Biased World = Biased AI

On By earickajIn Media
Forbes

Who hasn’t thought about the possible effects of AI? It is becoming an integral part of our everyday lives. The biggest fear most of us have is that AI is going to take over but I’d argue that a more concerning issue is that it is going to perpetuate the longstanding racist ideals that have been breathing air into our society for ages. 

Starting Point

For this assignment, I went the route of exploring the role bias plays within AI’s Generative features. I got the inspiration from one of our lectures where we discussed how when an individual inputs a specific kind of idea into an AI resource, the output is often biased or stereotypical. For example, older white men pop up when one types in “professor” even though that is not the only kind of representation of those that identify as professors in the real world. 

“Enough of Biased AI” by Earicka Jones

Why is it so Hard?

“breaches of representational fairness occur “when systems reinforce the subordination of some groups along the lines of identity” — whether because the systems explicitly denigrate a group, stereotype a group, or fail to recognize a group and therefore render it invisible” (Samuel, 2022). In this assigned reading from lecture, it details the multitude of reasons why it is so hard to make AI inclusive for everyone. Within my remix video, I focus more on this reason that is somewhat mentioned in this quote regarding systems maligning or forgetting about certain groups of individuals.

Stanford

Where to Start?

It is often people of color that are either forgotten about or disgraced within AI. I looked into index reports that gave percentages of the amount of individuals that have received PhDs in AI and it showed that the recipients were overwhelmingly white and men. One can infer that if it is primarily white men inputting knowledge into these AI tools, they are going off of stereotypes and biases they consumed about people of color in the real world because they don’t identify with them.

Dangerous Consequences

I wanted my video to be directed both those that currently have the power to alter the nature of AI and those who are being affected by it. “Given that facial recognition is increasingly used in police surveillance, which disproportionately targets people of color, a system that is better at identifying Black people may also result in more unjust outcomes.” (Samuel, 2022). I included a clip where a black man named Robert Williams was falsely accused of a crime due to the bias instilled in AI in order to stress that there are dangerous effects of biased AI as this situation could have easily ended in Williams being a victim of police brutality or unjustly having to spend his life behind bars. 

Joy Buolamwini

Advocating and Solutions?

Joy Buolamwini, an expert in biased AI, was a figure I spent a great deal researching for the remix as she was integral in the process of bringing light to AI’s biased tendencies particularly in regards to it misgendering famous dark skinned women and not being able to recognize dark skinned individuals altogether. She created the Algorithmic Justice League which I put in my video as a credible source because it centers around spotlighting how harmful AI can be and find ways to hold it accountable. I hoped to curate my remix in a way that illustrates to the viewer, vivid examples of AI’s bias in an effort to ensure they can’t shy away from what is happening and what will continue to happen if they don’t educate themselves.

 

Codedbias

Sources:

Leave a comment