Click Here To Listen To The Podcast

Podcast Transcript:

Hello, welcome to the first podcast episode of a limited series called Bias In Artificial Intelligence where I Rachel Smith will be discussing the growth in artificial intelligence systems and the biases they have. This episode will be an introduction to the biases in AI. 

As we all know AI systems are being more widely used. We see these systems everywhere and have access to them in our homes and on our phones. Some examples of AI systems are Chatgpt, facial recognition, and financial technology. AI systems are also being implemented into social media platforms such as Snapchat which has created a feature similar to Chatgpt that allows you to message AI and ask for help with different subjects or have a conversion. 

These systems are very intelligent but just like many other new technology they have their errors. AI systems have a bias against people of color and women, a Berkeley research paper about consumer lending discrimination In fintech found that White applicants with the same property and personal characteristics as minorities experience a rejection rate of 20%, compared with the minority rejection rate of 28%. And a study done by Joy Buolamwini showed facial recognition systems are unable to differentiate between Black and Asian individuals. Joy stated “The companies she evaluated had error rates of no more than 1% for lighter-skinned men. but For darker-skinned women, the errors soared to 35%.” Facial recognition is being implemented in policing so these errors can lead to deadly consequences. In addition, recruitment software is another form of AI with bias, Amazon had a recruitment system that would penalize resumes that include the word women such as women swim team captain. These errors can affect the way people of color use technology and they create unnecessary discrimination. 

The reason these systems have a bias is unknown each system is different so the reasons why these biases exist may differ based on the system. In Amazon’s case, they tried to change the code in their recruitment systems but the system still had those biases which resulted in them having to get rid of the system. For systems such as facial recognition, the lack of diversity in test subjects is one of the reasons for its bias. Other AI systems are biased because the systems are searching the web for information in order to output an answer but not all the information online is factual and reliable some of them are biased articles that affect the way AI responds. In addition, AI systems are unable to measure social systems such as racism and sexism as a result it becomes hard to regulate these systems and change their bias. To conclude these biases do not negate the positive aspects of these systems but it is important to be educated on these biases and for solutions to be created.  

Thank you for listening, below this podcast I have created a transcript that links my research paper and relevant sources on this topic so you are able to learn more about the biases in AI. 

At the beginning of my High School journey, I wanted to become a lawyer and planned to major in law once I attended college. Prior to being in my High School’s business program, I was in the law program and that program is the main reason I chose to attend my High School since the law program was one of its main attractions. As I started attending classes related to law I pushed myself to study all day and thought that it was worth it since lawyers get paid well and it was what my parents wanted. I soon realized I was not passionate about law but was hesitant to give it up because I did not want to disappoint my parents. However, I spoke to one of my friends that was in the business program and she motivated me to join the program. In addition, I realized money was not the only factor that should determine my future. 

My time in the business program helped me realize I wanted to attend a business school because I enjoyed working with my peers to create business plans and attending trade shows. The business program helped me realize I did not want to major in law however my interest in the tech industry stemmed from me being inspired by coders such as my aunt and attending stem-related workshops over the summer that allowed me to meet groups of other girls that were also interested in the stem industry and showed me that the tech world does not have to be a male-dominated industry. Furthermore, I chose Computer Information Systems as my major because I wanted the opportunity to learn more about coding and inspire people who look like me to pursue their interest in the tech field. 

In addition to being a CIS major, I am planning to minor in Communication Studies because I think many of us assume that we are good communicators but I would like to understand how to be an effective communicator in depth. Also growing up I have always been more quiet and reserved but after becoming comfortable with speaking in front of people I want to continue to develop my communication skills. Overall both CIS and Communication Studies are fields that I am passionate to learn more about and I hope I can incorporate the things I learn into a future career.