AI in recruiting really means ‘abdicated intelligence’

By Neil Patrick

What is advocated and marketed as technology-enabled recruitment processes increases the difficulty in finding and retaining great people. The best people are made not found. We don't need to get better at ranking people by increasingly tightly defined data points, we need to take ownership of our responsibility (and self-interest) to find good people and make them great. 

Artificial intelligence is the big topic in almost every professional field right now. From drones in farming to robot surgeons, the overriding narrative is about the tasks which AI will enable us to perform better, faster and cheaper.

HR and recruitment are no different. The application of AI is spreading like wildfire as new tools are developed which expand the range of tasks that AI performs to assist recruitment management processes.

Until now, it was relatively simple to integrate the digital world with our own professional world. Have a LinkedIn profile which is properly constructed. Expand our professional networking onto one or two social media platforms. Write some commentaries or blog posts. By these means, anyone checking us out could easily discover our credentials.

Some people understandably chose not to participate or did so in the most cursory way possible. They had no wish to participate in the online race. They had fears about privacy. They didn’t understand how social media worked. They had better things to do with their time. All these were legitimate grounds to not participate. But not anymore because…

This is now all set to change

The next wave of AI and big data is going to transform the processes of hiring way beyond anything we’ve seen to date. Traditional recruitment is a gruelling, complex process for employers and recruiters alike. Recruitment teams have to be heavily incentivised to commit to the heavy workloads involved. And this costs money. A lot of money. But AI will streamline and speed up these processes. It will be able to identify suitable candidates in a few seconds. It will message the chosen few and chatbots will perform initial screenings. Candidate selection decisions will be made on the basis of data and scoring algorithms rather than fallible human interactions. Very little human intervention will be needed.

Recruitment costs will fall even more. Hiring efficiency and speed will increase. Hiring choices will be validated and justified by the ‘scientific’ methods involved.

At least that is the vision. The reality is more worrying.

Why it’s flawed

Data is not science. The principal predictors of job performance cannot be discovered by algorithms. The first attempts to automate the selection process created a bigger mess than before. Online job boards and applicant tracking systems (ATS) drove application numbers sky-high and candidate quality tumbled. But acquiring 500 applications cost around $50. Using a professional head hunter costs about $30-$40,000 per hire for professional vacancies. These economics ensured that automated recruitment processes took hold and continue to grow in usage.

Nick Corcodilos explains in this video why the application of data driven metrics to recruitment ensures that employers miss many of the best candidates for any given role:





I agree with everything Nick says here. This process is flawed. It cannot find the best people because the available data points cannot determine that they will perform well on the job. And because quite a few of the very best people choose not to present themselves online in their professional capacity. Yet for all its weaknesses, automated recruitment is only going to expand because the cost differential is so compelling.

In a strong and growing economy, organisations can invest in quality processes. In an economy which is uncertain and faltering, when profits and growth are elusive, focus inevitably shifts to cost reductions. Cost trumps quality in such times.

Your career is at risk if you choose not to participate

It seems logical to me that the current and anticipated applications of technology and artificial intelligence in recruiting will continue to erode the quality of hiring decisions made. This may deliver short term cost gains, but will push up long term costs as turnover rises and employee performance falls.

Yet this is not my greatest worry. My fear is that the relentless advance of this technology will create a new underclass of smart, educated and capable people who have chosen for legitimate reasons not to present themselves online. These people will become completely invisible to the data capture bots. And that invisibility will slowly but surely eat away at their employment opportunities.

There’s not a ‘talent shortage’, there’s a leadership vacuum

 
For organisations, the deployment of recruitment AI encourages organisations to abdicate their responsibility to create and nurture their own human talent pool. This creates a downward spiral of ever increasing data point discrimination reinforcing the mythology of what employers disingenuously call their ‘talent shortage’.

And if the belief in the pseudo-scientific reliability of these systems persists within management, we will see the abdication of leadership’s responsibility for taking good people and helping them become great. Instead, the tools will be adjusted to cure perceived shortcomings, when the real shortcomings are rooted in the mistaken faith in progress through technology.

Artificial intelligence is well named. Because it’s not real intelligence…

PS. My good friend Marcia LaReau at Forward Motion Careers has a great post here about what jobseekers can do to avoid becoming a victim of this situation.