Recruitment will start boiling but this is how you do a great meal

Some believe that within 2 or 4 years, what it’s called “Machine Learning” (ML), will swallow the recruitment as we know it, given the fact that this such radical change might be inflicted by a new global economic crisis which is forecast to happen as soon as 2018. This, is implying Oleg Vishnepolsky in his article “Recruitment industry will die in 2018”, will have as a consequence, the evaporation of most jobs in recruitment industry. Seems very plausible, but with a few reservations that I want to point out to this alarming scenario. I put aside the speculation of a new economic crisis, which never comes when expected and triggers massive loss of jobs and slow recovery in pretty much any industry, not just recruitment, with or without the existence of Machine Learning (ML):

1 — most likely, Machine Learning (ML) will change recruitment the same way digital cameras did for the photo industry.

Along with the presence of the first digital cameras, hundreds of thousands of photo laboratories worldwide started to go bankrupt. All of a sudden, photography was very affordable and images of any size could be stored on computers, phones, DVDs, flash-drives etc. This didn’t mean the disappearance of professionals but rather made this field accessible to any person willing to become a pro or not. GIFs as motion pictures and HDR photography (High Dynamic Range) were invented, light field technology from Lytro also emerged allowing us to refocus a snapshot after it was taken and, in general, amazing photo editing tools have dramatically upgraded the photo industry. All this didn’t drive away the photography from art but on the contrary, there are much more professionals now than 15 years ago and with superior skill sets.

2 — Machine Learning (ML) is, at best, the automatisation of immediate deduction which represents detection of a social pattern and short sighted decision.

New buzzwords are VR (Virtual Reality), AI (Artificial Intelligence), ML (Machine Learning), AR (Augmented Reality). Not a single startup can afford not to put into its elevator pitch words like innovation or ML, no matter if they’re true or not. Everyone pretends doing it, but how good, efficient, documented, no one knows for sure. Indexing and filtering resumes using keywords so that recruiters can make a better use of their time it is not ML, but still automatisation. Mathematical diagnostic of a candidate as a reading result from a cv is limited to one’s ability of writing and bragging so that can draw attention. Moreover, such limited diagnostic cannot contain the overall understanding of character and flow events of a person’s life. This and many other limitations are the reason why the recruiting process is still very elaborate and crucially relying on human interactions.

If, let’s say, the algorithms will be “trained” to detect workplace aversion on social networks, soon enough people will adapt so that they don’t raise any red flags and put themselves at risk by being fired. The censored behaviours employees have now on slack when sending messages on company’s channels will be extended to any social network and digital tools will struggle to catch up, trying to interpret micro-signals of intention for leaving the current job. The algorithms will fetch, recruiters will persuade. (see also “To all recruiters — use machine learning to hire better candidates”) Just like in the past, when the work of a photographer also was about manipulating chemicals in the dark room and now their skills shifted entirely on the creative side, a similar process will also take place in recruitment.

ML is a continuous adaptive process, a hunter-venison like scenario. The algorithms will try to identify humans intentions as accurate as possible, which in turn will act on an evasion manner when they feel like there is something important at stake. No one is willing to let himself twist read by a big-brother eye from HR department, to give up control to some lines of code and, as a result, the online behaviours will be more contained and educated. People don’t want to be fired, but they want to have the liberty of quitting pretty much like they don’t want to be rejected from a relationship, at least not before they find a new partner.. ML is not Artificial Intelligence (AI) and will keep depending on perpetual input of programmers and product owners. The complexity of digital learning will be directly proportional with hypotheses and biased by the cultural background of its creators. Machine Learning (ML) means segmentations, interpretations, probabilities, permutations and conditionings, whilst Artificial Intelligence (AI) is supplementary encompassing conceptualisation, strategy, philosophy, determination — a rather complete autonomy from human mind.

3 — recruitment has many things in common with manipulation, persuasion and identifying potential, but Machine Learning (ML) is far from all these.

Until a universal reputation system is developed (via blockchain), on which everyone should painfully depend on, recruiters will strongly be needed and every person can put whatever wants on the resume. Unfortunately, many endeavours of such kind have slipped from reputation to popularity. (For the past years I’ve tried to build http://2ndwavers.com — a reputational system by adding an essential layer to ecommerce. But so far I haven’t succeeded making the right people dream next to me.) If it’s wise to hire for will and less for skill, if the most valuable life lessons are born out of failure, struggle and pain and if personal evolution is strongly correlated with the people you choose to have by your side, what ML will essentially be doing is to increase the thermal agitation coefficient on the job market. (e.g. after posting a positive pregnancy test on social networks, an algorithm will send out new job offers on the premise that a recently become dad is more prone to take on a new venture by accepting longer working hours or more free time. The thinking there is that an essential change in someone’s life triggers another one.)

In order to stay relevant, perhaps somewhere in the near future recruiters will become a mix of trainers, mentors and personal agents. They will need to nurture and believe in their “products” the very same way entrepreneurs put effort and passion in whatever they’re selling. Recruiters will need to have a much closer relationship with the workforce still uncaught by platforms like UpWork. The more the job market pass into a volatile state, the more solidified their relationship with the chosen workforce group must be in order to survive. Finally, the better they know their “products”, the better and frequent will sell them, reaching a point where earnings are made from both companies and employees. Some recruiters already have intelligent novel strategies, approaching you with the desire of teaching them about your work field, knowing your family, everyday vocabulary, personal history, expectations and values.

Conclusion

The sandbox mantra of “fail fast, fail often” of millions of startups across the globe is asking for workforce that cannot be invented overnight in huge numbers. Instead, is being displaced from other companies. Machine Learning (ML) brings targeting in the job market chaos. This chaos is determined, among other things, by the accelerated derailment of loyalty and engagement from employees to companies and vice-versa.

Furthermore, freelancing platforms (UpWork, TaskRabbit, Airbnb, lynda.com, Dreamstime, Iconfinder etc.) keep weakening the influence area of recruiters and already determined radical transformation of jobs, long before the emergence of any algorithm related to ML.

The new recruitment paradigm might bring more bonding and less deceiving or superfluous activity. Besides, the holy grail aimed by any recruiter is exclusivity — which, of course, in this case comes with bonding.