Graeme Dawes - Fotolia
Brandon Wirtz was supposed to be a fifth-generation teacher. Indeed, the founder and CEO of artificial intelligence engine developer Recognant is a teacher -- of robots, not people -- and not the factory floor variety of bots, either. Instead, Wirtz sees AI changing a very human process: human resources.
To reach the place where artificial intelligence and HR meet, Wirtz spends his days educating his various AIs about everything from how to order pizza to what an appropriate pickup line might be. His bots -- "Loki," "Lobby" and "Molly" -- are at different stages of independence and aptitude. Loki, who identifies as female, is perhaps Wirtz's favorite bot -- and the most likely to drive him crazy with questions.
The games Wirtz plays with Loki -- "I Spy" is a particular favorite -- might seem frivolous, but they serve as the basis for the bot's education in how humans think and communicate. And though it may seem unimportant that Loki understands Santa Claus, zombies and Instagram, all of that matters when it comes to artificial intelligence and human resources if she -- or a bot like her -- is going to work in HR dealing with prospective employees. "I know this seems creepy, but it isn't," Wirtz laughed.
There's no fooling a bot
In his world view, Wirtz believes HR is largely broken and AI is going to fix it. "One of the biggest problems in HR is that you have an interviewer, and they know nothing about the particular job they're hiring for," he explained. "Lots of times, an HR person is faking knowledge about the job, so they don't know enough to know what keywords to be listening for."
Even if a bot has never heard of, say, Photoshop, it can quickly search the internet and arm itself with enough information to know if an applicant using Corel Draw may lack the necessary experience, Wirtz said. "The AI doesn't have to understand the conversation but can pass the transcription on to the hiring manager and indicate this was not an acceptable answer," he added. "AI is a way to get deeper interactions with interviewees, and it doesn't matter what they talk about because the AI is an expert or at least a jack-of-all-trades."
A robot might have a more in-depth interview with a job applicant, while showing no bias toward the candidate, Wirtz reasoned. "Sometimes it comes down to 'This candidate reminds me of someone I didn't like in high school,' or 'This person and I have bonded over the same hobby,'" he suggested. "Computers don't have these biases." However, they can be programmed to search for applicant biases, such as racially derogative messages posted on social media. And with the right training, a bot can even help sort out the very subtle human characteristics like emotional intelligence, sense of humor and even ambition, Wirtz said. "An AI is not very good at making jokes, but it can tell when a human has made a joke, and that can help [the bot] decide whether someone has the right personality for the job."
If you're dubious about artificial intelligence and human resources, you're far from alone, but Wirtz has an answer for skeptics. "A well-trained AI will listen, and humans really don't," he asserted. "Say you are looking to hire a test engineer. That's a job without a lot of upward mobility. So you want to hire someone who's going to be happy to stay a test engineer. An AI will analyze how many times the person used the future tense and that's key information that a human would more than likely have missed."
Early signs of intelligent life
Yet that kind of human analysis is only possible when bots are patiently taught by someone who understands the building blocks of AI. Although Wirtz started coding at the age of 7, he took a hiatus from that for several years to be a YMCA camp counselor before returning to software development. He worked on mind simulation products, which provided key training in psychology and understanding how the human brain works. He then moved on to AI-based content creators.
- CEO of artificial intelligence engine developer Recognant.
- CTO and co-founder of search engine optimization company BlackwaterOps.
- Standards body voting member of the Society of Motion Picture & Television Engineers.
- Grew up on a farm in Michigan, where he learned to fix "physical" things like soldering, welding and changing car parts.
- Learned to type in kindergarten and began programming shortly after that; received a visit from the FBI in third grade class in '87 investigating how he encrypted an "unencryptable" hard drive.
- In high school, developed an interest in artificial intelligence for applications ranging from video games to flight engine simulations.
- Wrote landmark programming scripts during the Y2K era.
- "Wrecked my teeth" practicing kick-boxing, self-proclaimed bad golfer and fly fisherman, and an avowed Mountain Dew Throwback addict.
- Fathered three bots -- "Loki," "Lobby" and "Molly" (by golly).
"I was trying to 'game' Google and create content for fun and profit that wasn't great but didn't suck, either," he acknowledged. "So I learned about content creation, fact extraction and knowledge building." Those were the tools that formed the basis for Loki and her AI colleagues at Recognant and what Wirtz hopes will foster a new approach to artificial intelligence and human resources.
Thanks to AI, applying blindly for a job will become a thing of the past, Wirtz predicted. Instead, a chatbot will appear right after the application is submitted online and strike up a conversation with the applicant. That prescreening exchange will provide the hiring manager with immediate information about the applicant; the applicant may receive feedback from the hiring manager as well.
"From spelling errors to subject knowledge to attitude, an AI can ask and record this exchange," Wirtz said. "If this works, we can get rid of the biases, the emphasis on education instead of experience and a lot more unknowns. And it really is starting to happen today."
Does AI really have a role in business?
Do you need HR management software?
A primer on AI in HR
AI tools are intelligent but not all-knowing
- Building the Right Management Team for Growth –Oracle Corporation
- How to build an effective User Acceptance Testing (UAT) Team –ComputerWeekly.com
- 6 Key Challenges to Building a Successful Data Team –Dataiku
- Top 3 Challenges Your Team Faces when Building in the Cloud –TrendMicro