Despite fears by many, artificial intelligence might actually help us get hired and keep us employed—and help us find an even better position than we could on our own.
That’s according to Frida Polli, CEO of Pymetrics. Her company’s AI software lets job applicants skip the résumé-submission process and instead has them take a series of neuroscience assessments. The software tests for factors such as memory, risk-taking, and focus. The results help Pymetrics match candidates with job openings from more than 60 companies including Accenture, LinkedIn, and Tesla.
Polli spoke at this year’s Collision conference, which brought together engineers, entrepreneurs, and chief executives of high-tech companies to discuss emerging trends.
On average, Polli said, 250 people apply for any given job and 75 percent don’t make the cut because of their résumé just based on keywords. Yet the person who ends up getting hired fails at the job 30 percent to 50 percent of the time, according to Harvard Business Review. That’s because innate abilities, and not job experience or education, is what predicts whether someone will do well in a new role, Polli said.
AI IN ACTION
The software looks at applicants’ potential for different occupations by evaluating cognitive and emotional profiles.
Pymetrics develops an algorithm specific to the traits the employers are looking for. The data it uses is based on the assessment results of the companies’ top-performing employees. The algorithm then compares the traits of job applicants to those who are already successful in the roles.
The software relies on objective behavioral data and neuroscience exercises that are the gold standard of neuroscience research, Polli said.
“The current hiring method is using the wrong data and the wrong set of tools to evaluate people,” she said. “Machine learning will help people find their place in the world of work.”
When employers base their hiring decisions on the assessment results, it helps remove gender and racial bias. The open-source software has a built-in auditing system to check for biases in the algorithms.
If applicants aren’t a good fit with the job they applied for, they get a list of other openings that do match their innate skills, Polli said. “Instead of rejection letters, we’re offering candidates new opportunities,” she said.
The software also can help employed people find their next position. Accenture, for example, has partnered with Pymetrics to repurpose its talent, rather than replace workers. For those whose jobs are expected to be automated in the coming years, the company is using the software to determine what other roles the employees would be good at, Polli said.
Personal-care company Unilever did away with using résumés in the hiring process. It used the Pymetrics software to increase the diversity of its staff and bring in more applicants from various socioeconomic backgrounds, Polli said. The software also helped the company reduce its recruiting costs by 25 percent and time to hire by 75 percent.
Pymetrics is now working with philanthropic groups, including the Rockefeller Foundation, to hire people between the ages of 18 to 24 who are not in school and are unemployed. “This population is normally excluded from the workforce,” Polli said.
“Unbiased AI has an incredible ability to balance out some of the things created by humans that have made the world a less fair and equal place to work in,” she said. “Hiring should be based on your potential, not pedigree.”