Will A.I. determine your destiny? ?

At the beginning of the 2020 lockdown, momentarily, many companies faced feelings of paralysis

All future plans were moth-balled, and crisis management was the order of the day. For a while it looked like no one was going to hire. However, the food and beverage industry saw demand escalate and a different set of logistical concerns emerged. Within weeks I was part of the shift from conventional recruitment techniques to a new model that I suggest may well continue whatever happens to the infection rate. Photo: Charles Deluvio

Companies began making significant hires without having physically met in person

Actually, a couple of years ago I wrote about AI in regard to recruitment. I wasn’t particularly quick off the mark as AI within recruitment has been around for a decade or so. However, it has really demonstrated its worth during the pandemic when face to face interviews were impossible. Companies were making significant hires without actually having physically met in person. This was extraordinary and unthinkable even a year ago but is becoming normalised.

We are all looking for smarter ways to achieve the mundane

It is no surprise then, that demand for technological solutions for recruitment would be additionally attractive right now. We are all looking for smarter ways to achieve mundane aspects of our working lives. Recruitment software can take the grind out of sifting through CVs and also basic tests that evaluate personality and intelligence.

If AI is used to sift CV’s it means a candidate is being judged by their ability to write a CV

The type of qualities it searches for are also extremely important in this new world order. It is worth testing someone’s risk tolerance or their reactions in different situations for example. We are definitely looking for agile, expansive thinkers that can look at both the macro and micro level. However, if AI is being used to sift CV’s it means that the candidate is being judged by their ability to write a CV and not many people are versed at doing this professionally, hence the need to seek professional input when doing this.

Photo: Markus Winkler

One US company that has made strong inroads in AI for recruitment is Pymetrics

Their recruitment software aims to ‘fairly and accurately measure cognitive and emotional attributes in only 25 minutes’. It’s quite a claim, especially when we know algorithm bias does exist. Some time back Google did experiment with robot interviews but abandoned the experiment. Why? The robot felt males outperformed females. Interesting.

However, there have been developments and now, to screen out candidates on the first sift, AI software is often used. Certainly, some of the biggest employers employ this technique as it allows for less time to be used when looking at a wider pool of applicants.

Yet, I can see the attraction. Many candidates, that may prove their worth later on in the process, never even make the cut when there are huge numbers of applications. Their dismissal can sometimes be pretty arbitrary, so maybe AI might be fairer in those early stages. The pressure is on to find the exact fit for companies, as every hire really does count especially in these uncertain times.

But, also the pertinent question for me is: are organisations relying on job advertising to generate the CV’s in the first instance?

If this is the case, they will be selecting business critical staff based on who has actually applied for a role and not actually looking at the wider market of passive candidates, this is where we come in to our own. Question is do you want to hire the best candidates that are looking or the best in the market. Photo: Campaign Creators

Will AI ever be able to go and find candidates that have not applied for a job?

No matter how good AI is it will only be able to sift CV’s and will not actually be able to go and find candidates. It will be interesting to see how this plays out in the future. Tech is a wonderful addition to our lives but sometimes old fashioned ‘wetware’ does a great job that is hard to replicate.

An algorithm can reject point blank

In addition, there are other AI solutions to the interview style. Since 2016, HireVue, whose name reflects their product, records people dealing with questions at an interview. This is done remotely and then the audio stream is converted into text. This is then analysed by algorithm. It can scrutinise answers for certain keywords a company might thing are appropriate. At this point the candidate can be rejected point blank or the company can decide on the next outcome. This is not unusual, and the company has already conducted 19 million+ interviews in this way. Looking at its exponential growth it is no surprise to learn that in the next eight to nine years 16% or recruitment jobs will disappear. Looks like it might be time to retrain as a coder.

Can we ever truly rid algorithms from bias?

It’s fair to say that AI may well add to the mix and even offer more objectivity if the algorithm is free from bias. Pymetrics is keen to point out that is tests rigorously for bias and maybe it may well augment the whole process. After all, a cv is just one facet of a candidate’s skills and can only tell us so much outside of interview. Yet there are many examples of bias. Amazon abandoned an AI recruitment model when it came to its own conclusions that male candidates were preferable. This was down to the prevalence of tech industry experience rather than anything else. I guess these gremlins will be amended in the long term, but we have yet to design a robot that reads body language, semantics, proxemics and intuition. However, we can only imagine that it’s already on the conveyor belt.

Meanwhile if you cannot wait for a robot to find your perfect executive recruitment match then do let us use our considerable skills and experience to find the very best candidates for you. We think that, human flaws aside, we have a great track record of finding the people you need. Why not call us today for a no obligation discussion and see how we might help secure your next hire.