I, Robot [Recruiter]

Image by ergoneon from Pixabay

Share with:


Isaac Asimov's series, I, Robot, written in the 1940's fired the imaginations of a generation of post WWII kids. By 1950 Alan Turing had converted Asimov's science fiction stories about 'Robbie' into an engaging science challenge- taken up for the next few decades by pioneering computer software designers, inventors, dreamers, and academics who believed we could pass the point where computer conversations were indistinguishable were indistinguishable from human counterparts - see the Turing Test

More than 75 years later the debate has moved from whether we can approximate human interaction to what impact will AI have on...everything? The Science of Robotics has clearly advanced.

How far behind can recruiting be?

Has it already happened and, we just don't know it?

Rob MacIntosh's recent long-form post and call-to-arms, 70% 0f Recruiters Don't Care or are Clueless: Long Live AI Recruiting is a helpful summary and overview of where we are today- well worth the time invested.

And, by the way, while we are less likely to hear about 'robots' in the context of recruiting on any given day (ask me about Sophia in the picture above who was first showcased by Kevin Wheeler several years ago in Australia), the phrase, Machine Learning, is essentially the same thing and rapidly becoming Recruiting's biggest buzzword.  (Chris and I just returned from a user conference where every session used that phrase at least once.)

Here's an easy mind problem to consider. Is there a job open in your company today (only need one) where, for example, you would consider software designed to:

  • sense an opening the moment it received a formal 'approval', extract relevant data from the linked description, write a message using a pre-set job marketing template, and send the message to a distribution portal
  • then, when responses occur, respond to or extract from the responses sufficient data (or ask additional questions in a conversational way in series over time), rank the candidates against relevant weighting criteria (eliminating unconscious bias) and, make a conditional offer (or schedule an interview or ask  a human to check the software's 'judgement' - if the lawyers still had too much of a say).

Our point is to argue that the software already exists to disintermediate recruiters for (a good guess is 20%) of the jobs currently being done by them. By eliminating- even for 1 high-volume position, the choke point of time, cost and variability of recruiter performance (not to mention hiring managers- i.e. all humans), we think the data would show better quality hires, higher retention, significant savings and a measurably better candidate experience (fairness, baked in human touch, closure, opportunities to listen, clearer expectations, speed to completion).

Of course, no one is ready to do this for real- and so, no one has cobbled it all together...yet...or, have they?

As the newest machine learning tools emerge, we hope cXr Labs will interest them in getting feedback from TA leaders in the Colloquium who have volunteered to check out what is bubbling up. If you see something, say something. Let us know if you have the time and inclination to rotate in on our volunteer list.

Discuss this article & related topics on the CXR eXchange.