Everyone is convinced that AI is going to revolutionize their industry, and recruitment is no different. The question is, will it be a positive force that finally helps us to rid the process of age-old prejudices, or will the machines pick up our bad habits and embed them forever in their algorithms?
The answer, of course, is down to us. Will we use AI to free up human time, so that we can use our expertise, intelligence and empathy to understand individual candidates, their skills and motivations, and surprise employers with people who are better for the role than they ever believed possible? Will we allow virtual assistants to help us look at a wider variety of CVs than was previously possible, bringing less conventional candidates into the mix? Or will we simply transfer our prejudices into an algorithm and cement a machine-like reliance on conventional qualifications, experiences and ‘best fit’ workers?
Too many in the HR industry seem determined to do the latter. Forbes recently asked ten senior recruiters to say what they think the impact of AI will be, and although the answers differed, there was a worrying emphasis on letting AI match candidates and jobs through its superior capacity to process data.
While allowing a computer program to search through CVs does have the advantage of being able to consider more candidates, because of the processing speed available, we must be careful that this approach doesn’t go too far. Otherwise, there is a real danger that the new era of automation, where contact is initiated by algorithms and initial calls are made by bots, will see us losing the valuable touch of traditional human interaction.
The recruitment process is already more like the commodity business than we care to accept. In fact, we treat candidates worse that goods: products on Amazon get better marketing and visibility than people in a hiring situation. In order to compensate for this marketing deficiency, good recruiters are adept at looking beyond the resume, and reading between the lines of a cover letter, to see the potential of the person behind the paper.
Algorithms will not do this. Machines will inevitably select candidates based on historical expectations for a job, seeking to match traditionally desirable attributes and qualifications with what can be found in a resume or job application. If you are not a good writer or self-promoter, good luck.
Those who have come to their present position through an unusual career path; learned through sometimes difficult experiences; or take an unconventional approach to their work, will find this algorithmic recruitment process equally unforgiving. And yet, as we saw recently, these people are often the best candidates when it comes to doing a difficult job. Read blog.
Even those who are trying to use AI in a positive and responsible way to look at a wider field of candidates and select on skills and personality rather than just experience and qualifications face a very difficult task and could end up doing more harm than good, because the vast amount of data that’s used for data science is itself riddled with historic misconceptions, discrimination, bias and prejudice.
The historic data available for competency-based selection is simply not good enough. Instead, we need to get human interaction back into the hiring field, so that a recruiter can get to know the person, rather than just their potted history presented on a resume. This is not to say that we shouldn’t use a mix of both human intelligence and artificial intelligence: we just need to be clear about who does the most important things.
Last year, Undercover Recruiter rated AI as one of the five trends shaping global recruiting. But, while it said that 80% of executives believe AI can improve productivity and performance and such algorithms are expected to replace 16% of jobs within the recruitment sector over the next decade, it stopped short of suggesting that computers will take on the key task of candidate selection.
Instead, Undercover Recruiter recommends that AI be used to help automate more routine tasks, freeing up human time to concentrate on the thinking part of recruitment. It suggests using chatbots to answer routine inquiries and provide candidates with regular updates on the process. AI can also be a great personal assistant for recruitment executives, helping organise their time and completing routine paperwork on their behalf; and can play a role in the on-boarding process of new workers by checking they are receiving all the training, materials and resources they need to get started.
The idea is that the technology helps the experienced recruiter to spend more time and energy on the key tasks for which they are qualified: finding truly great candidates, and getting skilled and committed people the jobs they deserve.
A best of breed HR approach will include interactive resumes, but also interaction with candidates. With virtual assistants and chatbots in an assisting role, tomorrow’s recruiters may eventually be able to build a better data set, based on outcomes from prejudice-free hiring. In the meantime, they should not be depending on a computer that learns from our flawed history to make key decisions for them.