Bias in algorithms is a costly oversight. AI can contribute to diversity, equity and inclusion
History of Technology in Black Communities
Within communities of color in the United States, there is a history of mistrust in technology – law enforcement, social services, housing and health care have all displayed disparities and inequalities, especially during the COVID-19 pandemic. Cummings explains that even in recent history, many communities have been used as “guinea pigs” in research, so levels of mistrust are generational and trauma-based. Deploying new technologies, such as AI, requires building and restoring that trust and justice.
As AI is meant to assist and replicate human interactions, the last thing technology should do is maintain old human biases and perpetuate harmful and inaccurate stereotypes. The old systems that continue to undermine the future of individuals, especially Black, Indigenous and People of Color (BIPOC), are not a tool that should be considered for deployment.
Because AI has begun to help with functions such as deciding who gets scholarships, mortgages, and opportunities to build economic capital, it becomes apparent that AI must – without a doubt – be unbiased and inclusive. When reproducing human interactions, it should be the best humanity has to offer, rather than the worst.