Why AI Fails At Assessing Distinctly Human Skill Sets

It’s safe to say that AI recruitment tools are a major topic in the HR space right now. While they’re praised for their ability to reduce the time and expense involved in the recruiting and hiring process, currently available AI is also notably imperfect. As smart as current AI can seem, it’s also easy to manipulate. There are countless websites and forums devoted to helping candidates game the system and zoom right past the AI’s algorithms. Their known propensity to pick up and then deploy ethnic and racial biases is troubling, to say the least.

There are specific things a human recruiter can simply just do better because, to put it simply, they’re human. AI doesn’t see the person behind the résumé and cover letter. Consequently, that means AI won’t do well measuring the strength of the soft skills human recruiters most need to assess to determine if someone is a good long-term fit for a role.

The Top Soft Skills Recruiters Need To Assess

While there’s no definitive list, most employers are seeking to hire or train workers who possess listening skills, attention to detail, communication, critical thinking, empathy (often in the form of teamwork or interpersonal skills), and an ability and willingness to learn.

Understanding what these skills really mean and what they look like during the recruiting process can lead to better results in hiring, and that means taking a step back from relying on AI to do the hard work.

Active Listening Skills

Listening is about more than just the ability to repeat words and phrases. If that were all that listening skills required, one could easily use AI software to throw a quick test where applicants listen to recorded phases and repeat back what was said.

Instead, what companies need are workers who have active listening skills. That means they can listen to and understand others, interpret what was both said and what was left unsaid, and then reflect that back with phrases like, “Here’s what I hear you telling me, is that correct?” The thing with listening skills is that they require seeing and responding to both verbal and nonverbal cues. While AI tools might be close enough to sentience that they can express emotion, we’re not quite there yet.

Attention To Detail

Similar to listening skills, attention to detail is more than just being able to identify specific details about an objective or phrase and repeat it. It’s about the ability to pick up on and respond to exceptionally small clues and details, often related to interpersonal relationships. Importantly, attention to detail generally requires understanding and responding to ambiguity. Dr. Robert J. Marks covers this topic well in his 2019 Mind Matters News article. As he explains, even the word “it” can trip up an artificial intelligence system when the subject of a sentence is unclear (something we humans do naturally in both written and spoken language).

Communication

Can you hold a conversation with an AI? Sure. Those of us old enough to have lived through the ’90s will remember chatbots available in AOL instant messenger, like SmarterChild. Now, chatbots are deployed throughout business customer service arms worldwide, to varying degrees of effectiveness. But while AI can certainly respond to questions in a human-like way these days, it can’t assess the quality of someone’s ability to communicate. Unlike a human recruiter, AI can’t evaluate the exact word usage, tone of voice and style of communication that could matter to success within different roles.

Critical Thinking

Now, one thing AI can do is evaluate whether someone answers questions right or wrong—even exceptionally difficult and complex questions. What AI can’t evaluate is how effective someone is at seeing, analyzing and solving their way around different problems. This harkens back to AI’s inability to understand ambiguity. Critical thinking is an aspect of looking at a problem with exceptionally ambiguous parameters and often using creativity to develop a response or answer. It’s about more than just getting an answer. It’s about how that answer was solved. AI simply can’t assess the value of getting from point A to point B.

Empathy

This one almost needs no explanation. Empathy is purely based on emotion and understanding. Although there is a scientific aspect to empathy, and some AI systems can be taught to mimic empathy, the science of empathy isn’t what matters. It’s the person-to-person application and assessment of it during the interview process that matters.

A Willingness To Learn

An AI recruiter can be programmed to send a questionnaire. One could even send a questionnaire that asks, “On a scale of one to 10, how would you rate your willingness to learn?” What recruit—who actually wants to get hired—is going to say anything less than a 10? You might get the occasional recruit who will, in an act of humility, give a nine. But ultimately, AI won’t be able to assess the reality of that response. Human recruiters are needed for that.

Human Recruiters Excel At Soft Skill Assessment Because They’re Human

AI recruiting tools are not a replacement for human recruiters. They’re simply that: tools. AI recruitment tools are ill-equipped to assess the increasingly important, soft skills that lead to better productivity and better retention. In short, it’s better to leave what makes us distinctly human, to human assessment.

I agree that one day AI could very well be able to do everything promised, but I personally believe we’re still five to 10 years away from this possibility. Even Siri, who learns from 25 billion requests each month to provide predictive recommendations and information, doesn’t always get it right. Siri also has an advantage over the AI used in most HR technology in that it learns from our voices rather than an untrained voice, which is much less reliable. If the world’s best technology, designed by the world’s top engineers and supported by endless funding, has not adequately figured out machine learning, then I don’t believe HR technology is at a stage where it’s ready to be leveraged in the hiring process.

 

Original post: https://www.forbes.com/sites/forbeshumanresourcescouncil/2022/09/12/why-ai-fails-at-assessing-distinctly-human-skill-sets/?sh=1fa35a305ba8&s=09

Leave a Reply

Your email address will not be published. Required fields are marked *