In the current landscape of the 21st century, a pronounced digital revolution is underway; there has been a rapid acceleration in the adoption of Artificial Intelligence (AI) and Machine Learning (ML), catapulting them to the forefront of technological advancements. The essay centers on utilizing biased algorithms in AI/ML-based recruitment processes to streamline hiring procedures. Another notable focal point will be corporations that are becoming heavily reliant on open AI models such as GPT-4, Whisper, and DALL-E in their day-to-day operations.
According to Goldman Sachs, 300 million jobs will be lost by AI/ML’s ascendance, driving labor cost savings and raising productivity. Undoubtedly, stakeholders losing out as a result of such integration are people of color, those who belong to a specific race and are of a certain age that segregates them and puts them at risk of discrimination in the workspace and the recruitment process. The California Lawyers Association advises employers to exercise caution while integrating AI/ML tools into hiring processes. They emphasize the importance of conducting a “bias audit” and ensuring transparency in AI decision-making within the business environment (Lewis, 2023).
The exclusionary impact of AI/ML in the labor market
The implications of AI/ML on exacerbating exclusionary labor market patterns are concerning, impacting both the individual and their family. Biased algorithmic patterns lead to the perpetuation of existing societal inequalities, where candidates belonging to a specific group and race may be put at a disadvantage due to ongoing discrimination and cultural biases. Moreover, there may be a lack of diversity in AI/ML development teams, which translates to blind spots regarding potential biases and limited representation of different races and people of color altogether. CNBC’s workforce survey in December 2023 revealed that “53% of Asian and Black workers, and 46% of Hispanic workers are very or somewhat concerned,” compared to 37% of white employees.
Furthermore, the amalgamation of open AI models in business practices poses a threat of unemployment to the older population – in particular, those individuals who are unable to adapt to rapidly changing AI/ML algorithms, therefore becoming occupationally immobile based on their outdated skills and potential (Stypinska, 2022). Ultimately, those people who cannot adapt to changing technological trends and suffer from discrimination in the recruitment process and the work setting are laid off or quit and are no longer in a position to provide for their families.
How do we foster inclusivity in our economic policies to cater to these workers and their families at this critical juncture?
Policy interventions for inclusive technological adoption
1. Conducting annual bias audits
A starting point is to conduct bias audits annually and disclose the usage of any automated process to provide transparency in hiring procedures. New York City’s Bias Audit Law, 2023, strives to place all applicants on an equal footing during recruitment by doing regular audit checks for bias against an individual based on race, gender, and intersectionality. If such regulatory checks cannot be adhered to, then the law prohibits the usage of any automated employment decision tool (Government of New York City, 2023). This is a commendable step given that most applications are now screened with the help of AI/ML software, and it becomes difficult to discern their credibility and impartiality towards a distinct population segment.
However, the other side is that mere transparency is insufficient when economic disparities persist. We need robust policies targeting those individuals disproportionately harmed by biased algorithms. A crucial economic policy involves providing tax incentives for companies to rectify identified hiring biases. This could involve tax breaks for conducting regular, independent bias audits and implementing corrective measures like diversifying the training data and creating fairer scoring metrics. This eventually creates an economic incentive for companies to make an effort to actively eliminate bias, promote better hiring practices, and benefit a wider pool of qualified candidates.
2. Leveraging upskilling programs
Secondly, it is imperative to counter the idea of “AI-Ageism,” and this becomes a reality when companies invest in the skills of their employees, teaching them open AI models and aiding them in improving their skills. An effective economic policy to promote this initiative is encouraging government-subsidized upskilling programs specifically targeting the older demographics. These programs could focus on training tailored to open AI models like those provided by Schneider Electric USA.
Schneider Electric USA has launched an online platform called “Open Talent’s Market” (OPM) that aids in not only improving the tech-related skills of its employees but also gives them more agency in choosing the right kind of project that aligns with their skills. There is a trade of skills in the organization, and employees can learn from one another, ultimately catering to removing any existing skill gaps (White, 2023). AI/ML models can be complemented with human expertise and knowledge, where they exist in harmony rather than replacing human capital entirely. Therefore, this initiative by Schneider Electric is laudable in retaining employees while enhancing their capacities and promoting prosperity for them and their families.
In conclusion, navigating the stream of AI/ML requires responsible implementation – a mixture of robust economic policies such as tax incentives for addressing bias and government-funded upskilling programs are critical to ensure inclusivity and empower individuals to thrive. With better planning mechanisms, and foresight these policies can be tailor-made to provide an equitable future.
