Maren on AI
I’ve found myself struggling with the concept of AI lately. I have several clients who are working with the concepts of Artificial Intelligence, which means I am talking about it, and writing about it and having to relate these core concepts to more than just my encyclopedic knowledge of Terminator 2.
It’s generally admitted that AI is this year’s Big Data, meaning a large concept that we’d all like to understand but few rarely do. So, without setting myself up as an expert, just someone who’s been grappling with concepts far above my pay grade, I’d love to dismantle some of the myths and legends that surround using artificial intelligence for recruiting and hiring.
I invite debate around these statements in the comments!
Myth #1: AI and Automation are Samesies
I suffered under the delusion that AI was the same as automation for awhile. Why? Because no one corrected me and like anyone else, I tend to be taken in by what vendors say. But no more!
Automation is wonderful and most of us use it in recruitment, hiring, marketing, even sourcing. But it’s not the same as AI.
Automation does the same thing over and over in perpetuity.
For example, you create a sourcing string to scrape LinkedIn every day at 8:00am. So long as you never turn that off, it will do that forever. If the word “congratulations” is in your string and all the other indicators are ticked, the string won’t know (without human intervention) whether that person got a brand new job or just had a baby.
A calendaring software that allows hiring managers to schedule appointments quickly and blocks off selected times on two calendar is very convenient, but it’s not AI. It’s automation. It will block off that time, even if your personal calendar indicates you have kickboxing.
Brian Delle Donne, of Talent Tech Labs and Mitchell Martin, had this to say in a recent article trying to clarify talent acquisition and AI’s role therein:
At a basic level: automating process through computerization, for example creating automation that fires off different work flows are labor saving and even intelligent. Taking it one step further, using algorithms to find correlations and other relationships like making matches or triggering a response is also intelligent.
However, in these examples of using technology to mimic intelligence we are not using Artificial Intelligence, but instead, building intelligence around what are already known systems, known behaviors and generally known outcomes.
Real Artificial Intelligence (yes I realize the term is an oxymoron) learns.
That’s it. That’s how you tell what AI is and what it is not. Think of how IBM’s Watson beat Ken what’s his face at Jeopardy.
Myth #2: The Bots Will Take Our Jobs
As I write this, we are sitting in the aftermath of a contentious election cycle in the United States. For many, a motivating factor in their vote was “bringing back our jobs”. It is into this climate, that AI makes its tentative debut for the white collar worker (with some notable blue collar exceptions, like truck drivers). Watson is attractive when he’s beating some smug know-it-all on TV, less so when he’s taking jobs from the already smarting republic.
But this too is a myth. While true some jobs are gone for good, and more jobs will follow, there are jobs being created too. But perhaps “artificial” is the wrong word for what we’re seeing. Chris Cancialosi writes:
IBM prefers the term ‘augmented intelligence’ over artificial intelligence. They believe that rather than computers taking over for humans, cognitive technology will serve a critical role in augmenting the humans it supports. This augmented intelligence will create unparalleled opportunities to draw out the full capacity and potential of the human spirit by relieving people of transactional and mundane tasks that take up so much time today.
It’s a real fear. From Skynet to well, mostly just Skynet, we’re all a little afraid that what we create will rule over us, if we do too good a job. Elon Musk himself is pouring over a billion dollars into an initiative designed to make the research we’re doing safe.
Most people, Musk included, believe while AI will make some jobs obsolete (as technology has since time immemorial), it will also create more strategic jobs, more interesting oversight and potentially more exploration of work. CEOs believe around 5% of their current workforce will be replaced by AI; however, this stat doesn’t take into account the people who will manage these systems and use the findings they reveal to make changes within the workforce.
An HRTechWeekly article cites Stanford and Deloitte as sources in this predictive paragraph:
Rather than AI leading to a jobless future, the 2016 report from Stanford University’s One Hundred Year Study on Artificial Intelligence suggests that AI will be regarded as a ‘radically different mechanism for wealth creation’ replacing ‘tasks rather than jobs’ and leading to the creation of new types of jobs.
To give this some context, a reported 60% of existing retail jobs have a ‘high chance’ of automation by 2036 but a new sector of e-commerce has emerged in response to this change. As predicted by Deloitte, high risk jobs are being replaced by more creative low risk jobs with each new job paying a salary £10,000 higher than the one it replaced, in the process adding £140 billion to the UK economy. This shift is also apparent in the rising demand for specialist tech skills in areas like data analysis across all sectors.
So the bots will take many of our SUCKY jobs.
Myth #3: AI, like Big Data, is too unwieldy for recruiting
Okay, maybe no one ever said that. But based on the number of articles dumbing down both concepts, it’s clear many people think AI (and Big Data, those of you reading this in 2015) are too BIG for our industry.
Wrong. We do need to understand these things and even innovate in their direction. Recruitment and its staid older sister, HR, shouldn’t be pulled kicking and screaming into an era where we don’t understand the trends and shifts impacting our roles. Here are some companies who are actively leveraging some form of AI in recruitment and talent acquisition:
- Mya: A chatbot that guides candidates through the application process, Mya asks about skills and administers assessments, among other things.
- Wade & Wendy: Wade and Wendy chats with both recruiter and candidate to find better fit through the application process and promises to “build better relationships”.
- Textio: A “predictive engine,” Textio analyzes job listings and hiring outcomes data to find meaningful language patterns for those writing job descriptions and advertisements.
- Karen: Karen combines the chatbot innovations listed above for both recruiter and candidate, with candidate rank to analyze resumes submitted to a job posting and compares them against the job description, company, team personality, and culture fit.
There are companies who are leveraging the advances in AI for YOUR job, today! Not just automation, but true learning. Karen founder Noel Webb, has leveraged IBM’s Watson and helped it to learn recruitment. Textio has created a case for more diverse language in recruitment marketing and been used as a bedrock for understand language that attracts and deters applicants around the world.
What are some of your thoughts around using artificial or augmented intelligence in recruiting?
via HR Examiner http://ift.tt/PgrQXS
November 27, 2016 at 10:35PM