A.I. (artificial intelligence) has been a topic of conversation in the HR world for a long time. We want to automate some processes, but we also have employees who worry about their role if some processes are automated. What happens in recruitment if screenings are done by AI? What if staff are able to get those easy benefit or leave questions answered by AI and not a person? What if we didn’t need someone to process paychecks? We are in the adaptive age and we need to help our employees not only be ready for these changes, but embrace and adapt to what their work can and will be.
At SHRM19, I was able to sit in on Ben Eubanks‘ presentation on AI. He has written a fantastic book “Artificial Intelligence for HR” which is a wonderful deep read on how we can use AI to support and develop a successful workforce.
We like to think of AI as “smart,” but its not. AI does exactly what we tell it to do, whether we think we are or not. Think about Netflix or Youtube – they learn what you like and make suggestions based on what you’ve watched. Think Google suggestions. Google has learned what you’ve (and others) searched for in the past and uses that to make suggestions as to what you might be looking for this time. But it’s not able to go beyond what you or others have done.
AI promises 3 things to add value to your day to day work:
AI is not just about doing things faster – we’ve all been a part of (or heard of) an implementation of technology that didn’t improve the process, it only took the current labor-intense process and made it faster. It might have automated some of the processes, but without the personalization or augmentation, we are merely doing a bad process fast. While you might be happy for the extra time, you will eventually be frustrated that the process isn’t better.
Think about your day to day work and what would you automate and how? During the session, many people brought up scheduling: meetings, interviews. How much time is spent attempting to coordinate calendars? We talked about that initial prescreen of candidates for minimum qualifications. When I think about how many phone screens I’ve done with candidates who aren’t able to work the schedule of the position posted. What if we had a chat bot to walk candidates through some of those initial questions? If there was a way for the candidates to easily input and update their availability and AI to go find the positions that match their qualifications?
In 2017 at SourceCon, they had a contest between humans and a machine to see who could better source candidates for 3 positions. And a human won. In 25 hours. The machine came in 3rd, but got very similar results in mere seconds. So before we celebrate and cheer the humans – note the humans DID use some machine work to whittle down their possible lists. I think this is the key to AI – it is not all or none. Let’s use a machine to narrow down candidate pools so we are focused on the right candidates and not spending 20+ hours on those who aren’t.
The other area we are seeing more AI pop up is in the use of Chatbots. A chatbot can take the place of a human answering questions on line, that initial line. A chatbot can answer 80% of the questions that come into HR. Think about the questions that come into HR that are currently answered by a human:
- Where do I find the form for FMLA?
- How do I request PTO?
- What is the deadline to sign up for PTO?
- Where is the job application?
- What jobs do you have that I would be qualified for?
The chatbot allows for a level of personalization that can be hard when 100s of calls are coming into HR on a daily basis. People are actually saying Thank You to the chatbot at the end of the conversation! This helps create a comfort level with the organization, a familiarity that starts the engagement from the very beginning of the employment relationship. Now the chatbot cannot completely replace the human in recruiting. Ben’s research found that the closer you get to the actual hire, the more people wanted a human on the other end. And even more interestingly, the further away you get from the hire date, the more employees were okay with talking to a chatbot.
In addition to the concern around robots taking our jobs, there is a lot of concern around the bias that shows up when you use a machine to find candidates. But the machine is stupid. The machine is going to give you exactly what you ask for. And we, humans, are biased without even knowing the biases we have. We inadvertently put these biases into the machines. We need to work with those developing the AI we want to use to ensure these biases are accounted for and mitigated.
Ben shared a lot of examples, both good and bad, of AI in the workplace. We can and should use AI to eliminate what we can and enhance the rest. Task-oriented, repeatable tasks are the best tasks to give to the machine. The soft skills, that are not repeatable, that involve a curiosity, are not done well by a machine and need to stay with the human.
Spend time developing the right skills in your workforce to be ready to adapt to the new workplace that AI is taking us. Encourage your staff to learn & grow!
Want to learn more about AI? Check out Ben’s book and his Twitter feed. There is seriously good stuff there!