Museum discusses ethics of big data

Tech industry in ‘moment of soul searching’

Three contributors to the MIT Museum exhibit Big Bang Data convened to discuss to discuss big data justice at the museum Tuesday. The panel discussed their views on the biggest issues facing society with regards to big data and how best to move forward.

The panel included Jose Luis de Vicente of Sonar+D, visiting curator of Big Bang Data, Joy Buolamwini G of the Media Lab and founder of the Algorithmic Justice League, and Catherine D’Ignazio, assistant professor at Emerson College and research associate at the Center for Civic Media.

de Vicente said the impetus for the exhibit was the transition from data being created by large institutions to being driven by social activity.

When asked about their view on the biggest issue in big data justice, D’Ignazio emphasized the need to prioritize the perspectives of those most marginalized throughout the decision-making process. She mentioned reading a paper regarding a participatory smart city, where the only stakeholders at meetings were researchers, developers, and construction firms.

Buolamwini also noted a lack of representation, describing an issue in facial recognition with “supremely white data put into predominantly male analysis.”

In discussing ways to educate people, Buolamwini described a ProPublica investigation which raised public awareness of racial bias in machine learning algorithms and said that the Algorithmic Justice League she founded lowers the barrier for more people to make similar discoveries.

D’Ignazio said that emphasizing data literacy in education and having journalists describe their methodology when presenting data can help the population better determine whether an analysis is trustworthy. When convincing people that data usage is important, she recommended telling stories of data misuse.

de Vicente argued that the need for political pressure to regulate usage of big data is more critical than increasing awareness, an example being the lack of transparency when Volkswagen used code to fake emissions tests.

D’Ignazio responded that to get public will for political change, engagement is necessary.

When asked about tools for teachers, D’Ignazio described the importance of informing students of issues while not disempowering them. She listed a few resources: the AdNausium browser add-on, which deliberately clicks on all advertisements to obfuscate browsing data, the LightBeam browser add-on, which visualizes third party tracking of internet usage, the site, which lowers the bar to use data and has lesson plans for educators, and data aggregator Acxiom, where users can view the collected information on themselves.

de Vicente said that the tech industry is in a “moment of soul searching.” The industry thought its innovations were for the general good, but the good world envisioned did not materialize, he explained.

While none of the panelists described a line between ethical and unethical uses of data, Buolamwini said that she would definitely classify lethal autonomous weapons with the ability of facial recognition as unethical.

The Big Bang Data exhibit is available until the end of March. MIT ID holders receive free admission to the museum.