Addressing Bias Against the Poor in Artificial Intelligence by Observing Individual Bias in Twitter
Publication Date: Spring 2023
Abstract
With an increase in Artificial Intelligence (AI) usage, there is an increasing demand to address biases and implement fairness in AI. This white paper focuses on the importance of first recognizing and addressing societal biases against the poor to address biases against the poor in AI. The results of the study are from the first collection of data collected from Twitter. The data is used to provide an index of how people in societies perceive poor people.
Why AI skeptics should care
- Bias, particularly bias against poor people, is not just present in AI. Humans also display biases.
- Addressing human biases can help address and minimize AI biases because algorithms are programmed by and learn from human behavior.
- AI has already been adopted in many industries like healthcare and finance. So, the focus should be less of eliminating AI and more of improving it.