For a few years now I have read, listened and wondered about what impact Artificial Intelligence will have in the recruitment industry. Artificial Intelligence is slowly but surely changing the world and the ways in which we live and work. Most of its effects will be transformational and of great benefit to the human race, while others will have consequences that the majority of us probably don’t want to think about too deeply. In almost every sector, as AI encroaches on the working day the human practitioners make the case for their continued involvement, while conceding that other people may lose their jobs. Some of this will be wishful thinking. In the course of preparing this article I read a blog which said, “The interactions that people have with one another can never be replicated by a machine.” That’s what copywriters thought until JP Morgan replaced them with an algorithm.
AI is going to cost people their jobs, albeit it will create others. The deficit/surplus of jobs lost/gained and the ways in which working practices change and society adjusts will colour our attitudes, but it’s what we do about it – how we adapt to it – that ultimately matters.
This is the first of a series of blogs on this subject and its impact not just on recruitment generally but specifically in the key markets in which Cedar and its customers work. We’re going to start by looking at AI and recruitment, before moving on to Finance, Procurement and Transformation & Change. But for all our clients and candidates, the key questions will revolve around how recruitment companies like Cedar use AI to make themselves more efficient and productive.
It is perhaps helpful to begin by defining AI. At the most basic, general level, AI is the software that allows computers/machines to carry out tasks in a way that we would consider “intelligent.” AI can also be divided into two areas: applied or general. Applied AI covers the kind of things we’re fairly used to reading about: it’s the basis of the systems designed specifically to drive and steer autonomous vehicles and intelligently trade stocks and shares. In contrast, general AI refers to systems/devices which can, in theory, be applied to any task. While less commonly reported than applied AI, this is the area where the really cutting-edge stuff takes place, especially in the development of Machine Learning (ML). ML is, in many ways, a subset of AI, but the difference is that ML involves the machines thinking for themselves and it’s this that gives some people the heebie-jeebies.
Crucially, AI is not the simple application of machinery and technology to an industry. Since the invention of the wheel, humankind has applied technology to make life easier. In recruitment, we’ve seen technology increasingly applied in myriad ways over the last 15-20 years. The introduction of the job-board, as opposed to the old-fashioned newspaper (itself the product of perhaps the second greatest advance after the wheel, namely printing), brought thousands of easily searchable jobs cheaply into one place where candidates could browse them. The last 15 or so years has seen the development and refinement of the ATS (Applicant Tracking System), video interviewing and a raft of online applications, all claiming to make the recruiter’s life easier. None of these made much use of AI: they simply used more sophisticated technology than before. The difference is that until relatively recently we simply thought in terms of building machines/computers/robots that we taught/designed to do tasks to make our lives easier. AI has changed all this by getting the machines/computers/robots to “think” how to “learn” and perform these tasks themselves.
Today, go along to any recruitment event and you’ll be endlessly assailed by rec-tech companies whose kit purports to solve one or more recruitment problems. However, in my opinion, the essential business of recruitment – finding and matching quality candidates with the jobs available – is often forgotten in the rush to find the killer recruitment app that will make £millions for its inventors.
Recruiters have two semi-distinct pools in which to fish for candidates. Those who want a job – the active candidates – are the most obvious ones. It’s not too hard to find them and in fact they often will seek you out first.
Much harder to trawl is the second pool, containing those who are not actually looking to move on but might be persuaded if the right job is put in front of them – the so-called “passive” candidates. Sometimes passive candidates move to the active pool of their own volition (a bad day at the office, a chance conversation about jobs at a competitor). More often though, given the current market’s skills shortages, finding and then moving people from the passive pool into the active one is more difficult than it has ever been; that’s where Cedar comes in…
Currently, recruiters start most candidate searches on LinkedIn. Experience and Boolean strings help separate the average from the exceptional. Other social media are also used to source candidates and there are several rec-tech firms that will sell you software that claims to nurture and persuade candidates that your firm is the place to be. Yet all such software and methods of applying it still rely mainly on LinkedIn and other online sources to create a database of candidates. To date, no-one has really come up with software that will allow resourcers to find everyone they want. The ability to key a job and person spec into a computer and then wait for it to come back to you quickly with a list of ideal candidates is the holy grail and this is where AI has the potential to make a major breakthrough for the recruitment industry. ML can, and will, learn where to find and how best to connect with both passive and active candidates. The techies are working on this, naturally, because there is a stack of cash to be made. And just like the early days of electronic share trading, the companies and their customers who are first to market, will be the ones who make their fortunes. Weld on some of the undoubtedly excellent screening, interviewing and onboarding software (and in particular the use of chatbots, with younger people more comfortable with their use) that currently exists and you have an end-to-end model that will change the world of recruitment massively. For a start, the need to maintain a database of candidates will no longer exist if you know you can create bespoke databases of quality talent for each vacancy.
The obvious question is, “where does this leave the human element of recruitment?” In the medium term, by which I mean the next decade at least, although resourcers’ jobs may be at risk, I think the skilled and experienced human recruiter will be even more important. Yes, algorithms will assess and verify candidates, but until we stop employing humans entirely then companies will still want to know that a real person has looked the shortlist in the eye and satisfied him or herself that they’ll cut the mustard.