Zufelt’s talk is a part of the 2019-2020 Madison Smith presentation series[a][b][c][d][e] organized by David Fox, Chair of the Department of Interdisciplinary Studies. Madison Smith was an 1873 graduate of Andover who was born into slavery in North Carolina but escaped with the Union Army during the American Civil War, according to Fox. The presentation series was created for faculty members to share work and explorations of systems of power and identity.
Zufelt discussed how companies collect data and use it for profit. He emphasized that people who create artificial intelligence to analyze vast amounts of data should be responsible for their actions and take care in what they make.
“One of the things that caused Snapchat to get to its level of popularity was this idea of ephemerality of the snaps, they go away. What that does is give you a sense of privacy, that this thing is gone and people can’t see it anymore, so life is good. Only, it’s not the case, right? Why would it be the case? Why would the company delete the snaps? They have no incentive to delete the snaps,” said Zufelt.
“Zufelt continued, “In fact, they have a huge incentive to keep the snaps. The first is advertising. They make money by knowing you, and they know you because of your snaps. The other fact is that they’re required by law to collect data, because the law enforcement can come in at any time and request data for suspects for crimes. So you have several thousand people that are gathered around this data set and you tell them don’t touch this data set. And of course [Snapchat] failed to do this, so they were fined by the Federal Trade Commission in 2014.”
Zufelt’s next point was how data visualization could be manipulated, considering an example of giving economic support to one of two hypothetical congressional districts. By changing the buckets in the x-axis of a bar graph of incomes, people can make it seem like the overall better-off community is actually struggling more financially. Audience member Frank Zhou ’22 found this segment particularly insightful.
“I especially enjoyed [Zufelt’s] analysis of the income distribution of two hypothetical congressional districts, as it was a very revealing exercise in the deceptive, misleading potential of data to spin twisted narratives. His analysis showed us that data could easily tell the truth while also deceiving the viewer; that was a very insightful observation,” said Zhou.
While Zufelt noted the different perspectives that people hold about data and large tech companies, he stressed the importance of bringing nuance to conversations about data and engineering. Zufelt pointed out the two extremes of data security and how engineers nor tech entities are exempt from blame.
“[There are] people who feel that large companies like Google and Facebook that are highly leveraging data are terrible, and yet they use them every day, right? I hope that I swing people a little bit more to the nuanced middle narrative. I think it’s not as simple as ‘I’m just an engineer,’ on the coding extreme side. And it’s not as simple as “Google is terrible,” on the opposite extreme either, but I think the truth lies in the middle and we should allow the narrative to be complicated if it needs to be” said Zufelt in an interview with The Phillipian.
Finally, Zufelt brought governments into the picture. He brought up the example of Mark Zuckerberg’s testimony in front of Congress to show that politicians need to be more well-prepared. In addition, he noted that governments should keep their data open so citizens can keep them accountable. This topic reflects Zufelt’s evolution of interests as a teacher—seeing how his technical skills can be used for social justice and to positively impact the world at large.
“Why I chose that topic is that it’s sort of where my own interests are going to as I evolve as a teacher. When I first got here, I was really excited about learning and teaching the technical skills and coding skills. And now I’m really interested to explore, ‘Okay, what can we do with that? Can we do things that actually help people?’” said Zufelt.
Audience member Emma Slibeck ’20 found Zufelt’s presentation style particularly fun and engaging due to the audience engagement. She was glad the talk could clear up some mysteries surrounding data and its use.
Slibeck said, “He’s super engaging and really dynamic as a speaker. There are points where he was walking across the stage, and there was a lot of points where he was calling on the audience. His whole presentation style is also super fun and engaging as well. I think the main takeaway was taking away some of the mystery behind what data is and how it is used, and giving examples and ways of how we should be using data for change, or how our data visualization can be a really positive and powerful thing, and what are the expectations that we should have for our own data.”