CHICAGO—What does artificial intelligence (AI) mean for the future of recruiting?
About 90 percent of the tasks recruiters do will likely be automated in the years ahead, but that doesn't mean that talent acquisition professionals will become extinct. What they spend their time on will change, said Kevin Wheeler, founder and president of the Future of Talent Institute, a San Francisco-area think tank. He was speaking at CareerBuilder's Empower 2017, a conference for recruiting professionals held by the recruiting software company.
"AI is at the top of the hype cycle right now," Wheeler said. "You can ask Alexa what the weather is and it tells you. But all it did was look up the weather report online, turn that data into words and relay it to you. AI will get better, but right now, AI tools are really stupid. No recruiter should fear that they will lose their job within the next five years. Some low-level, repetitive jobs may be in danger, but we still need people to make decisions."
How AI Will Improve Hiring
AI is an umbrella term for machine learning and using algorithms, Wheeler said. "It's a tool that takes a huge amount of information and crunches it into something we can use. In the hiring context, AI will be able to calculate the probability that one person will be a better hire than another person by looking for patterns that come closest to the criteria you input."
That's basically what recruiters do in their heads now when considering candidates, he added: weighing some attributes more than other attributes, maybe at a subconscious level.
It's in crunching unstructured data where AI is beginning to make itself useful. "We all use structured data, like numbers on a spreadsheet that can be counted and multiplied," Wheeler explained. "If I say 'we interviewed 33 people, made 12 offers and had four accepts,' that's expressing structured data. But being able to look at what someone says or writes and understand it and make decisions based on it, such as in reviewing resumes and online profiles—that is using unstructured data. That's where AI is attempting to do better than we can."
AI can also process data much more quickly than the human mind, of course. But it needs a lot of data to be of value. "The more data you have, the better your decisions," Wheeler said. "You need thousands of data points to get accurate predictions. If you're only using hiring data from 100 or 150 people, you're getting highly skewed results."
How AI Will Affect Recruiting
By 2025, AI will be more integrated in recruiting, Wheeler said. "We're at a tipping point, where we'll begin to see a lot happening quickly. The tech won't be that great yet, but it will be there." Some jobs will disappear as more recruiting functions are automated, he said. Sourcing, screening, assessing and reporting are already highly automated today. Administrative work will certainly be automated in the future. "But how many people like to do admin work?" Wheeler asked.
Chatbots are an example of emerging AI on the market. The most basic chatbots are useful for initial candidate interactions and can improve careers site retention for example, but most are "really stupid," he said. "All they do is read a list of FAQs. The average chatbots can schedule interviews and screen candidates with basic questions. More sophisticated ones can match candidates and do basic search."
AI will enhance workforce planning and make using predictive analytics routine. That's where AI can be incredibly helpful, to find patterns among data, Wheeler said. "AI will transform recruiting from largely a transactional function—scheduling, interviewing, clerical routine stuff—into something that truly adds value to the company."
But will talent acquisition ever become 100 percent automated? "I don't think so," he said. "Recruiting will go through what the travel agent world went through 15 or 20 years ago. When you booked travel in the past, you called up a travel agent. Today, it's all pretty much automated. But let's say you wanted to take a complicated trip around the world, with multiple stops—you'd still use a travel agent. Because what you want is highly complex. It's the same thing for recruiting. If you want to hire a retail sales clerk, you'll push a button and it will be almost 100 percent automated. But if you want to hire a CEO, I don't think so."
AI will lead to more jobs being atomized—separating one role out into several more distinct roles, he added. "Recruiting is already becoming atomized, for good and bad. Less and less recruiters do it all anymore. Today you can just be a sourcer, or work in branding or screening and assessments. There was a time when a recruiter did all of these things."
So let the automation work for you, he said. "Let it do the transactional stuff, schedule the interviews, chat with candidates initially. That's good. Let it. Don't waste your time. Embrace it."
Wheeler said recruiters should allow AI tools to do what they do best while talent professionals focus on the human side of the job—building networks, devising strategy, collaborating, communicating with candidates. "Those are the skills to have," he said. "And those are the skills AI is not good at. The question is, how do we use the intelligent augmentation to help us become better recruiters?"
The Rise of Robots Will Lead to Complicated Questions
AI has the potential to enhance the hiring process, but it must be used with caution and human judgment, Wheeler said.
For example, it has the potential to help phase out human bias, but there are some issues with using algorithms that must be addressed. Machine learning relies on humans to input the data and craft the algorithms. "Algorithms can sometime reinforce human biases," said Jed Kolko, chief economist for job search engine Indeed, speaking at Indeed Interactive in May.
Was this article useful? SHRM offers thousands of tools, templates and other exclusive member benefits, including compliance updates, sample policies, HR expert advice, education discounts, a growing online member community and much more. Join/Renew Now and let SHRM help you work smarter.