Artificial Intelligence and Human Resources are as opposite as the north pole is from the south pole, as different as matter is to antimatter, right? Artificial Intelligence (AI) is all about impersonal, robotic, computerized number crunching and is, quite frankly… not human. Human Resources (HR) on the other hand… is. So, how can the two be used in one sentence?
That’s a good question, and one that is garnering a lot more talk by humans these days. Many people fear AI, not just because they’ve watched too many movies where robots are able to outsmart humans and take over the world, but because they really fear for their jobs. A case in point that has been going on for longer than many people realize is computer automation replacing humans on assembly lines. AI operated kiosks are now being rolled out in the “fast food” industry, even as I write this.
The reality is, more and more jobs that were once done by humans are now being done by machines that are able to do them more accurately, faster, and without any need to stop production for things like breaks and sleep. But while organized labor and their friends have been warning us for more than fifty years about how jobs are going to be lost to machines, the unemployment rate in developed, industrialized countries is actually very low. In fact, over the last 50 years, the rate has stayed virtually level, even though the population has grown considerably.
How can this be?
The Economy has Changed
The fact is, the economy of the developed industrialized world has changed. There are fewer people doing repetitive manual labor than there were 50 years ago. In fact, there are a lot fewer people doing those jobs. However, it’s really hard to justify having a human spend 8 or more hours per day doing a completely mindless and repetitive job that could eventually drive him to alcohol, drugs or outright psychosis, when a machine can do it.
The development of machines to do highly repetitive or dangerous jobs instead of humans has not created the nightmare of high unemployment that was predicted. Instead, humans today are living longer, healthier lives than ever before, while continuing to enjoy employment levels that haven’t been significantly effected negatively or positively by automation.
What has happened has been a shift. Humans who once worked doing tasks that are now done by machines have moved on to do other tasks that machines have not yet been designed to do. More people than ever before are not only finishing high school, but they are going on to trade schools or universities, and from there to productive work, perhaps not even imagined in their parents’ day.
In a nutshell, the creation of better, faster and yes, smarter machines has not created unemployment. It has opened up whole new industries and development as people – human resources – have been freed up by very capable machines to use our minds and motor skills to do bigger and better things.
AI isn’t going to cause massive unemployment, any more than modern machinery and automation have. In fact, AI is already being used in HR departments to help place people in jobs, rather than making them redundant.
When AI is added into the HR department, rather than making jobs disappear, it can actually add positions previously not thought of while finding people to fill those positions. The advantage are twofold. Employees whose present jobs may no longer be economically viable are quickly moved into new positions where their skills are most valuable. The company becomes more profitable and productive than before.
Does AI really benefit HR?
The short answer is a resounding YES!
In the past, a small employer would post a job, then look through all the resumes that came in to make a decision on who should be interviewed.
Obviously, somewhat larger companies needed to post more jobs, meaning they would have needed to wade through more resumes and also conduct more interviews. This manes they would need to have more HR staff to handle the larger load.
For a large company, this became a major operation. Large companies continually post so many openings and receive so many resumes to sort through that HR staffing, alone, accounts for a major expense. Not only must the HR team be much larger, but the overall office space and equipment for HR must be larger. Obviously, it was a major expense that potentially could have been better invested, if it was possible to reduce it at that point in the chain.
Modern applicant tracking software (ATS) makes it less necessary for a talent acquisition specialist to go through every resume. An ATS system now allows a single HR specialist to search through literally vast numbers of resumes by keyword, education, location, experience level, etc.
Notice that I wrote “less necessary.” ATS software isn’t perfect and may not return the best and most relevant search results. But, what is THE most relevant result? Is it just a question of hard skills and soft skills? I don’t thing…
This is where AI shines. Artificial intelligence available, today, can sift through quite literally thousands of applications quickly, very efficiently, and at a fraction of the cost of human staffing. Because AI works by using machine learning technology, employers can enter pertinent system teaching sets and information such as who the best current employees are for a position when it opens up, as well as resumes of the best present industry performers along with their history and career paths.
When AI is used to search through all the resumes, it is now possible to provide the recruiter with candidates that mean predetermined criteria, such as their location and previous experience. Furthermore, they can be ranked for the recruiter based on other important characteristics or backgrounds matching the highest performers among existing employees.
AI may compute many sources of information, not only the CV or LinkedIn profile of candidate but also of employees. AI may parse all information provided by social networks. Analyze photos posted by a candidate/employee on Instagram, analyze the tweets, analyze all Facebook and LinkedIn activities, analyze web sites/blogs, etc. That’s not all, AI is able to read and analyze facial features. It can determine the predominant mood, emotions and behaviors of a person.
Without any human interactions, AI will match the candidates (or everyone – not only applicants) with the company environment (culture, business, growth, employees,etc.) to identify the best revelant people that benefit the company.
Eliminate human interactions?
On one hand, this process doesn’t eliminate a meeting between a real human being and a candidate for an interview, but it greatly enhances the potential outcome for the recruiting of better candidates in less time at a lowered cost.
But on the other hand, the machine may avoid cognitive biases (a derivation of rational thought). A recruiter who had an car accident this morning, probably will miss the best candidate with whom he has the interview this afternoon.
We can list some human cognitive biases, let’s have a look at Julie ATLAN’s article:
1 – Recency effect: it is easier to remember the latest information we were confronted with.
2 – Effect of simple exposure: the fact of having already been exposed to a situation or of having met a person beforehand makes their feeling towards them more positive.
3 – Effect of primacy: it is also called the bias of the first impression or mental anchoring bias because the information received first will determine the general impression that one has of others. You will then tend to conceal information that does not support your impression.
4 – The bias of projection or false consensus: we tend to think that most people think and reason like us, even see that they agree with us. In terms of recruitment, it is assimilated to membership or endogroup bias which corresponds to the tendency that individuals involved in the recruitment process to favor the candidates in which they find themselves – whether it be a hobby, a personality, common values. The trap is that we let ourselves be seduced by a candidate for the wrong reasons!
5 – The bias of stereotype or association or excessive generalization: we consider certain information as sufficient to make a decision because we make associations between character traits or specific information that make us draw hasty conclusions such as for example belonging to a group of individuals – automatically giving them common characteristics.
6 – The halo effect: close to stereotyping, it is often assimilated to the adage “he sees only what he wants to see”. This bias consists in generalizing the set of characteristics of a person from a single one. The first impression on a candidate will have a chain effect on the rest of the interactions.
7 – The framing bias: you have a very precise idea of what you expect from an exchange and you need to reinforce your prejudices. The tone and the turn of your questions will be the perfect proof.
8 – The contrast effect: we have a natural tendency to always want to compare everything because comparison is often the basis of our judgment. It can also manifest itself when information is presented from a certain angle.
Some companies are also using AI chatbots to “chat” with potential candidates, before they even send in their resume. They are able to answer questions about such thing as privacy concerns, social media profiles, flexible hours, and many of the other questions that would not be part of the job description in an employment posting. This solves a problem that can be epidemic for some companies, where they experience a continuous need to replace employees who discover shortly after they start work that they really don’t like the constraints of the new job.
With intelligent chatbots, job seekers and companies are often able to discover ahead of time if a potential applicant is suited for the job in a more pleasant, relaxed and less stressful manner, and without hiring more HR staff. Furthermore, if a position is already filled, a chatbot can politely and pleasantly let applicants know, rather than the applicant receiving no response, as has often been the custom in the past. This helps provide a good experience, even to someone who does not get hired, leaving the door open for a skilled candidate and potential valuable employee to apply for another posting in the future.
Eliminate Discriminatory Hiring
AI can go a long way in eliminating discriminatory hiring. While it would be wrong to suggest that most humans involved in talent acquisition consciously practice discrimination based on certain stereotypes, every human being has prejudices built in that they probably don’t even realize they have. It is human nature. It is only necessary to read what success coaches such as Jack Nasher write to know that things like physical appearance matter to human recruiters!
AI, being unemotional machine intelligence, can help to eliminate this unconscious discrimination. Instead, AI uses objective information to rank candidates based in the needs of the employer and the candidate’s potential to fill those needs.
When AI is combined with video during the interview process, the candidate’s body language, which includes posture, eye movement, facial expression and potential more that AI will eventually pick up, certain things about the candidate can be quickly analyzed. It can quickly be determined whether the candidate is being honest, how comfortable he/she is with the process and even how detailed and accurate the answers are. Such technology has the potential, not necessarily to eliminate the need for real human interviews of qualified candidates, but almost certainly to greatly streamline the process of narrowing the field.
Free Up Talent
As mentioned above, there is still a certain amount of fear about the machine replacing the human. People do worry about how AI might negatively effect their job.
However, there is an upside that people employed in firms using AI should really be thinking about. After all, doing the same thing day in and day out, no matter how interesting it may have been in the beginning, is, quite frankly… boring.
AI provides real potential to change that. Once you get past the understandable discomfort of realizing you’re being watched continuously by a machine, there is a real upside to this. First of all, that machine is impersonal, so the fact that it’s “watching” you shouldn’t bother you nearly as much as those dreaded meetings with the manager for a yearly performance appraisal. In fact, those could be virtually eliminated.
The big upside is that AI can potentially free up talent that is being underused by over-qualified personnel. Everyone who has any talent, after doing a certain job for any length of time, learns a surprising amount of new skills that could be better and more profitably used elsewhere in the company.
Such talents are often easily overlooked in a traditional setting. As long as everything is going along smoothly and the job is being done, nobody notices or says anything. In fact, in a traditional workplace, it is very often that the only time someone is noticed is when something goes wrong!
When AI is integrated with HR and the system as a whole, including all departments, customer service and acquisition, sales departments, etc., it is a continuing possibility that new areas of business will be found to be developed, requiring human resources that may already exist in the company, being underused or even wasted on positions that are either redundant or that could be effectively filled by junior staff at a lower cost.
Furthermore, by moving existing talent to areas that are more stimulating, their loyalty and productivity will likely increase, making their contribution even more valuable.
Summation of Benefits of AI to HR
Leave the boring, talent-killing busy work to AI. A bot never complains (so far!). One of the best ways to destroy promising new recruits is to bore them to death with repetitive tasks that software can do better, faster, and… endlessly.
Streamlining of the screening process is easier, faster and probably much more accurate with AI. A company naturally wants to hire the best talent it can, but that usually requires posting to large job boards where there can often be thousands of applicants. AI can wade through them at the speed of light and without getting bored or careless.
Human bias in recruiting is obviously reduced, at least throughout much of the screening process. (It is noteworthy that a LinkedIn study reported in July, 2017, shows that 56% of candidates still prefer to hear from a real human hiring manager.)
AI can really help in onboarding and training. Learning modules for new and continuing training are now not only a good idea, but they exist and are effective. Employees can now learn at their own level and timetable with efficient tracking of their progress. Indeed, AI can also answer many of the common questions asked by employees, greatly reducing the amount of time spent on such mundane communication by the HR department, while freeing up managers to deal with more complex issues in an unhurried manner, thus increasing the overall quality of coaching in a given firm.
In any discussion of training and expected outcomes, the dreaded topic of performance analysis arises. AI is now making it possible for this to be an ongoing process. Rather than having semi-annual or yearly performance assessments, it is now possible for defined objectives to be set in place broken down into easily tracked increments. This helps improve productivity and also detects negative performance indicators before they become a bigger issue.
Big data and metrics is no longer something to be feared. AI makes it easier and faster to crunch the numbers and analyze the data. In turn, this helps manager match the right opportunities to existing and new employees.
Furthermore, problems may be identified earlier, making it easier to recommend possible solutions, including further learning opportunities, thereby reducing turnover and likely having a happier, healthier workforce.
Predictive recruitment is THE future.