In order to eliminate bias, I would develop the app “Bias Decoder.” This app’s main object is to analyze job descriptions for biases and suggest edits to ensure a fair hiring process.
Currently, if a certain ethnicity or gender rises through the ranks of a company unproportionately, they will likely hire people who are similar to them. This is a problem because more diverse workplaces lead to increased success, not less diverse ones.
Bias Decoder will mitigate these biases by using AI to remove vocabulary, descriptors, or duties that may hinder certain applicants unfairly. Removed topics could include gender, ethnicity, religion, and age. Elimination of these unfair descriptors will enable recruiters to evaluate applications solely on qualifications.
In addition to its use for online applications, it could also be applied to virtual interviews. The software could partner with other applications like Zoom to remind interviewers of bias-based pitfalls. When interviewers or interviewees do or say certain things that may induce biases, the AI could identify this and use messages to remind the interviewer to remain impartial. While this may seem unnecessary, biases are often subconscious, so these messages will force interviewers to remind themselves of impartiality.
To address Cathy O’Neil’s algorithm audit and ensure Bias Decoder passes the test, we must evaluate its data integrity, accuracy, long term effects, and definition of success. First, the data being used in the app is likely to be inherently biased, which will hopefully be removed. Second, success for the platform means that biased data (such as a job application or interview) is removed from any biased information to enable recruiters to evaluate a job candidate fairly. Third, the accuracy of this app may be difficult to determine. The app will likely go through multiple versions to fine tune its accuracy. Finally, long term effects of this app could include filtering out valid candidates by accident, which would hinder the recruiter’s job.
Image source:
https://www.peoplematters.in/site/interstitial?return_to=%2Farticle%2Ftalent-acquisition%2Fhow-ai-is-transforming-the-hiring-process-16517
