Anand-qb1wp
It's idiotic. The reporter looks like a reasonable person, he cannot get as low as 37% ... unless there are other 63 very attractive con-artist applicants.
That's so stupid, I may not look like as a movie star but I'm necessary in my workplaces. I should receive a score close to 100%, but an AI can rank me anywhere.
The AI is coming. Better get ready
Take me to Mars
Yes, I'm ready for the Overlords with open arms!
I think your an AI Elon
Cool this is nice, been three years this was done any follow up which shows how much this evolved ever since
So glad I am close to retiring. The future of work looks brutal.
8:00 EXCUSE ME!? ITS ILLEGAL FOR COMPANIES TO ASK ABOUT OUR SOCIAL LIVES!! YES its illegal to mine our data to profile us based on our free time...
Facial expressions are not reliable nor, a developed science that I have ever heard of.
Who developed this technique and when?
Certainly, nothing I have ever read about in depth psychology. Nor anywhere else. Need more information to make an analysis of this unproven conjecture. 🤮 Puking would be my first reaction, however, I would have to verify my initial judgment. Contempt prior to investigation is unwise.
"Is it legal? "Yes" "Is it ethical?" "Well..." lol.
This system is an example of innovation but it is interesting to know what aspects the robot will analyze and even the time that the applicant will have to answer the questions. It has its points in favor but it also has others against it.
Wow this sucks and shouldn’t be done.
Skylab time? Hello Arnold do you want my jacket, my gun and my boots.
People still confusing "legal" with "ethical." Glad this video at least acknowledges this.
This eerily reminds me of the late 19th century criminology theory that you could identify potential criminals based on certain facial features. It is absurd to think that a person’s job performance is linked to “micro expressions.” There are just too many factors in the realm of human anatomy, emotion, and expression to simplify things down to that degree. And it gets really disturbing when you hear all of the stories about the unintentional biases that tend to be baked into these data sets. With every premature push of modern AI tech it becomes more and more clear that we are currently living the sci-fi dystopian narrative. This is the world we are actively creating.
One reason they maybe unwilling to share, is because they themselves do not understand how their AI learned how to choose. (As is one of the problems with deep learning!)
Or, that they are faking it till they make it, and are using crude gaming type AI, (+ basic deep learning?) and letting people think that they have made a new discovery. When in fact they are just using an older tool in new ways.....
How to remove human bias from interviews using AI.
1. Program said AI with biased humans to be "unbiased".
2. Use "unbiased" AI to hire.
3. ???
4. Profit!
7:25 and 7:31 are my favorite parts of the video ;)
This will cause companies to hire just the same people over and over and big companies will be even more boring. Diverse pool of people and perspectives bring even more improvements
Nice development in the world of "Knowledge is limited" (Jiddhu Krishnamurti, we Brockwoodian used to lived together while he was still alive as: Krishna Ji). Thank you for sharing this video to some one used to be one of those lunatics, but still interested and delving into this AI notion.
great video, i personally always get jobs or interview 2nd stage when the 1st stage is phone or face to face, AI can do what people see
If AI is going to do most of the hiring process and play a role in HRM too, don’t we run risk to forget that HR is always about working with human beings and not inanimate resources?
One major problem with this software is that the ai can’t tell you how a candidate will change during training.
I wonder if the managers and software "engineers" that came up with this system could be hired by the same computers they created.
Computers cannot form an opinion about a person. That requires emotion and reasoning, which computers will never have.
Sure, they can come to a consensus based on the data it collects but that consensus is based on what the computer is programed to collect.
Humans are more than the sum total of their parts. A computer can never comprehend what a human feels when walking barefoot on the beach holding the hand of a loved one.
Yeah you would think wouldn't you
Removing human bias" for "artificial intelligence bias"
Which is worse ?
That's crazy!!!
Using robots to hire people, whom they plan on replacing with robots anyway...''genius''
Now people will have to start training for the interview with the bot and stop making comments on social media, like YT. Oops.
Your training attempts with the bot will also be taken into consideration.
Hard to find good help these days. Humans are so last century. Maybe there is something that will be better than human individuals with all their cumbersome needs and wants. Realy, who do they think they are wanting good healthy food and homes of their own and ....freedoms... Disgusting.
Wow! loosing my job from A.I. will come sooner than I thought.
Funny thing is, by looking for candidates that match a certain pattern among your "top" performers, you are ingraining more strongly the stereotypes that the company already has that determined who are the top performers. So, instead of removing stereotypes, I strongly suspect that you are enforcing stereotypes far more. I would consider this to be a weakness in a company doing this from an investment stand point. The reasoning behind this thought is that as you try to more closely match your sample population that you have determined is your best, you selectively weed people out who might approach problems from a view point that your selected population would not, leaving you more likely to be blind to situations. I would think that a far better use might be to weed out individuals who have undesirable traits, such as swearing, poor grooming, or other negative personal traits. I do see this becoming a potential obstacle to the ADA as people with facial or speech issues would be likely to not match the standard normalized traits. Just my thinking.
My anxiety just went through the freaking roof.
heres hoping a quadcopter with c4 ziptied to the chassis detonates next to kevin harper of hirevue's head
I remember having to do a deepsense personality test for a job
What AI-driven hiring procedures often do is to remove human bias by introducing another human bias - that of the people, who configured the darn thing. :P
Scary.
Hiring people based on personality how totally ridiculous. They're just building their ai as well.
A name like Wong automatically place a candidate in last ranking. Having the “wrong” name even without AI often eliminates people
Yes, and I'd rather carry a shovel behind those elephants.
Companies currently use random selection, this is far better no firm has resources to vet millions of peopel that apply and keyword searches still limit. This also eliminates bias, as a robot doesn't care if you are black, a woman, a jew or whatever.
Doubt it can be successfully used in a wide range of professions.
er... what do they do if you don't have a twitter or Facebook or other such accounts? or if one does, but doesn't use them? As for Facial expressions, what happens if you don't follow the norm in expressions, due to injury or other medical conditions?
It would be interesting to see if behavior during the interview can be reliability linked to say, intelligence.
It's very biased because whomever programs the AI components or the robots will interject their OWN biases!!!!
Thanks for the information. Really cutting edge. I'm sharing it to Bonus Videos page on Facebook.
So what if you don't have any social media, or if you use it anonymously?
How creepy
Just another gate that ppl have to climb thru, find LinkedIn profile of top performer and mirror their facial expression #boom Hired. Create a twitter profile tweeting about hard work is good and being policy complaint #boom hired. In all seriousness, these tools need better data points than facial expressions or witty twitter comments to predict accurate KPIs.
Interesting but I would refuse to use it myself. I go for interviews regularly but one thing I will not accept is psychometric tests. I am just not good at them. I went for one job a few years ago and beforehand I told them what I will and will not accept. At the end of the interview they said just one more thing we have this online test and then called me a few weeks later and got mad at me for not completing it. It didn’t matter as I took another job in those weeks. You have to know your strengths and weaknesses.
Wow I thought this was completely absurd!
These kind of test is very acutate, people can react complete different in situation.
These corporations, need to be shut down completely!
Why ? whats the problem with choosing a different way to hire people for THEIR company ?
Cool, but what if the model picks up certain features during training that you don't want it to, say all the top employees in a particular company are hispanic and it gives higher weights to that.. A lot of previous softwares have failed because of that Compass being a noteworthy example.
That describes why Inference is really important in ML and AI.
Good AI takes years to develop with continuous evolutions unlike what the companies nowadays are doing developing it aggressively for the sake of staying competent in the market
So... the end of the world is near !
I think you are trying to say end of humanity.
Christian rapture then Great Tribulation. Find Jesus Christ quickly and repent to be saved. No more warnings.
@@trentsc4929 cult
So he (Kevin Parker, CEO) says their platform removes human bias, but what if the (training) data is biased?
i suck at interviews already, why do you wanna make this harder for me
lol who says Ai wouldn't be biased?
they analyse data objectivly
which will lead only to biased outcomes.
It funny how it looks like the baseline scene from Blade Runner
It's bad enough that one cannot always be at A-game level before humans, who may or may not give u a chance to explain why you answered how u did, now AI wont even think twice about u.
Either human adapt to the overlord algorithm of hiring AI in various aspects (Job skills, Related life skills, Psychological skills etc ) or just get left out of labor market. AI is going to BETTER with more PUBLIC and PRIVATE DATA. Every single MICRO seconds are judged instead of action or work that is very strict control.
The 146 dislikes are from HRs all around the planet.
i like the idea
The pencil and paper personality assessment industry has been held accountable for decades to provide valid and unbiased results in determining fit for countless job profiles. This continued when the assessments went online. So when do we see this become a priority in AI-based observations and decision-making? Basically, the same EEOC and APA rules and ethics apply here, no matter the medium being used. So, big data, show me the coefficients!
Can an Entrepreneur run their Business Model/Idea through an AI program to make it or determine if it will be successful?
"There's no clear established pattern of what facial expression is needed for any job"", dope. Ifeoma Ajunwa gets my vote for any future AI regulatory body, if there will ever be one.
What about the people who have autism or have a mental disorder? Are they not allowed to the same opportunities as the rest? Facial movements doesn't make any difference if you are good with the job requirements.
Strange.
0:53 If you can read human expressions and micro expressions, you can mind/thought read people. 1:14 Except that HR managers RARELY hire the best person for a job. And never find out till the probationary period is over. 1:26 Well 'ACCURACY' is not gonna be an issue!
7:30 , this video was worth watching
ha ha I'll try anything once...bring it on
yes I did realize everything I write and say on public forums could be used against me lol.
-"you got 37%"
Jason:"oh that's good, right?"
he did 30 interviews out of which 10 where video interviews.
Jason: " *half* of his interviews where video interviews" xD
Wow is Jason bad at math xD
Time to push for universal income
It raises many ethical questions
wow 😱
We have been warned about AI by many leading tech experts. Their impact will soon be felt. Whether it will be a good or negative impact we have yet to know.
UNETHICAL to an extreme! It will backfire. Not because it is wrong. But because a lot of people are going to turn away from these companies. Companies that respect people more will naturally have more access to talent.
Not in America. It's funny that you think corporations care about you to begin with. They don't care about you, at all. They only care about profit and will exploit you at a moments notice if it helps their bottom line. Nothing will change until we change. Revolution is coming.
I dont think they care. But the most successful ones will probably be the ones that become symbiotic with their employees instead of parasitism.
So i take in your country there are no interviews or that people are not rejected and all companies run at a loss and are owned by governments/people for public service?
Who are you talking to?
Kevin Burris
It seemed different but the whole concept of microexpressions is hard to buy.
Hire Vue is Cyberdine.
10 out of 30 isn't "about half of them" hahahaha at 4:48
next time try saying "exactly one third"