Artificial intelligence (AI) in HR is a very hot topic. But is it effective in helping HR professionals find suitable candidates, or match employees to opportunities? In a recent browsing session for talent management software, I ran across an article about a woman who ran into challenges applying for a job where AI was used to filter ‘qualified’ candidates. The crazy AI solution that stood between her and her next job seemed to operate without any common sense. Only after she employed some novel and counterintuitive tactics was she finally ‘seen’ by this organisation. Not an ideal situation, and not a confidence booster for AI in HR.
HR Leaders who combine rigorously trained AI-based models with human judgment can recruit and upskill people for professional career growth and the future of the business.
gettyAI’s role in recruitment is not to thoughtlessly reject candidates. Applied correctly, meaning with the right human touch from an HR perspective, AI and in particular a subset called machine learning (ML) can support human decision making. Finding the right balance may yield the desired results many organisations are looking for when they attempt to use AI in HR.
What is ML and why can it be problematic?
SAP defines ML as “focused on teaching computers to learn from data and to improve with experience – instead of being explicitly programmed to do so. In ML, algorithms are trained to find patterns and correlations in large data sets and to make the best decisions and predictions based on that analysis.”
The main potential problem with these technologies lies with the data that people use to train the AI model. As the old saying goes, ‘garbage in, garbage out’. If the data is biased or incomplete, then the results or ‘advice’ that the AI gives will be equally biased or incomplete. There is a famous article about judges and law enforcement professionals using AI-based risk assessment scores to determine if someone would stay in jail while awaiting trial, or perhaps whether that person should be investigated. But like many applications of AI, they failed to understand the difference between correlation and causation. The result was misery and injustice perpetrated on a scale only AI could support.
Building a better workforce future using AI and ML
Despite the stories of mishaps and anomalous results, it’s increasingly clear that AI can be very effective in matching employee skills to opportunities and jobs. To date, these match ups were just clever algorithms that performed a one-to-one match between, for example, competencies assigned to a job role versus competencies attributed to an employee. However, the talent landscape has evolved over time, and now there are many more career and employee-related variables to consider. Managing the interplay of these variables between someone’s capabilities and job opportunities is getting so complex that humans tend to miss key skill sets while trying to fill project rosters. Budget and increasing time pressures only add to the need to automate. HR leaders and experts need to focus on choosing the data points and variables that make up a rich and useful data set for the AI to use. Leaders also need to continually monitor the output, perhaps surveying employees to make sure the matching and interactions are productive. With this approach, AI can be a great ally supporting career development and building a better workforce.
The role of AI in attracting top talent
A recent Gartner study showed that 17% of organizations use AI-based solutions in their HR function and another 30% plan to do so in 2022. There’s no question companies are using AI that continuously learns from interactions with HR professionals to support human decision-making. For example, SAP SuccessFactors will introduce machine learning in its Opportunity Marketplace. Other solutions like Gloat and Phenom Gigs already use this technique to match employees with opportunities in organisations.
The takeaway is that if organisations are careful with the input, and monitor the results from their AI, the output could potentially provide more career options for employees. The tool could help match someone to an opportunity that is a perfect fit, perhaps even for skills that the employer didn’t know the employee had. Recruiters will enjoy more and more qualified candidates automatically matched to increasingly complex job roles. This could translate into happier, more fulfilled employees and reduce voluntary turnover, saving organisations big dollars. I, for one, embrace our new job hunting, skills screening overlords. After all, they may very well hold the keys to my successful future.
