predictive

ARTIFICIAL INTELLIGENCE: ARE HR PROFESSIONALS AT RISK?

Latest article published in InsideHR

 ROB SCOTT — JULY 22, 2015
would HR professionals be as enthusiastic about HR technologies if they contained Artificial Intelligence (AI) capability

Are we ready to be pushed down the proverbial pecking order of importance by sophisticated AI technology? asks Rob Scott

Very few HR and talent professionals would refute the value that technology has brought to their operations. HR functions have leveraged these tools to become efficient, effective, collaborative, engaging and more accurate. But would HR professionals be as enthusiastic about HR technologies if they contained Artificial Intelligence (AI) capability that could predict more accurately and make better business decisions than the highly educated, people-focused HR practitioner?

At what point does software that is able to pick the best applicant, predict who is most likely to resign or identify the best mentor for a talented employee, become a legitimate replacement for a highly paid HR practitioner?

Most HR professionals I engage with don’t believe this will transpire, citing the complexities of human behaviour, personal choice and the absence of universal logic in managing people in the workplace. In the short term I agree with them, but not for the same reasons they mention. In fact, when I look at how most HR functions rely on standard processes to manage certain events, I have no doubt that near-future HR technology will do a better job than humans in executing these rule-based processes. Our flawed minds can never achieve the same level of efficiency.

“AI in HR is maturing; we are seeing interesting algorithm designs, predictive analytics and automation solutions coming to market”

This is not to say that our current HR technologies are anywhere close to being artificially intelligent. Right now there is a lot of hype-spinning by software vendors about the predictive prowess of their tools, but in reality these are immature tools. We should, however, be under no illusion that sophisticated AI for HR is heading our way. As it becomes more credible and capable, it will displace employees who are focused on maintaining standardised HR processes and mundane transactional work. There is, however, a far deeper and fundamental reason why I believe AI will, in the short term, find a home as a digital assistant rather than as a replacement for HR professionals. It goes to the heart of a human emotion – fear. Having artificially intelligent machines making sophisticated and important people-based decisions feels threatening and generates a level of anxiety about our status as human beings. We are not ready to lose our “superiority” to machines, no matter how intelligent they become.

As an example, Microsoft recently released a small tool which guessed one’s age based on a picture you uploaded. The results were mostly wrong, however, the tool went viral. Why? The reasons lie in the notion that while the technology is inaccurate, we feel less threatened by it and are able to maintain our dignity and humanness.

This is a powerful lesson and opportunity for HR software developers. Building AI software that is too accurate and human-like is likely to be rejected or underutilised, not because its outcomes are incorrect, but because it pushes human beings down the proverbial pecking order of importance and insinuates that the work they are doing is demeaning and unnecessary.

“Building AI software that is too accurate and human-like is likely to be rejected or underutilised”

Of course, we shouldn’t forget that technology enhancements have been at the heart of mankind’s industrial revolutions and progress. New machines with capabilities that outshine human ability have typically been met with resistance from those affected, at least until new work opportunities borne from the new technology become evident. AI in HR is maturing; we are seeing interesting algorithm designs, predictive analytics and automation solutions coming to market, but future job clarity in a digital and AI age is still blurry. Until then, AI tools for HR will develop into great digital assistants under control of HR professionals. At least for now the role of the HR professional remains in demand. 

5 key takeways for HR 

  • AI is a growing phenomenon in HR. We are increasingly seeing the inclusion of decision algorithms, predictive analytics and automation tools in HR software.
  • Basic AI tools will have the ability to manage standard HR processes with little to no human intervention, ultimately displacing employees from these mundane roles.
  • Complex AI tools which can make human-like decisions are likely to be rejected in HR because of the implied threat to our status.
  • Whilst it seems far-fetched, HR professionals should start thinking about how to “manage” and integrate artificially intelligent machines in the work environment.
  • Digital HR assistants are already with us managing workflows, finding information and managing large amounts of data. We don’t need to fear AI.

Image source: iStock

Advertisement

Your (HR) data will find you

I’ve never been a fan of telling other people what to do, think or say. I find it arrogant and demeaning at both a personal and professional level, and aside from situations warranting it (e.g. your immediate safety), you quickly lose respect and credibility. In many ways HR reporting has committed this same offence. For some reason HR leaders continue to produce standard HR reports and dish these out at regular intervals to management and executives for examination and supposed insight into their business operations. I don’t know about you, but I haven’t met too many leaders or managers who are excitedly waiting for the next HR report to land on their desk or email in-box.

Standard HR reporting has reached retirement age, and should be put out to pasture. It’s a reactive way of looking at your business and people management, and has its roots in a time when HR had to spend hours and days collecting, fixing and consolidating information into spreadsheets before distributing a report that added little value. Managers don’t need a report to tell them they have 3 vacancies unfilled, or that it took 47 days to fill a position or that 7 appraisals are still outstanding. They know all of that before they get the report. Sadly I still see many HR functions fixated on producing their “monthly report”.

Most modern HR systems have dashboard, trend analysis and mini analytics that replace the need for Standard HR reporting. This is much more effective and removes the constraints (and arrogant assumptions) of standard reporting by offering line managers a choice of information to support their decision making, and more importantly, they get it immediately. At a minimum you should be providing this approach to managers.

This approach doesn’t however go far enough. If line managers don’t know what questions to ask or what data or information is relevant, they won’t get the best outcome. This shouldn’t be seen as a loophole for HR to get back into a telling mode, but rather an opportunity to define how information finds the line manager based on their people related and enterprise social on-line activity.

Its good to see some of the leading HR vendors moving into this space together with strong predictive analytic tool-sets. The algorithms behind these tools are complex, but also configurable to suit your environment and solutions. It does however require a significant rethink about decision making in general, not just related to HR information, but including the interplay between other internal and external data sources.

Check out my previous blog “Is WFM  the new HR?“

%d bloggers like this: