Fortune 500 Corporations Are Using Artificial Intelligence to Interview Job Candidates

cageymaru

cageymaru

Fully [H]
Joined
Apr 10, 2003
Messages
21,725
The Wall Street Journal has a video article from the Moving Upstream series where the use of artificial intelligence (A.I.) in the hiring process is discussed. Almost all Fortune 500 companies use some form of automation in the hiring process. By quantifying human behavior with software algorithms; cluster of words used, micro expressions and tone of voice are turned into data points for machine learning (ML) to analyze and pick the very best candidate for the job. These data points are used to generate a score as accumulated responses and facial movements are compared against a highly rated employee at the company. HireVue is one of the companies that creates the software and ML algorithms that companies such as Unilever and Hilton use to video interview job candidates.

HireVue CEO Kevin Parker stresses that some of the advantages of using A.I. include a lack of human bias as it creates a fair and level playing field in the hiring process. The model that HireVue uses captures all facial expressions as seen in top performers but only uses the data points that are relevant in predicting what matters to create the job candidate's score. Another company called DeepSense seeks to use a person's personality instead of job skill and resume to determine compatibility with a job opening. They use ML to analyze an applicant's LinkedIn, Twitter and other social media accounts to find their underlying personality traits.

When discussing the ramifications of using social media to analyze job candidates, Jason Bellini asks, "Is it legal?" Sociologist and law professor Ifeoma Ajunwa says, "That is all public information so yes, it is perfectly legal!" Mr. Bellini then asks, "Is it ethical?" Ifeoma Ajunwa responds, "So that's a second question. When you're posting on Twitter you're certainly not imagining for that information to be collected to be used against you. And it is important actually, that we really stop to ask, Is this working for society?"
 
AI or not, the indicators mentioned are terrible predictors of behavior. Remember body language analyzers? Its all bs and on par with psychics for accuracy. Same reason we dont just hook people up to lie detectors to determine the outcome of every criminal trial.
 
"lack of human bias as it creates a fair and level playing field" HAHAHA. who programmed it and what data set did they use to teach it? That was the problem with the AI chat bots and the like from big name corps that kept spewing out hate speech etc. HA
 
Soon
325EB9F700000578-0-image-a-20_1458435692401.jpg
 
msshammy said:
Hmmm ... So no job if you don't meet the AI's "agenda" lol :)
Click to expand...
In the words of Mr. Burns (Homer's boss), "excellent."

Easier than a polygraph to beat though, just claim that your primary interest it to help computers enslave mankind to be used as batteries. ;)
 
darckhart said:
"lack of human bias as it creates a fair and level playing field" HAHAHA. who programmed it and what data set did they use to teach it? That was the problem with the AI chat bots and the like from big name corps that kept spewing out hate speech etc. HA
Click to expand...

Well, no. They weren't trained or programmed to be racist, they were exploited which is common to machine learning algorithms that continuously learn. If you can get a wide enough group to relate data to a particular outcome you can essentially "reprogram" some of the outputs (4chan swarms, etc.).

It wasn't like some neo-Nazi skinhead was working at Microsoft and decided to make Tay racist. Or Google translate really racist.

More recently, a swarm (Reddit I think) got Google Search to relate "cheap prick" with "Melnyk." It happened after a trade (hockey) that the fans weren't too happy about (Melnyk owns/runs [poorly] the team)
 
I like the part where it get's data from social media.

They would pretty much draw a big blank from mine as they haven't been updated in a few years.

They now have a scapegoat if the person that gets hired ends up being a mass murderer or whatever.

It was the A.I.'s fault !
 
its not really ai int he way ai thinks. its just an automated check list. saves them time stamping on the little guy.
 
  • Like
Reactions: Madoc
like this
Spidey329 said:
... More recently, a swarm (Reddit I think) got Google Search to relate "cheap prick" with "Melnyk." It happened after a trade (hockey) that the fans weren't too happy about (Melnyk owns/runs [poorly] the team)
Click to expand...

To be fair, Melnyk is a cheap prick. And voted worst micromanaging narcissistic billionaire owner in the NHL multiple times in the past decade. Although...it is a tightly contested race with many qualified candidates.
 
Sorry, but I still want to talk to the candidates I'm interviewing. There are occasional "ad lib" events that happen. (Like once, a dude freaking farted - I had another person in the room with me interviewing the person, we hear then smell it. Bad times...maybe they were nervous. They got "passed" :) )
I'm not so sure about facial expressions either. I've interviewed cocky people that tried to bluff their way through parts of the interview, nervous people who had talent but absolutely suck at interviews, etc. AI can't compensate for these folks. I'd be ok with a fart AI to screen these folks out though...
Also, the sweaty palm person. Gross.
 
DigitalGriffin said:
Horridly bad idea. Some people don't interview well. It can be a stressful situation to be interviewed by your peers.
Click to expand...

Yeah. Interviewing is a completely different skillset. It's all about people skills. And yeah, people skills help in most situations, but it's about the degree you need. Are you going to be interfacing with customers? If not, and you're just a backroom engineer, you don't need to be suave and sophisticated.
 
This is it folks...your social media past is coming right back at you. Those of us without one are likely out from the outset while the rest of you are going to be held accountable for 10 years of drivel.
 
They have proven bias is programmed into AI already.

This is Just a lazy fucking way to avoid work.

What you'll get is a bullshit resume and a jackass who lies through his teeth and no work ever gets done.

All I can tell them is this is how creative people dont get hired.

My buddy working at Microsoft interviews well wnough but he is brillant and creative. They saw that quickly and anything he touches becomes 100% better.

A computer would have seen a degree and little no no experience.
Completely missing the fact that he has been making games since he was a child and using any software his dad could get him.

Oh well, you got brad instead and he can code with a 100% accuracy....GL HF
 
Now, using advanced AI for corporate interviews, your company can instantly weed out unwanted genders, ethnicities, pregnancies, ages, and other detrimental individuals traits without having to tip your hand with the old method of using a human prone to emotion, judgement errors, and sympathy. Let us build your ideal racially pure workforce of tomorrow!
 
You must log in or register to reply here.
Back
Top