By Brian Thomas & Laura Caty
Imagine seeing your dream job posted on a company’s website. You throw caution to the wind and submit an application. Almost immediately, a Chatbot pops on your screen with a series of automated questions designed to verify your interest in the job, your job history, and your available start date. After keying in your answers, an algorithm sends you a series of predetermined questions. You submit your answers and then receive a short logic game, which is supposed to judge your personality. After playing the game (and without ever speaking with a human) you’re sent a form offer letter inviting you to start your new job next week.
Does this sound far-fetched?
It’s not. Amazon tried to automate its hiring process in 2014. Their goal was to create a streamlined, efficient hiring system. But automation comes at a price. Since Amazon historically hired more men than women, its algorithm minimized women’s resumes and prioritized men’s. The algorithm even penalized resumes that included words like, “women,” “women’s chess club captain,” and even dinged applicants for attending women’s colleges. In seeking efficiency, Amazon didn’t appreciate that computer programs are created by humans and the programs often reflect the biases and prejudices of the programmers. So does that mean we’re going to give up on AI?
Probably not. An IBM survey found that almost 70% of the CEO’s it surveyed believe that AI will significantly increase the value of human resources. Many still believe AI will reduce bias, help solve the talent crisis, and improve candidates’ experiences. It’s not just talk. In 2011, companies invested $282 million in AI. That number jumped to $2.4 billion in 2015. It’s a good bet that this number will continue to increase. Smart employers should think about when and how they use artificial intelligence.
A logical first step is to recognize the inherent limitations of artificial intelligence. AI can never replace human instinct and intuition. Ben Eubanks, a leading voice in AI research, says the critical difference between humans and machines is that computers can quickly analyze massive amounts of data, while humans excel when it comes to creativity, curiosity, collaboration, compassion, and critical thinking. One reason successful companies allow multiple team members to meet with a job candidate is it produces a more informed, comprehensive, and realistic view of the candidate’s skills and personality.
Think about it –a candidate could drink too much during a lunch interview, berate the waiter, chew with his mouth open, and tell a bunch of inappropriate jokes. No reasonable hiring manager would hire this candidate. Could artificial intelligence make this same assessment? Can you really know how a candidate will interact with people without observing how she actually interacts with people? Can an algorithm really tell you whether someone will be a great (or terrible) employee?
We need to answer these questions. Companies will continue to use AI to increase productivity and efficiency and many will also use AI in their hiring decisions. Maybe one day artificial intelligence will catch up to human creativity, compassion, and curiosity. But we’re not there yet. So until that happens, let’s keep the human in human resources.