AI as Talent Scout: Unorthodox Hires, and Maybe Lower Pay (nytimes.com) 58
An anonymous reader shares a report: One day this fall, Ashutosh Garg, the chief executive of a recruiting service called Eightfold.ai, turned up a resume that piqued his interest. It belonged to a prospective data scientist, someone who unearths patterns in data to help businesses make decisions, like how to target ads. But curiously, the resume featured the term "data science" nowhere.
Instead, the resume belonged to an analyst at Barclays who had done graduate work in physics at the University of California, Los Angeles. Though his profile on the social network LinkedIn indicated that he had never worked as a data scientist, Eightfold's software flagged him as a good fit. He was similar in certain key ways, like his math and computer chops, to four actual data scientists whom Mr. Garg had instructed the software to consider as a model.
The idea is not to focus on job titles, but "what skills they have," Mr. Garg said. "You're really looking for people who have not done it, but can do it." The power of such technology will be immediately apparent to any employer scrambling to fill jobs in a tight labor market -- not least positions for data scientists, whom companies like Google, Facebook and Amazon are competing to attract. Thanks to services like Eightfold, which rely on sophisticated algorithms to match workers and jobs, many employers may soon have access to a universe of prospective workers -- even for hard-to-fill roles -- whom they might not otherwise have come across.
Instead, the resume belonged to an analyst at Barclays who had done graduate work in physics at the University of California, Los Angeles. Though his profile on the social network LinkedIn indicated that he had never worked as a data scientist, Eightfold's software flagged him as a good fit. He was similar in certain key ways, like his math and computer chops, to four actual data scientists whom Mr. Garg had instructed the software to consider as a model.
The idea is not to focus on job titles, but "what skills they have," Mr. Garg said. "You're really looking for people who have not done it, but can do it." The power of such technology will be immediately apparent to any employer scrambling to fill jobs in a tight labor market -- not least positions for data scientists, whom companies like Google, Facebook and Amazon are competing to attract. Thanks to services like Eightfold, which rely on sophisticated algorithms to match workers and jobs, many employers may soon have access to a universe of prospective workers -- even for hard-to-fill roles -- whom they might not otherwise have come across.
OMG, what a concept! (Score:5, Insightful)
The idea is not to focus on job titles, but "what skills they have,"
Do you ever get the feeling that most business administration is making random decisions, and anything that's slightly better than random, no matter how obvious, is a revolutionary concept?
Particularly for a title as meaningless as "data scientist."
Re: (Score:2, Insightful)
Hiring is the most important thing that an organization does. Yet many, perhaps most, companies are astoundingly bad at it. It is common for companies to give authority to reject incoming technical resumes to a young liberal-arts major in HR with a nose ring and a pierced tongue.
Re: (Score:2)
Re: (Score:2)
I have a paper saved somewhere that looked at the financial decision making spreadsheets for a good portion of the fortune 500 and came to the conclusion that spreadsheets were so hard to debug, and consequently there are so many errors in all of those studied, that companies' financial decisions are effectively random.
Other research finds that higher paid executives are *less* likely to perform well.
It seems odd that competition wouldn't weed out this kind of incompetence.
Re: (Score:2)
Regarding your reason 1, sure, inside a company. But capitalism suggests that some upstart company that isn't a total screwup should come along and eat the other's lunch. But that doesn't seem to happen. Wonder why?
Reason 2: there is is.
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
I think I'm even more cynical than you are. It sounds like a company founded to cash in on the latest bubble (it's got ".ai" in the name after all). They just happen to have stumbled onto something that should have been blindingly obvious from the beginning.
Re: (Score:2)
With this kind of tech, we could predict elections
Re: OMG, what a concept! (Score:2)
Are you serious? (Score:1)
You mean to tell me that when looking for someone to work, you want to actually hire someone that works!?!?!?!
Hm? (Score:2, Funny)
Wouldn't this approach be at odds with the time-honored tradition of required qualifications along the lines of "Must have 10 years experience with technology X which has only existed for 5 years"?
Re: (Score:2)
Must think "outside the skull". (Personally, I suspect zombies would be a good choice for the HR department. Unlikely to be worse than the current lot).
Lower depends on point of view, not for long (Score:3)
It might be lower pay compared to hiring other more experienced data scientists.
But if you are basically taking an almost-data-scientists and hiring them on as one - they are very probably getting what they would consider to be a sizable pay boost.
Not to mention if they turn out to be any good, they are quickly going to be getting large raises to match "real" data scientist, less they get poached by some other company... so that pay being "lower" is either very temporary or non-existent depending on the angle.
Re: (Score:2)
It might be lower pay compared to hiring other more experienced data scientists.
Most likely they are not getting hired instead of the more experienced scientists. Both are getting hired. The myth that there are a fixed number of jobs is the Lump of Labor Fallacy [wikipedia.org].
If more workers become available, unemployment doesn't go up. The economy expands.
Doubtful, the reason is that you can't find both (Score:2)
Most likely they are not getting hired instead of the more experienced scientists. Both are getting hired.
Read the article again. The whole reason they are considering the non-datasci people is - they literally cannot FIND a "real data scientist" to hire! Google et. al. hire them all out of the market with wages smaller companies cannot afford to match.
Sure they probably have room for both, but since they can't find any they at least can have something by hiring in the person that has a good base, just no
Re: (Score:2)
"Read the article again. The whole reason they are considering the non-datasci people is - they literally cannot FIND a "real data scientist" to hire!"
Because, for them, a "real data scientist" is, more or less, like a "devops progammer": an entelechy. I've developed neural networks for behaviour adaptation; I used Haussdorff dimensions to get quick insights on systems changing patterns; I managed systems entropy, a variety of statistical tools (anova, principal component analysis...)... and, still, I am n
Superior to most human HR (Score:4, Insightful)
Re: (Score:3)
I'd be all for it if... (Score:2)
... it put the dain bread recruiters who emphasize job titles out of work.
Not that I ever did but I would have long ago stopped counting the number of recruiters who ask the question: ``What job title are you looking for?'' I'm sure they become completely confused when you tell them that titles are meaningless. I worked in the IT group of a bank's treasury group years ago where anyone and everyone was a vice president. I was once a
Good job, Eightfold. (Score:4, Insightful)
Re: (Score:2)
You must be new here.
This is the second step.... (Score:4, Insightful)
As of now, at least in techy western WA, public school children are issued chromebooks at grade-school level, and are set to use them all the way through high school. Assignments are given, graded, and studied from the google classroom portal. This is giving tech companies unprecedented access to the scholastic performance of our young people, and it scares the shit out of me. I've read the agreement, and it looks like the only concession made is the promise not to send ads to the users while they are on this specific device. I have reached out to the school, the principle, the district IT manager with my privacy concerns regarding this arrangement, but the decision makers love the arrangement, it makes tracking everything easier, keeps the students on task, and provides computer systems to the less fortunate students int he district. None of them even seem to be aware of (or refuse to consider) the possible downstream effects of providing such a complete picture of an entire generations scholastic performance, or the actual value of the massive amount of personal, and scholastic data these systems are providing to a commercial company whos whole model is monetizing this sort of data.
The reasoning I've been provided from the district while working this problem for my own student basically amounts to, "Well nobody else is complaining", and I've been told it is completely voluntary, and I am free to revoke my permission to use the device, which dooms my student to a classroom with no computers at all.
I've given my son permission to use his own device in class, setup a vpn to his workstation at home, and instructed him to use his mobile hotspot for any personal "webbing" he does while at school. So far nobody at the classroom level has taken issue, and they assume I've had success working with the district in securing my student permission to use his own device. As long as he is not caught fucking off in class, or playing AAA games while the rest of em are walled into mathblaster, this should at least last until the end of this year.
I see a future coming into focus that has Alphabet holding nearly perfect information on the perceived abilities of every single body entering the workforce. Algorithms will select candidates for hire (or uni selection) before they even apply, based on perceived abilities gleaned from this arrangement.
I fear that once the effect of such an arrangement is clear, it will be to late to do anything about it.
And yet, many people with experience have problems (Score:2)
So you have stupid things like this at the same time that many experienced people have problems getting interviews, just because the "AI" that recruiters are using skips over many resumes just because of formatting. I know that I've seen and heard from many people who don't get ANY response to their resume, not because of a lack of experience, but because of some mysterious reason. The real problem is when you take humans out of the resume parsing process and you now need a formula resume that is design
Re: (Score:2)
I remember talking to a recruiter and he asked me if I had experience in blablabla, which I did, and he said it was odd that it didn't show up. I pointed him to the relevant section and he said that it looked like I'd worked on yaddayadda.
I told him that blablabla is another name for yaddayadda. He said the software didn't know that, and gave me the advice that if something has synonyms find a way to work all of them in.
Though this was some time ago, things have probably improved since then.