WHY THIS MATTERS IN BRIEF
AI is becoming pervasive in the hiring process, so it’s refreshing to finally see a company that’s trying to put people’s potential first.
Employers the world over have long dreamed of harnessing technology to widen their recruitment net and speed up the process of finding and hiring talent. And, if the AI hiring startups in the space are to be believed, employers also want to put an end to human recruiter subjectivity, and use technology to help them recruit a more diverse, representative, and sustained employee base.
The reality though is that recruitment technology and AI seem to be doing the opposite and actually increasing the amount of bias in the hiring process.
One of the major reasons for this ironic state of affairs is that the people developing these systems are training them using very narrow data sets which invariably mean that prejudices creep in. So far this seems to have mostly disadvantaged women, minorities, and more elderly applicants, and is well documented. For example, Amazon canned a promising AI program which, in its little whittling process, was consulting 10 years’ worth of successful applications. No heed was given to the fact those applications were all male dominated until it was too late, and that the machine was unfavorably rejecting women because of this.
Beyond bias, another objection to the broader uptake of recruitment AI is the fact that last year, UK regulators pushed forward a regulation that requires businesses and organizations to explain the decision made by AI. With historical frailties rife, businesses are likely to be risk-averse when it comes to using AI in their recruitment – if a mistake, bias, or trend is undetected then questioned, a big fine could now be on the cards.
All that being said, the use of AI in recruitment – both on a grand scale and in smaller businesses – will help to address the fact jobs are receiving more applicants than ever before, especially during the current pandemic.
One of the ways that AI tries to find a potential match for a role is by matching a candidates competencies with that of the role, and now a Silicon Valley startup called Eightfold have announced their AI is using its smarts to try to not only find a match but to also try and identify candidates with the best fit and potential – something that so far seems unique as other companies try to use almost surveillance-like systems to help them identify potential job hoppers or fakers.
“Today the half-life of skills is just 4 to 5 years,” because of the rapid pace of change as I’ve explained many times before which is accelerating,“so for all the companies looking to develop their talent for the future, just looking in the rear-view mirror – in other words, what skills people have listed in their resume – is not enough,” says Eightfold’s President, Kamal Ahluwalia.
“We recognized this is a fundamental flaw in recruitment practices. Unfortunately, today’s legacy solutions and investments that companies have made are inadequate for today’s business needs.”
Someone who has done a job in the past may not have the skills needed to be successful in that same role in the future. Instead, the startup is looking to make a candidate’s capability – not their direct experience – the first consideration for recruiters.
The key, Ahluwalia explained, is understanding the context: is a particular individual able to learn the skills required for a particular job? What is this person lacking? Can those gaps be filled quickly by shadowing an expert, or taking a class? Are there other individuals within the company or outside who have all the capabilities needed, or are these new capabilities?
“All of these questions and more are addressed by our AI models to make recommendations,” said Ahluwalia. The AI techniques used are enabling customers to tackle diversity, equality, and inclusion issues. Algorithms are EEOC compliant, so personal information around sex, age, ethnicity, pedigree is not considered by the AI.
“We are looking to solve problems for both the employer and the candidate and employee. Resumes are self-attested documents. Our algorithms are able to surface the skills an individual has and can quickly learn to compare these skills against the skills needed for a role. This information allows the hiring manager to run a focused interview process and make an informed decision. It also allows the candidate or employee to take a class or complete a project to address their weak areas,” Eighfold’s president said.
“More importantly, this transparent AI enables the individual to determine what career they want to pursue, and enable them to do what it takes to pursue their aspirations.”
Eightfold say their approach is empowering for both the individual, who is given an indication of areas for their own self-development, and the employer, who is given considerations for more diverse candidates based on capability rather than just prior experience.
“We can do all this with a single AI platform,” Ahluwalia said. “That is very different from the various point solutions you find in the market.”
With more businesses recognizing the importance of a hiring revolution – and developing the tech to make for more comprehensive recruitment decisions – there is plenty of hope for AI in this field in the future.