[Campus Conversations] Welcome to the Machine: Racism in Artificial Intelligence

October 14, 2020

Every year, Gwynedd Mercy University offers Campus Conversations, a series of lectures, panel discussions, and presentations whose purpose is three-fold: to generate conversation as part of our commitment to intellectual inquiry and lifelong learning; to address relevant, timely, and meaningful topics that speak to our mission and core values; to highlight the expertise of our faculty and other scholars in relevant disciplines. This year’s theme is Systemic Racism.

The second event of the series, “Welcome to the Machine: Racism in Artificial Intelligence,” focused on  machine bias, why it exists, and the challenges of creating a technology that is unprejudiced and impartial.

Cindy Casey, Program Coordinator of Computer Information Science at GMercyU, served as the moderator and introduced the two presenters: Ian McAndrew and William Butler from Capitol Technology University located outside of Washington, D.C.

McAndrew kicked off the presentation with a slideshow explaining AI bias, providing real examples, and setting the stage for the evening’s discussion.
  
“We need to be proactive much earlier on and not need to be detecting AI bias, but removing it, removing the possibility before it can happen,” said McAndrew, who serves as Capitol’s Dean of Doctoral Programs “Trying to find bias, I always believe, is incorrect. It’s stopping it which is important.”

Following McAndrew’s presentation, Butler, Capitol’s Cybersecurity Chair, shared a compelling video about gender and racial bias in AI. The video, entitled “AI, Ain’t I a Woman?”, depicted how AI consistently failed to recognize black women, including iconic women such as Michelle Obama, Serena Willams and Oprah.

“All of this is ruled by algorithms. Many of us have no idea that of the little decisions that are made in our lives, many of them are not made by humans,” explains Butler. “How do you think you get the car loan decision so quickly? It’s not a human in the loop, it’s all algorithms.”

Professor Casey then took over to direct the discussion portion of the event, asking McAndrew and Butler to share their expertise on questions relating to if a machine can be biased, how we as individuals can hold tech companies accountable for their bias, and more.

The key takeaway from the presentation is that AI bias is not a result of machines themselves, but rather the humans behind the machines. That’s why it is crucial for the next generation of programmers to recognize, address, and remove their inherent biases.

“If you’re majoring in data science or computer science right now, you have a tremendous responsibility to face your biases and not let bias come into your code or data,” said Butler. “You must be ethical. We all have some bias that we’re conscious of that we have to deal with.”

As the presentation came to a close, Professor Casey opened the floor for participants to engage in the discussion and ask their own questions of the speakers.

To view the entire recording of this Campus Conversation, visit GMercyU’s YouTube Channel.

Categories: