The role of artificial intelligence has been steadily increasing and there’s no shortage of experts hailing it as the next big transformation in the way Human Resources departments are run.
There’s more than a grain a truth in those predictions because any significant tool does affect business as usual.
But I’ve written before about some of the problems with using AI in recruitment, including the risks of perpetuating rather than eliminating some of the bias problems that have plagued hiring processes.
Aside from recruiting, some companies are looking at the potential of AI for improving the employee experience, mining worker data to optimize potential perks. Those kinds of programs offer promise, but as I’ve mentioned it’s important to keep an eye out for specific issues.
A recent report about an IBM program now used to reverse engineer job market challenges, and thereby identify overlooked talent, represents the next chapter in the ways AI could impact HR. The idea of the program, called SkillsBuild, is to find potential recruits in disadvantaged groups by focusing on skills rather than background, education or pedigree.
The idea of focusing on skills rather than education is a novel approach.
Obviously, no true Human Resources professional ever made a hire of a candidate with a certain educational background for the sake of that background alone. But a candidate’s education is taken as an indication of hers or his critical thinking ability, drive and affinity for learning.
Unfortunately, there’s the risk of allowing certain kinds of degrees to signal other aspects of a candidate’s background — including pedigree, which is how a preponderance of those with privileged backgrounds lead to homogeneity in the workplace.
Targeting for training
One of the benefits of the program is that it doesn’t set out necessarily to find attained skills within overlooked populations. Rather, it looks instead for the kinds of traits that would enable success and then targets those candidates for further training.
The program relies on a certain kind of candidate profiling that bypasses traditional indicators of attractiveness. “It takes a full assessment of the individual’s cognitive abilities, and creates a comprehensive competency profile that’s unique to them, then uses algorithms that match them to jobs that require different competencies,” Denise Leaser, president of the program, told CNBC. That includes characteristics such as empathy and deductive reasoning.
Meeting a need
One of the strongest promises of this program is that it focuses on segments of job-seeking populations that may not otherwise be plumbed for competitive candidates.
The first part of the program is being unveiled in Europe, with an emphasis on asylum seekers and other disadvantaged groups.
It’s a particularly effective approach due to the shortfall in competitive candidates for an increasing number of tech jobs.
AI: as good as its engineers
As with any artificial intelligence program, though, it’s important to remember that AI is susceptible to systematizing the unexamined biases of its engineers. That means it’s crucial to submit programs to an iterative assessment process to identify and correct failures.
It’s also critical to solicit input from people with a broad base of perspectives when developing AI programs. But as the IBM program shows, it’s a shift that holds as much promise as it does risks.
Source : https://it.toolbox.com/article/why-recruiters-need-new-specialized-crm-platforms