Mechanical Nun Parses Sins, Dispels Your Guilt Via API

Sinners, you are not alone. That is the message a Mechanical Nun art project currently underway in San Francisco hopes to deliver.

The nun will record your confession, Parse your sins, and link those sins to a location. The data will then be made available in an API for visualization on a map so sinners can see they are in good company, and hopefully, as a result, be able to let go of some guilty feelings.

The work will debut November 2 at the Yerba Buena Center for the Arts as part of Project Nunway, a charitable fundraiser by the Sisters of Perpetual Indulgence, an order of queer nuns devoted to community service. By late summer, the Mechanical Nun will make her way to Burning Man.

I learned about this project through Oren Schaedel, the project's lead, whom I met at API World earlier this month. Schaedel, a burner, a scientist, and a programmer, came up with the idea while hiking in Marin County with a group of friends, among them Sister Baba Ganesh, who is a member of the Sisters of Perpetual Indulgence.

During the hike, Sister Ganesh (who is male) told Schaedel he wanted a different type of art project for Project Nunway, one that would engage more burner, geeky types of artists.

What immediately sprung to Schaedel’s mind was the Electric Monk, a character from Douglas Adam’s novel “Dirk Gently's Holistic Detective Agency.” In the book, humans were too busy to believe in anything anymore, so they outsourced the job to robots.

Sister Ganesh liked the idea, so Schaedel pulled together a group of nine other burners, scientists, and startup enthusiasts to volunteer their time on the project.

Details of the project

As Schaedel explained to me, the project is still coming together, but the team is at a point now of where they are simply “connecting the dots.”

The nun herself is a telepresence robot dressed up as a nun, which an operator can speak through and control from a distance. The nun’s creators are still trying to figure out if the nun will listen to confessions or simply point people to a confession booth.

“The venue where we are launching it at is going to be very loud,” said Schaedel. “It is more likely the nun will point people to a confession booth where an Android tablet will record the confessions.”

An app on the tablet will ask you where you sinned and prompt you to select a category of sin from a sin tree. You will then speak your confession.

The confession and location information are then sent via Wi-Fi to a Raspberry Pi stationed in the booth. A credit-card sized computer, the Raspberry Pi uses a Machine Learning algorithm to parse the language and categorize sins.

“Most confessions are not about a single sin but a group of sins or just a guilty pleasure,” said Schaedel. “When a person selects a category [from the sin tree], it helps the algorithm pick the keywords from the spoken confession and assign each confession to at least one category.”

The Raspberry Pi then generates a JSON object for each confession and sends that data to a database in real time.

“We are still prototyping the database,” said Schaedel. “It will probably be a simple SQL server but the idea is that people can pull up the data to visualize, categorize, and use the information in different ways.” He emphasized the information will remain completely anonymous with no personal information attached to the confessions.

He further explained that the first visualization will be a simple map of San Francisco overlayed with neighborhoods where the sins took place. The map will be available for online viewing shortly after the Mechanical Nun debuts at Project Nunway.

Eventually the app will go online as well, so it can begin collecting sins from all over the world, Schaedel said. At some point, the Mechanical Nun API will become public, too, but the team will release it in phases depending on how comfortable they are with the anonymity of the data.

Schaedel said people React to the project differently. Some find it interesting, while others find it alienating. “Some people want a human to sympathize with them instead of an interaction through a machine,” he said.

He points out, however, that it is not uncommon for people to anonymously write about something they feel bad about online either in a forum or through a service like PostSecret. People simply want to “get something off their chest and say something about what they are feeling,” he said.

Hopefully, the Mechanical Nun will help them do that.

Be sure to read the next API article: Kodak Launches Developer API to Help Monetize Photo Apps