15th March 2024
While the BIAS survey has provided us an overview of people’s attitudes towards AI applications in the labor market, it is still of great importance to tell the personal stories. To collect people’s thoughts and own experiences on AI used in labor market, BIAS project is now in the phase of conducting ethnographic interviews and observations in Norway, Turkey, Italy, Iceland, and Netherlands. By talking with stakeholders involved in hiring practices as well as the labor market, we hope to know more about what is going on in the field right now.
The fast speed of AI deployment in this field is surprising. When we started conducting interviews last September, there were only a few of AI-powered tools for HR in the market. Although it was always a heated topic to talk about the future of AI among the HR circle, not many HR practitioners used AI to facilitate their work. Things then get so different from this spring, not only do several developer companies add AI co-pilot functions in their SaaS platforms, but HR people are enthusiastically trying ChatGPT in their daily work as well. The imaginaries, the expectation and the fear towards AI then become the daily interaction with such smart tools, how such discourses stay or change is thus something interesting for us to explore.
Nowadays, developer companies are excavating the full potential of AI, applying this smart technology in different stages of human resource management. For example, it can be used to generate and summarize any text, also screen and rank candidates for a specific job position. When it comes to bias mitigation in the hiring and recruitment, AI is mostly used to neutralize the job description. Based on different expertise such as psychometric test and the detection of gendered wording, AI can analyze the potential stereotype or discrimination existed in the job ad, then suggesting an alternative expression to make it less biased and more inclusive. During our fieldwork, we find out that this utility is the most popular AI application in the HR field, and more and more companies have used such tools provided by different developer companies.
Under such status quo, We are trying to cover a wide range of stakeholders in our ethnographic interviews, from participants from developer companies, HR practitioners who are responsible for hiring and recruitment, to job seekers and current employees that are under the management of AI-powered tools. Nevertheless, with entering the field site deeper and the snowballing from our kindhearted participants, we are surprised to find that this network is more complicated than we think. For example, third parties, like trade unions and civil society organization, are leveraging in the game to mediate the relationships among stakeholders. Such expanding of network then allows us to depict a more complete picture of AI usage in human resource management, advancing the study of AI and bias in workplace.
There still remains a lot of new practices, perceptions, and norms to be discovered. The ethnographic interviews are not only for data collection in BIAS project, but also provide a channel for relevant voices to be heard. Individual stories do matter, and it matters more than just a number, or a box ticked in the evaluation sheet. We hope to demystify so-called omnipotent AI with messy, complicated interaction in the real world, presenting the entanglement of AI and society in this world we live in.
