Companies know the importance of finding the right person for the right job. And with talent shortage the top concern for employers this year, many companies are using some form of AI to keep step with the ever-changing digital landscape and how they access talent.
At the same time, there are well documented cases of AI programs displaying signs of inherent bias. And many of the innovations emerging for talent identification are still very much a work in progress. So, to what degree can AI-based hiring tools help reduce bias in our talent identification processes? Let’s take a closer look.
Talent crisis and the impact of proper talent
It is hugely important that organizations find the right person for the right job; McKinsey & Company found that superior talent is up to eight times more productive. But going into 2019, businesses were increasingly concerned about a serious talent gap.
For organizations worldwide, talent shortage moved to be the top emerging risk in Gartner's latest “Emerging Risks Survey”. Moreover, a new Conference Board report identified the talent shortage as one of CEO's biggest worries for 2019, and a recent XpertHR survey cited high-quality talent as the top challenge for HR professionals this year.
With mounting pressure on businesses HR team’s talent acquisition methods and processes, there is a strong business case for innovating with AI. Of course, this comes with some important considerations.
AI & bias
There are many salient cases of AI programs displaying signs of inherent bias. Think, Amazon’s recruiting system not rating candidates in a gender-neutral way, or Microsoft’s AI chatbot learning profanity, racism bias and inappropriate language, which it then began using.
Machines are fed massive amounts of data and instructed to identify and analyze patterns. Ideally, these patterns produce an output of the very best candidates, regardless of gender, race, age or any other identifying factor.
But as AI systems, often based on real-life data, begin doing as they’re trained and start making decisions, the prejudices and stereotypes that existed in the data are actually amplified.
When real-world data has patterns exhibiting bias, the AI algorithm will eventually incorporate this into its functioning. So, it’s important that hiring managers and recruiters evaluate the data. For example, the number of candidates, who was picked, interviewed, the outcome, and why. We can also address this by teaching AI to predict relevant and objective outcomes, rather than mimicking human intuition.
Benefits of AI-talent tools
In recruitment, artificial intelligence presents significant opportunities in the areas of scalability and automation, as AI democratizes the predictability and objectivity of talent data.
People sometimes mistakenly believe that AI simply automates repetitive tasks, but sophisticated AI programs can tackle complex problems, just as any human could. When fed the right data, AI can actually make more objective, predictive decisions. And the Harvard Business Review says there are reasons to expect AI-talent tools to be more accurate and predictive than humans.
For example, they say automating all unstructured and humanly-rated interviews would reduce bias and nepotism, while increasing meritocracy and predictive accuracy. While in-person interviews are difficult to standardize, conducting them via video allows us to put people through the same exact experience, capturing millions of data points on their behaviors, while removing human prejudices, in the process.
Another advantage of AI is that it can be trained to ignore traits, like race and gender, focusing solely on the relevant signals of talent or potential. But in order for this to work, organizations must identify real performance data to train the algorithms.
So, can AI make hiring fairer?
By training algorithms to emulate, predict or anticipate human preferences, they not only reproduce biases, they make them worse. That’s not to say AI is biased in the way humans are, regarding emotions, feelings and opinions. But when AI is trained with biased data, not only does it emulate those human biases, it makes them far more efficient.
When AI is trained to identify actual drivers of performance, then we can expect a fairer, more accurate and replicable assessments of talent, beyond what any human could accomplish.
Whether you’re competing for talent in today's ever-competitive environment, trying to maintain a flexible workforce, or attempting to empower your people with tools that make their jobs easier, HR teams in all industries are embracing intelligent automation, ensuring their companies remain profitable and competitive.
Here are just a few ways that Level can drive value across your enterprise.