If you didn’t hear enough about artificial intelligence in 2019, don’t expect that to change any time soon as AI technology continues making its mark in 2020. Chat bots, scheduling, resume screening, all are becoming more integral to the recruitment process as those technologies improve.
We’re not out of the woods just yet though. AI is still a long way from fully disrupting the recruitment process. And that’s simply because it’s not true artificial intelligence. For AI to be true artificial intelligence where it’s making decisions on its own, there is still A LOT of data needed. And that data is just not there yet.
Artificial Intelligence, in the sense that it’s being used today, is still very much a buzzword, and even more to the truth, a shinier, more compelling way for marketing cons to say “computer program,” which at this time is all any AI is, as far as it relates to recruiting tech.
But let’s play along and look at what to expect from recruiting AI in 2020.
What Does AI Recruiting Tech Look Like?
Right now there’s an amazing 40% of organizations who already believe in the promise of AI and utilize it in some form in their daily operations.
These automated tools are powering conversations with candidates in the form of advanced-programming chat bots, interview-scheduling programs, and candidate prequalification via resume screening.
Currently, recruiting platforms and candidate-engagement platforms are the most popular types of AI-driven solutions.
Glassdoor suggests AI technology will even take a seat in the manager’s chair by using big data in real-time coaching: monitoring employee tasks, providing feedback, and making recommendations. This will free up employers to manage more traditional roles and big-picture problems.
Of course the goal of AI, from a business perspective, is to lower hiring costs and increase productivity in the workplace. The backend boon being more time for recruiters to achieve more with their day and the tertiary result of an improved candidate experience. Tertiary because, let’s face it, few organizations pay attention to that part of the recruiting process or even consider there is anything wrong with their current system.
With the Good Comes Some Bad
What could go wrong, you ask? Well, AI can have its biases.
Take for example tech that uses a candidate’s computer or phone camera in an interview to analyze facial movement, speech pattern, and voice inflection to score and rank talent.
Suppose the interviewee has the ideal skills for a role yet was nervous on camera, or perhaps is a non-native speaker, or has a stutter?
These “aberrations” could lead to the AI program passing these individuals over for more “normal” metrics observed in other individuals, even though the individuals with aberrant patterns are otherwise qualified to fulfill the role.
Not that the AI does this intentionally, of course, but acting merely on the parameters by which it is programmed to judge a quality hire.
These small but important distinctions could possibly lead to the scrutiny of an organization’s brand over time if quality candidates are consistently flagged as bad hires.
This in turn creates a terrible candidate experience for applicants, which will eventually leave requisitions open longer and cause your organization to have increasingly costly times-to-hire.
Or worse, it could call into question an organization’s ethical behavior should a litigious party decide they were purposely interviewed unfavorably.
The point is there is simply not enough data currently for these AI programs to accurately assess quality talent without human intervention.
To put this lack of data into perspective, remember Google’s 16,000-computer network deep learning experiment? It was designed to browse YouTube and identify cats pictured in ten million video thumbnails from which it only achieved a 76% accuracy.
Now how much data do you think it takes to accurately identify a quality hire?
But AI Presses On
And it should. You’ve got to start somewhere.
And what AI technology exists so far has not deterred HR teams from early adoption.
In fact, AI is becoming a crucial aspect of many HR departments and talent acquisition initiatives with fewer and fewer organizations showing reluctance to adding an AI strategy to their recruitment process.
Research by IBM found that only 23% of HR professionals today have worries about AI and the potential biases it might create, as with the example above.
Something to think about with bias though is, AI doesn’t have any bias in the screening and selection process. It interprets the input of its designer. The machine learning algorithms are limited by the datasets and the programming choices of the developers.
In other words, AI is built and maintained by far-from-flawless humans who have inherent biases in how they believe the AI should respond.
The only way to remove bias is via a constant stream of incoming data and persistent testing to model outcomes. Cat pictures!
So, the numbers of those averse to AI are dwindling. A study from CEIPAL, a SaaS group specializing in staffing, found that an amazing two-thirds of all staffing firms in the polled in their benchmark report will adopt an AI-driven ATS by the end of 2020.
From all of this, it appears that AI is here for the long haul. Pandora’s Box is open. Just how effective AI will be in the recruiting process depends on the strength of the datasets.
Look, AI is still new. And it’s going to have that new car smell for a long time.
In the short term, AI technology will eliminate the mundane tasks that prevent recruiters from making a larger impact and help create an optimized recruitment process. Over time this will enable better hiring decisions, create a better candidate experience for job seekers, and close gaps on time-to-hire.
In the long term there is no question that artificial intelligence will be game changing in how we attract, hire, and retain talent. There will be bumps, as with anything new, but in time I believe it can disrupt the industry for the better, bringing in flexibility, productivity, and intelligent insight.
The real trick, and subject for another post, will be to find a balance with keeping the human touch alive for candidates where interaction and communication with artificial intelligence may create a negative candidate experience.