We are now seeing the emergence of the first generation of embodied hiring robots, who could in time replace human recruiters. It’s important to make the distinction between embodied and disembodied robots, as there are already plenty of hiring algorithms out there in the marketplace and corporate space that are hiring candidates in the thousands. For example, the Fortune 500 company Xerox has handed over it’s call center hiring decisions to automated hiring system that makes it assessment decisions based on smart, data analyzing algorithms. Human recruiters have a very limited influence on the call-center hiring process at Xerox now. This kind of disembodied robot recruiter often goes under the radar because the perpetrator, the algorithm is invisible, but when we start seeing more fully embodied robot recruiters the threat becomes apparent, putting us in alert mode.
And one thing that recruiters should be alert to is Matilda the Hiring Robot. She is able to read human emotions and while the potential applications of Matilda are endless, she is currently being deployed as a job interviewer for sales positions. She currently has an arsenal of 76 questions at her disposal to assess candidates and is programmed to execute a 25 minute interview, during which she can read emotions and play relaxing music. The Robot is about 30 cm tall.
One of the reported strengths or advantages of a robot recruiter over a human recruiter is that they can form decisions, free of whim or prejudice and without bias. As most of us in the industry know unconscious bias is a real problem in the hiring process and is proving even more problematic to eradicate, as many people don’t even know it’s occurring, be they victim or perpetrator. This study from Yale has shown how unconscious gender bias has crept into recruitment of scientists with the study showing that male scientists were scored higher and more likely to be hired.
Matilda has the potential to hire and assess without prejudice, since they are assessing based on very controlled criteria. In fact, they are assessing candidates based on massive big-data-based computational analyses that are far superior to what a human recruiter could achieve. Actually, robot recruiters have the power to render accepted but flawed professional wisdoms completely defunct. For example, the Xerox disembodied hiring robot/algorithm found that the best predictor of customer-service worker’s loyalty is that they live nearby and can get to work easily – factors previously ignored by human recruiters – and when they incorporated this statistical insight and other key findings to their selection criteria, attrition rates were cut by 20%.
Robot recruiters really have the power to cut through many of the myths perpetuated by the analytically limited human mind. For starters, many employers will shy away from candidates from criminal records yet research shows that for certain jobs, there is no correlation between criminality and work performance. Apparently, those with criminal records actually perform better in customer support positions. Job hoppers are also frowned up by employers – being seen us unreliable – yet research from over 100,000 call centre workers, shows that job hoppers are no less loyal than those who hadn’t previously jop-hopped
Matilda might not look or seem like much and may not have too many modern recruiters quaking in their seats. But remember, this is a prototype, first generation product, and within 2 or 3 generations, I’d expect hiring robots to be more than a match for modern human recruiters.