Facilitators
Huiling Ding, North Carolina State University
Description
Recruiting has been largely automated today, with different AI systems screening resumes, evaluating candidate performance and organizational “fit” in automated video interviews, and scraping the web to build comprehensive social profiles for individual candidates. As an emerging technology, algorithmic video interview evaluators are opaque yet high-stakes black boxes that process interview recordings radically differently from their human counterparts.
Structure/Format
This workshop will examine how biometrics, i.e., facial recognition and voice recognition, are used in automated video interviews to cut down the costs of interviews, speed up the interview processes, and reach out to larger pools of candidates.
We will start with a quick overview of traditional interviews, video interviews, one-way video interviews, and automated video interviews to understand the different functions of these interview formats. We will introduce biometrics, AI ethics such as transparency, privacy, and accountability, and potential biases and discrimination introduced by the use of biometrics (Diakopoulos, 2016; Eubanks, 2018; Gallagher, 2020).
Then we will simulate AVIs by walking participants through the setup of the AVIs before sending them to record their video interviews via Zoom or phone while responding to predefined questions.
Participants will work in small groups to evaluate one another’s video interviews using the metrics of biometric algorithms, share their experiences and struggles doing AVIs, and explore possible ways biases and discrimination can creep into and get augmented in this automated process.
References