A Blog by Jonathan Low

 

Sep 27, 2018

AI and Neuroscience-Powered Job Recruitment and Assessment Is Growing

The issue is not whether it should be stopped, because usage is growing. 

The more important questions are how to make sure that it doesn't reinforce existing biases - and how technological, social and/or legal remedies can be deployed to address the inequities that will inevitably arise. JL


Khari Johnson reports in Venture Beat:
Using neuroscience and artificial intelligence to recruit job candidates, 60 companies use (them) in hiring, including Unilever, Hyatt, and Accenture. Top performers in a position a company is hiring for play games to track their performance and responses in different scenarios and gauge the traits associated with that job. Candidates are compared to these results. Since people already hired may come from a homogenous group, a bias detection tool prevents results that reinforce existing bias. In some instances, companies have seen a 20% increase in diversity and a 65% increase in retention rates.
Pymetrics today announced it has raised $40 million to expand its work with companies using a combination of neuroscience and artificial intelligence to recruit job candidates. The funding will allow Pymetrics to continue its product development and expansion to markets outside the United States.
More than 60 companies use Pymetrics in their hiring practices today, including Unilever, Hyatt, and Accenture. In some instances, companies using Pymetrics have seen a 20 percent increase in the diversity of hires and a 65 percent increase in retention rates.
The $40 million funding round was led by General Atlantic, with participation from Salesforce Ventures and Workday Ventures, as well as existing investors Jazz Venture Partners and Khosla Ventures.
More than 1 million job candidates around the world have played Pymetrics games using the company‘s Android and iOS apps.
Pymetrics begins by asking the top performers in a position a company is hiring for to play a series of games in order to track their performance and responses in different scenarios and gauge the specific traits associated with that job. Candidates are then compared to these results.
Recruitment and the use of AI to match companies with candidates have been on the rise with solutions from companies like Indeed, ZipRecruiter, Vervoe, and Plum, but concern with bias being baked into the process has also increased.
Last week, senators Kamala Harris (D-CA), Patty Murray (D-WA), and Elizabeth Warren (D-MA) sent a letter to Equal Employment and Opportunity Commission (EEOC) members asking questions about how the EEOC assesses claims of discrimination based on algorithmic bias and the use of facial recognition software in the hiring process.
Since people already hired for positions within a company may come from a single, homogenous group, Pymetrics augments its service with a bias detection tool developed internally to prevent results that simply reinforce existing bias. In May, Pymetrics open-sourced a tool named Audit-AI.
“We look at what traits make that population [top performers] unique, and sometimes those traits might be predictive not of job performance but the homogeneity of the people who went through it,” Pymetrics lead data scientist Lewis Baker told VentureBeat in an interview in May. “And so we use Audit AI to make sure that we don’t overweight any traits that are actually more predictive of a certain demographic group.”
In addition to helping companies find new employees, Pymetrics is also used to help companies target existing employees for new positions and to identify potential areas for improvement.

0 comments:

Post a Comment