Nieman Foundation at Harvard
HOME
          
LATEST STORY
From shrimp Jesus to fake self-portraits, AI-generated images have become the latest form of social media spam
ABOUT                    SUBSCRIBE
Jan. 27, 2015, 12:36 p.m.

At Datalore, data plus storytelling means empathy, humor, and games

At the MIT Media Lab, teams of designers, developers and storytellers pulled stories from eight different data sets.

A handful of Cambridge-area media institutions — including The Non-Fiction Cartel, StoryCode Boston, Harvard’s Bok Center, the MIT OpenDocLab, and the MIT Center for Civic Media — joined forces this weekend to host a hackathon called Datalore that focused on storytelling and data. Around 50 participants, each of whom applied to be there, split into eight teams for the three-day event. Each team worked with a data set supplied by one of their team members; the idea was to “brainstorm and prototype an interactive narrative experience that tells a story with data, around data, or about data.”

As one of the event’s organizers, HarvardX’s Nadja Oertelt, pointed out, some of the groups had to deal with especially emotionally charged data. The consideration each group gave to the sensitivity of their respective topics is reminiscent of last week’s essay on Source, “Connecting with the Dots,” by New York Times software architect Jacob Harris. In it, Harris writes about the experience of routinely taking human tragedy — in this case, “massive sectarian cleansing” in Iraq — and turning them into numbers, datasets, and, ultimately, dots on a map. The challenge he describes is similar to what two Datalore teams in particular had to face.

One of those groups worked on merging two data sets that dealt with state executions in Texas: one that provided names of prisoners and another that documents their last words. Their task was to use this information to highlight the racial and economic injustices of the death penalty system in Texas.

What they came up with was a grid display of prisoner’s photographs; when users click on a prisoner, they can view information about them, as well as play audio clips of some of the prisoners last statements. A text animation types out fragments of these final words; a constantly blinking cursor evokes lives cut short, people executed who still had more to say.

Another group, this one working with traffic fatality data, also faced a the challenge of bringing empathy to a dataset in order to turn it into a story. One of their team members, Amanda Casari, works with Fatality Analysis Reporting System data in her day job as a data scientist. Casari said the process of working with storytellers at Datalore drew her attention for the first time to the data points that are swept aside as outliers in a routine analysis. “I didn’t think about the stories of what the numbers represent,” she said. Together, their team created a website that used interactive visualizations and video components to open a dialogue around individual, personal tragedy.

On the flip side, there was also a pair of teams that dealt with their data set through humor. One of those teams worked with data pertaining to genetic editing, specifically with a set of “all of the double stranded breaks in the genome of cells treated with CRISPR-Cas9.” Dressed up in lab coats, CRISPR team members pitched the audience on a futuristic startup that would allow them to explore and alter their own genetic codes as never before. The joke, one team member said, was meant to highlight the overblown statements about genetic editing in traditional media and on websites like Reddit. Stories about designer babies and the like may generate a lot of attention, but as the team’s data analysis suggested, the science simply isn’t there yet.

A second group, this one working with datasets on for-profit college revenue and student loan debt, turned their data into a parody of an investment pitch. To highlight the predatory nature of these institutions — and the way in which the government supports them — they created The Proprietary Career Institute. The site has helpful information, such as where you might want to open a for-profit college, how to get rich off of government loans, and details on courses like Copywriting & Obfuscation and Juking the Stats 101.

Screen Shot 2015-01-26 at 3.14.09 PM

Both of the Datalore parody projects can be seen as a critique of the media: What comes across as boring on a map or with a data visualization might reach more readers if professional storytellers and designers aimed for laughs.

Other teams experimented with a range of platforms, from plotting atmospheric data on a map to interactive visualizations to an audio and video player that mashes up footage and narration. One group took an extra risk by turning their dataset into a game. Working with data on annual U.S. aid dollars flowing into Indonesia, the team built a board game called Forest Flip that recreates the competing interests of farmers, conservationists, industry and urbanization when it comes to land use. With every role, another year’s worth of aid money is doled out to players, who can spend it on flipping tiles to their advantage. As team member Tin Geber of The Engine Room put it, “It’s basically a bidding war.”

As with all hackathons, Datalore was less about building finished products than about allowing participants to stretch their skill sets and spend time learning how practitioners in other fields think and work.

“The point of the event was to encourage alternative processes and methods of consumption of data,” said Oertelt in an email. “The excitement of the participants and the shockingly sharp execution of projects was proof that people from all different backgrounds — journalism, film, web, and backend development, UX, design, data science — crave creative stimulation in a form that diverges from the standard work day, the standard hierarchy, the standard management structures and corporate influence. It was an amazing thing to watch and be a part of.”

Photo of hackathon organizers Dalia Othman, Nadja Oertelt, Sean Flynn, Heather Craig, and Johnathan Carr by Eric Gulliver of The Non-Fiction Cartel.

POSTED     Jan. 27, 2015, 12:36 p.m.
Show tags
 
Join the 60,000 who get the freshest future-of-journalism news in our daily email.
From shrimp Jesus to fake self-portraits, AI-generated images have become the latest form of social media spam
Within days of visiting the pages — and without commenting on, liking, or following any of the material — Facebook’s algorithm recommended reams of other AI-generated content.
What journalists and independent creators can learn from each other
“The question is not about the topics but how you approach the topics.”
Deepfake detection improves when using algorithms that are more aware of demographic diversity
“Our research addresses deepfake detection algorithms’ fairness, rather than just attempting to balance the data. It offers a new approach to algorithm design that considers demographic fairness as a core aspect.”