Can a computer close the gender pay gap? Google says its analytical tool to determine salary is gender blind. The Department of Labor is not so sure. Either way, the conversation on how tech companies can avoid gender pay gaps is gaining momentum.
A study on speech patterns in the Supreme Court reveals female justices are interrupted three times as often as their male Associates, and gender on the court is roughly 30 times more powerful than seniority. The outcome may be a less civil courtroom: the study found female justices learn to conform to male speech patterns by dropping polite refrains and disclaimers like “Excuse me” and “May I Ask”.
A recent review of government websites by the Information Technology and Innovation Foundation (ITIF) revealed many sites are not up to accessibility standards. This affects government employees as well as members of the public who use the sites. One suggestion for improvement is 30-day sprints to get agencies moving on accessibility updates.
We know that Artificial Intelligence can learn. This study shows that it can learn to stereotype, too. When researchers gave the Implicit Associations Test to a computer, they found it had biases just like humans. The researchers say the next step may be to program the AI with “an extra layer of judgement” to account for biases.