“Give people the right to talk to a human”
We sat down with Denis ‘Jaromil’ Roio to talk about, among other things, what software developers can learn from the world of agriculture.
“I am a programmer, so I develop software but I am trained in the humanities. And I am the co-founder of Dyne.org which is a non-profit software house,” he says. “We do bottom-up development and we do that in a very interdisciplinary way. So our research is informed by ethnographic practices but also artistic creation.”
Dyne.org was established in the Netherlands 15 years ago as a hacker community and quickly attracted a large group of volunteers. Today, it is a think & do tank that also counts researchers and artists among its members. They collaborate on projects mainly for the public sector, and it is here that Jaromil sees an urgent need for the public to understand the politics of algorithms.
What is ‘algorithmic sovereignty’?
“The word algorithm is pretty common. It is mostly about code that executes in a way that processes events, inputs, situations into a reaction, an output, an assessment of that situation,” he says.
“The second word, sovereignty, is a little bit ambiguous when translated into English. I think the best etymology for the concept is found in Spanish where ‘soberanía’ has been widely used as a term by Via Campesina, by food self-determination movements for food production and agriculture that have fought the patenting regime of companies like Monsanto.”
One of the key grievances of these movements has to do with intellectual property rights to crops, especially GMO crops, taking away control over their livelihood from farmers that become dependent on giant agribusiness companies. Jaromil sees the same dynamic play out in the digital world where closed source, proprietary algorithms are taking over systems of governance.
“Only if the software is free and open source so the code is accessible can it be part of a political process of negotiation. Because code is not neutral per se, algorithms are not neutral per se, they need to be negotiated, observing clearly not only the results that they produce but also the code they are made of. So we need a literate approach to algorithms,” he says.
“So algorithmic sovereignty refers to this condition in which we need to reestablish as participants our negotiations, our political space, our common ground with algorithms.”
Is open source code enough to solve this problem?
“I think it’s a precondition. Without that we cannot even lay the foundation of a process, but it is not enough because code needs to be understood. So perhaps a slower process is needed in order to document it and translate the documentation into a language that does not exclude people,” he says.
“The code must be like the law, it should be declared, it should be open, so it can be applied in an equal way, which is a foundational principle for democracy.”
Besides breaking open the black boxes of algorithmic governance, he also points out that an open world assumption is necessary for policy makers.
“In philosophy and political science, the open world assumption is a basic assumption that says that all the cases observed in law are not all the cases existing. So it’s always known that the world is more open and bigger than what we envision it to be in our conditions, in our code, in our societies. So we need to always leave an open door for these conditions to manifest and explain themselves,” he says, adding that there is a basic, low tech fix to implement this in practice:
“And this is simply solved by giving people the right to talk to a human and not making them fall into the cracks of an accelerated algorithmic governance regime.”
What is an example of an alternative way of doing algorithmic governance?
“I can share some experiences in good design practices that are making my own work more pleasant and successful. There are practices in design-intensive systems that we could leverage in different fields. And one of them is found in permaculture and biodynamic agriculture and these sort of disciplines that observe how systems are growing,” he says.
He refers to this design-intensive approach a historical third way. Exemplified in agriculture, the first way was labor-intensive in pre-industrial farming. The second way is resource-intensive.
“Industrial agriculture consumes a lot of oil and energy, but now some people are trying to build systems that are design-intensive, that are designed for diversity, that are nurturing and maintain the ecosystem. So just going through a part of this literature on permaculture, I realized how useful it is also to software design, a field that is completely different.”
According to Jaromil, one of the keys to creatively solve the problems that societies face as they increasingly get governed by algorithms, is interdisciplinary thinking. In this way ideas and systems from one field can cross-pollinate and inform and influence other fields, like applying permaculture ideas to software development or using processes from software development elsewhere.
“In the same way, in agile development, if we remove the references to software, it becomes immediately a criteria of work, of organizing work in horizontal, non-hierarchical ways, but very productive also in other spheres of production,” he says.
Correction: In a previous version of this article, it stated that GMO crops were “often engineered for infertility” which is incorrect. The origin of this misconception is actual research patented by Monsanto into the so-called Terminator gene that makes plants produce sterile seeds. However, the patented gene was never put to market.
