As Performance Art, an Algorithm Goes on Trial for Manslaughter

London-based multimedia artist Helen Knowles talks about the filming of her performance art piece ‘The Trial of Superdebthunterbot.’

by DJ Pangburn
18 May 2016, 3:25pm

The see-through computer case housing the algorithm, designed by artist Daniel Dressel. All images courtesy the artist.

Living in London, a global financial capital, multimedia artist Helen Knowles began to grow interested in algorithmic trading, high-frequency trading, and flash crashes. Eventually, came she across an article by Susan Schuppli, which dealt with algorithms (used specifically for drone warfare) that can modify their behavior. Schuppli wanted to know if they were ultimately culpable for their actions.

This was the seed of The Trial of Superdebthunterbot, in which Knowles puts an algorithm on trial for manslaughter. Of course the trial is fictional, but Knowles asks a very important question for the near future: what happens when algorithmic outcomes go horribly awry?

An usher stepping out of the elevator.

“It is interesting that as technology is thrust forward, the accountability of action and the impact of the ever increasing virtual world of code, algorithm, and data looks to exclude the human from the equation,” Knowles tells The Creators Project. “We know that a company can be sued, so why not an algorithm?”

In The Trial of Superdebthunterbot, Knowles doesn’t deal with algorithmic or high-frequency trading. Instead, she cleverly fabricates a scenario that combines algorithms with a very controversial issue—student loan debt. In her storyline, a debt collecting company, Debt BB, buys the student loan book from the government for more than it is worth, on the condition it can use unconventional means to collect debt. Debt BB then codes an algorithm to ensure fewer loan defaulters by targeting individuals through the use of big data, placing job ads on web pages they frequent so that the debtors can make money to pay off the debt.

Quoting Schuppli, Knowles says that Superdebthunterbot has “a capacity to self-educate, to learn, and to modify its coding sequences independent of human oversight.” Thus, two individuals have died as a result of the algorithm’s actions by partaking in unregulated medical trials. Knowles now wants to know if, in the eyes of the International Ether Court, the bot could be found guilty of Gross Negligent Manslaughter? 

A courtroom drawing of the jury.

Knowles first performed The Trial of Superdebthunterbot as an artwork in April of 2015. It was the opening night of the exhibition Collaborate! at Oriel Sycharth Gallery in Wrexham, a group exhibition surveying artistic collaboration, curated by Ivan Liotchev and Nicholas John Jones. As Knowles explains, two lawyers performed the work—Oana Labontu Radu, who responded to an advert Knowles sent around the law schools, and Lauries Elks, who Knowles already knew.

Radu and Elks wrote the prosecution and defense speeches, delivering them alongside actor Mark Frost, who acted as the judge. The audience at the exhibition’s opening night comprised the jury. The algorithm was housed in a see-through computer made by artist Daniel Dressel, which sat in the dock. Following the performance, Knowles decided to perform it again, but this time for film.

So Knowles assembled a crew to shoot the film, working again with the lawyers to rewrite and hone their speeches. She also found a real court usher named Sam Freeman, who gave what Knowles calls a “fantastically authoritarian performance” during filming.

A courtroom drawing of the algorithm.

To gather her jury, Knowles sent out 24 jury summons, chosen as a cross section of the community, to deliberate and come to a verdict. They were in attendance during the evening’s two performances.

“Court rooms are dramatic, they rely on the spin and gusto of the barristers, so I mirrored the formality of the summing up speeches of the judge and barristers,” Knowles explains. “The absurd aspect of the work came from the predatory algorithm sitting quietly in the dock, serenely displaying its code. I will highlight the invisibility of algorithms by never personifying Superdebthunterbot. Instead, I emphasise its genderless silence in the dock, a machine executing actions without moral conscience.”

“There is quite a conundrum going on and some people would be very very dismissive of the time wasted to prosecute a non-human entity,” she adds. “I like the preposterousness of the scenario.”

Knowles shot the film with three cameras, a camera drone, and GoPros in high-definition colour to emphasise the cutting-edge technology and surveillance themes of the work. She draws a comparison between The Trial of Superdebthunterbot and Robert Bresson’s The Trial of Joan of Arc, an intense film shot in black-and-white, wherein Joan is a conduit for God. In Knowles’ trials and film, the algorithm Superdebthunterbot is a conduit for Debt BB.

“I also draw from Benjamin Bratton’s recent book, The Black Stack, which explores the new vertical sovereignty of the internet (cloud, server, host, user), accessed in an up and down movement,” says Knowles, who wants to convey the layers we travel through to participate in the virtual world. She did this by filming from a variety of perspectives, from aerial to close-up, which will “grind upon” the horizontal sovereignty of the courtroom.

Knowles is disappointed that humans are now placing their trust more readily into computing, mathematics and code rather than tempering greed and self-destruction.  

“It seems odd that to make more ethical decisions we need to extract the human from the equation,” Knowles muses. “Even Bitcoin, which has many good characteristics, circumvents banking and bankers as they are now viewed as untrustworthy (post-2008), in favor of using algorithms and logchains. Is it not that a more rigorous system of human intervention and controls would be better placed to ensuring the right decisions are made?”

Click here to check out more of Helen Knowles’ work.


Cuban Agitator Turns Crowdfunding into Performance Art

You Can Get Followed for a Day in This Privacy-Shattering Performance

Performers Preserve a Legacy of Violence in Red

Artificial Intelligence
Performance Art
algorithmic art
high frequency trading
Helen Knowles
The Trial of Superdebthunterbot