Machine Learning and Algorithmic Bias by Kristine Conway https://padlet.com/kristinelconway/6uknzbpd91vcnptv Algorithms are the specific set of steps used to solve any computational problem by breaking it down into explicit steps. When computers perform algorithms, they can take a problem that would take humans a lifetime to perform, instantly (Wachter-Boettcher, 2017). Although algorithms can help to make our lives easier, it can also reinforce existing social inequalities as human bias makes its way into the machines learning model. When using these algorithms, it's important to question where and when the data is coming from and the unconscious biases of the programmer (Wachter-Boettcher, 2017). Below I will be comparing algorithmic bias in two popular search engines: Google Search and DuckDuckGo and will analyze the nature of bias across each search engine. en-us 2022-10-12 18:44:48 UTC 2025-11-17 22:08:14 UTC hello@padlet.com "Mathematician" kristinelconway https://padlet.com/kristinelconway/6uknzbpd91vcnptv/wish/2337558627 2022-10-12 19:01:11 UTC https://padlet.com/kristinelconway/6uknzbpd91vcnptv/wish/2337558627 Google Search kristinelconway https://padlet.com/kristinelconway/6uknzbpd91vcnptv/wish/2337558961 2022-10-12 19:01:27 UTC https://padlet.com/kristinelconway/6uknzbpd91vcnptv/wish/2337558961 "Indigenous Women Canada" kristinelconway https://padlet.com/kristinelconway/6uknzbpd91vcnptv/wish/2337559873 2022-10-12 19:02:12 UTC https://padlet.com/kristinelconway/6uknzbpd91vcnptv/wish/2337559873 "Caucasian Women Canada" kristinelconway https://padlet.com/kristinelconway/6uknzbpd91vcnptv/wish/2337560300 2022-10-12 19:02:31 UTC https://padlet.com/kristinelconway/6uknzbpd91vcnptv/wish/2337560300 Google Search kristinelconway https://padlet.com/kristinelconway/6uknzbpd91vcnptv/wish/2337562544 2022-10-12 19:04:12 UTC https://padlet.com/kristinelconway/6uknzbpd91vcnptv/wish/2337562544 "God" kristinelconway https://padlet.com/kristinelconway/6uknzbpd91vcnptv/wish/2337562865 2022-10-12 19:04:25 UTC https://padlet.com/kristinelconway/6uknzbpd91vcnptv/wish/2337562865 Google Search kristinelconway https://padlet.com/kristinelconway/6uknzbpd91vcnptv/wish/2337562968 2022-10-12 19:04:31 UTC https://padlet.com/kristinelconway/6uknzbpd91vcnptv/wish/2337562968 Duck Duck Go kristinelconway https://padlet.com/kristinelconway/6uknzbpd91vcnptv/wish/2337565008 2022-10-12 19:06:10 UTC https://padlet.com/kristinelconway/6uknzbpd91vcnptv/wish/2337565008 DuckDuckGo kristinelconway https://padlet.com/kristinelconway/6uknzbpd91vcnptv/wish/2337565274 2022-10-12 19:06:23 UTC https://padlet.com/kristinelconway/6uknzbpd91vcnptv/wish/2337565274 DuckDuckGo kristinelconway https://padlet.com/kristinelconway/6uknzbpd91vcnptv/wish/2337565523 2022-10-12 19:06:35 UTC https://padlet.com/kristinelconway/6uknzbpd91vcnptv/wish/2337565523 Analysis kristinelconway https://padlet.com/kristinelconway/6uknzbpd91vcnptv/wish/2337571321 The narrative Google Search and DuckDuckGo presents here is that mathematicians are predominantly white, seemingly affluent, males from the past who write formulas on chalkboards. Although Google search did include 3 females, one white, one Asian and one black, 83% of the pictures are still predominantly white, males from the past. There are no male people of colour represented. This type of narrative is dangerous as it may dissuade these minorities to pursue a career in this field. In his interview “Why is Everything White?” Muhammad Ali describes how as a child, he questioned if he would go to Heaven because there was an absence of black representation in the media (BBC, 2020). For kids, representation is meaningful as they search for cues in order to develop and shape their understanding of ethnic-racial groups (Rogers, 2021). Children may be more inclined to choose this career if they see themselves represented because, “It’s a different feeling when you get to see somebody who looks like you in a career you’d like to pursue” (Klein, 2021). ]]> 2022-10-12 19:11:06 UTC https://padlet.com/kristinelconway/6uknzbpd91vcnptv/wish/2337571321 Analysis kristinelconway https://padlet.com/kristinelconway/6uknzbpd91vcnptv/wish/2337571503 Interestingly, the Google Search and DuckDuckGo were quite opposite with their results. The Google search showed the first 17 images as the popular Christian depiction of God as a white male or an image of a cross, whereas the DuckDuckGo search engine showed predominantly Eastern depictions of God as Hindu, Egyptian, Buddist and Greek. Both depictions seem to be extremely unbalanced when it comes to ethnically diverse representation. There are over 4,000 recognized religions in the world and many of them believe in one or more Gods. Why are more not represented in these searches? The lack of diversity in the search may make people question their own beliefs, much like how Muhammad Ali questioned his own religion, due to a lack of representation of black people in religious media (BBC, 2016). A lack of ethnic and religious representation in these searches may lead a person to questioning their religion.   ]]> 2022-10-12 19:11:16 UTC https://padlet.com/kristinelconway/6uknzbpd91vcnptv/wish/2337571503 Analysis kristinelconway https://padlet.com/kristinelconway/6uknzbpd91vcnptv/wish/2337571610 After searching “Indigenous Women Canada,” in both Google search and DuckDuckGo both had similar results. Both were quite political, with strong keywords such as “crisis,” “genocide,” “murdered,” and “stolen” and many of the pictures portray Indigenous women as sad, angry or stoic protestors who have had enough, this was especially more prominent in the Google search. DuckDuckGo included images of smiling indigenous women dressed in traditional clothing, however the captions below include many of the strong keywords listed above. The narrative this suggests is alarming. Though these are issues many Indigenous women are facing it should not be the only pictures we see. Indigenous women are more than just their political views and they should portrayed that way. As the article, “media portrayals of missing and murdered Indigenous women suggests, “media representation of indigenous women has had an impact on how they are seen as victims of violence” (MediaSmarts, 2022). These images only further portray indigenous women as victims or protestors, and is only another construct perpetuated by the media.]]> 2022-10-12 19:11:21 UTC https://padlet.com/kristinelconway/6uknzbpd91vcnptv/wish/2337571610 Analysis kristinelconway https://padlet.com/kristinelconway/6uknzbpd91vcnptv/wish/2337571796 After searching “Indigenous Women Canada” in both search engines, I thought it would be interesting to see what would come up if I searched “Caucasian women Canada.” And there was a clear contrast between the two. The images from both search engines do not have many political pictures, apart from 2 images from the Google search which depict happy and smiling women doing their jobs and many pictures of women smiling and holding the Canadian flag. The narrative we get from this is Caucasian women are very patriarchal and happy to be Canadian; a stark contrast to the way in which Indigenous women are represented as angry, sad and stoic victims of oppression. This emphasises Yasmin Jiwani’s assertion that the “fallen women stereotype [that] makes Indigenous victims seem less sympathetic to media” (MediaSmarts, 2022). Also the captions underneath the images of Caucasian women are captioned with words such as “beautiful, ”gorgeous” and “student,” there are no images of Caucasian women protesting or standing up for their rights.]]> 2022-10-12 19:11:29 UTC https://padlet.com/kristinelconway/6uknzbpd91vcnptv/wish/2337571796 What are some solutions to addressing algorithmic bias? kristinelconway https://padlet.com/kristinelconway/6uknzbpd91vcnptv/wish/2337608229 The use of algorithms is not something that will be going away anytime soon and though they are not yet perfected, there may be a way to ethically use these algorithms to our advantage in the future. Some ways to improve these algorithms and reduce bias are as follows. First there should be transparency on where and when data is being collected, “reliance on historical data is a fundamental problem with many algorithmic systems” (Wachter-Boettcher, 2017). The data used should be free and accessible to everyone so we can see “which assumptions might be encoded in it, and whether it represents users equally.” (Wachter-Boettcher, 2017). There should also be equal access to courses and training so we can all understand the data we are seeing; “if we live in a “scored society” (at 25) then everyone in the legal system must understand how that score is tallied. Training for lawyers and judges must be available to ensure data literacy” (Silver & Christian, 2020). Though I do not believe this should be limited to only lawyers and judges. We should also make sure there is a diverse workforce helping to develop the technology as it begins to have a bigger and bigger impact on society. Programs like the Princeton summer camp (Klein, 2021) can help diversify the data being used, rather than just one specific segment of the population. We need to not only “demand” the tech industry “takes responsibility for the data it collects” (Wachter-Boettcher, 2017) but we also need to take responsibility for ourselves, educate ourselves so we can understand and change the narrative. ]]> 2022-10-12 19:43:04 UTC https://padlet.com/kristinelconway/6uknzbpd91vcnptv/wish/2337608229 References kristinelconway https://padlet.com/kristinelconway/6uknzbpd91vcnptv/wish/2337608897 BBC. (2020, June 10). Muhammad Ali: 'why is everything white?'. BBC News. Retrieved October 13, 2022, from https://www.bbc.com/news/av/world-us-canada-52988605 
Klein, A. (2021, January 15). A summer camp with a long plan: Keeping bias out of artificial intelligence. Education Week. Retrieved October 13, 2022, from https://www.edweek.org/teaching-learning/a-summer-camp-with-a-long-plan-keeping-bias-out-of-artificial-intelligence/2019/08 
Rogers, O. (2021, October 20). Why representation matters in kids' media. Common Sense Media. Retrieved October 13, 2022, from https://www.commonsensemedia.org/kids-action/articles/why-representation-matters-in-kids-media 
Silver, Lisa and Gideon, Christian. “Harnessing the Power of AI Technology; A Commentary on the Law Commission of Ontario Report on AI and the Criminal Justice System” (November 18, 2020), online: ABlawg, http://ablawg.ca/wp-content/uploads/2020/11/Blog_LS_GC_LCO_Report.pdf
Media portrayals of missing and murdered indigenous women. MediaSmarts. (2022, September 26). Retrieved October 13, 2022, from https://mediasmarts.ca/digital-media-literacy/media-issues/diversity-media/indigenous-people/media-portrayals-missing-murdered-indigenous-women 
Wachter-Boettcher, S. (2018). Chapter 7. In Technically wrong: Sexist apps, biased algorithms, and other threats of toxic tech. essay, W.W. Norton & Company, Inc. 
 
 
]]> 2022-10-12 19:43:44 UTC https://padlet.com/kristinelconway/6uknzbpd91vcnptv/wish/2337608897 Google Search kristinelconway https://padlet.com/kristinelconway/6uknzbpd91vcnptv/wish/2339293113 2022-10-13 18:50:27 UTC https://padlet.com/kristinelconway/6uknzbpd91vcnptv/wish/2339293113 DuckDuckGo kristinelconway https://padlet.com/kristinelconway/6uknzbpd91vcnptv/wish/2339293318 2022-10-13 18:50:37 UTC https://padlet.com/kristinelconway/6uknzbpd91vcnptv/wish/2339293318