Revista Anfibia | “Ancestral technologies and green algorithm. Erase my data, why not?”

Anfibia FAIR Collaboration

Revista Anfibia | “Ancestral technologies and green algorithm. Erase my data, why not?”

October 25, 2023

By: Marina Otero | Art: Javiera Cisterna Cortés | Source: Anfibia Chile

Data, space, storage, emptiness, memories. Information, memories, notes, postcards. The production and storage of digital information reaches an unprecedented volume. The demand for data storage could soar up to 50% by the end of this decade. What is its impact on social heritage memory, on the environment, on neocolonialist relations, on our subjectivities? Marina Otero proposes “getting rid of our data through mourning processes” and implementing new paradigms for the production, consumption and storage of digital information.

MCHAP 0780 is the largest quipu on display at the Chilean Museum of Pre-Columbian Art in Santiago. It was found in an Inca cemetery in Arica. Dated 1500, this information recording system used by the Incas is composed of 586 knotted camelid fiber cords, organized in 8 sections of 10 sets, each with up to 13 sub-levels containing 15,024 units of data whose meanings are still unknown. State officials made them to store census records, accounts, genealogies, poems, songs and even the exploits of prominent people. In this way, they transferred information throughout their territory. Each string reflects millions of different combinations of color, fiber type and knot direction to house an enormous variety and quantity of knowledge. These sequences create a coding system whose mathematical organization anticipates the operations of a computer.

Throughout history, empires have relied on communication networks to circulate products, information and knowledge, manage resources and their complex social organizations. For the Inca empire, the quipu was what digital infrastructure is to contemporary global markets.

Today, as computers and data centers encode and process immense amounts of information – at the cost of consuming enormous amounts of energy, water and emitting large volumes of CO2 – the quipus remain virtually indecipherable. If we could read the stories stored in their knots, they would transform our understanding of the Inca empire and its social organization. In the absence of written records, their story has been told mainly by the Spanish colonizers, who imposed Hispanic writing over that system of strings. Later, around 1570, quipus would achieve colonial recognition as records and begin to be admitted in legal and administrative contexts. However, quipus also provided strategies for local resistance against colonial rule, subverting the Catholic imposition of confession or facilitating the planning of uprisings through messages unreadable to colonial officials. Similar episodes triggered what has been described as “the colonizer’s fear of the knotted rope,” which eventually led to the banning of the quipu by the Catholic Church in 1583, during the Third Session of the Limense Provincial Council. Not many survived. And despite their occasional transcription, these records were not preserved along with the quipus or their illustrations, making the relationships between the strings and their meaning illegible today.

Today’s digital infrastructure has been identified as the key to unlocking the secrets of quipus and protecting them from the passage of time. The more than 850 quipus digitized worldwide constitute a database that could be studied using trained computer algorithms and macro-analysis. From the study of these reproductions and simulations, patterns and similarities could emerge that anthropologists, archaeologists, historians, linguists and mathematicians could study. However, the attraction to their enigmatic nature and the desire to unravel these chronicles of the Andean past resorts to the same logics that put indigenous and local landscapes and communities at risk. Academic and scientific efforts to archive, digitize, analyze and decipher them are part of a historical lineage that connects to the European colonial order and its material, cultural and epistemological genocide. Transforming the quipus into databases implies subjugating, once again, indigenous knowledge and cosmovision. What was destroyed by colonization was not only the knowledge to encode and decode the quipus, but the social and cultural relations organized around them, leaving the surviving cords as empty signifiers and evidence of the violence of erasure deployed by colonialism.

Someday our data will be quipus

Over time, our digital files become undecipherable like quipus. If we were to leave a folder of photographs on a storage device for years, the computer’s perception of the code that allows it to read the information in the files would change, rendering the data unreadable. The corruption and disappearance of data demonstrates the interconnectedness between the digital and physical realms. Eventually, our millions of files will suffer the same fate as ecosystems affected by unsustainable storage practices.

From fear of loss, most market-driven scientific innovations for data storage hold the empty promise of preserving information unaltered by the passage of time. But they fail to recognize that data centers, like archives, are immersed in states of proliferation and decay. Memory is not what is contained and stored, but a practice that involves acts of remembering and forgetting.

The decomposition of data reminds us of the fragility of the digital and physical world, of our environments, bodies and histories. Rather than aspiring to a storage medium that allows for unlimited accumulation, I advocate for a practice that acknowledges our dependence on data and accompanies us in processes of detachment, loss, mourning for digital data, individual and collective.

Just as we mourn the loss of physical objects, places and people, letting go of data also requires a mourning process that recognizes its emotional and cultural significance. The obsession with backups fails to recognize that any archive is continually renewed through maintenance practices, that meanings are updated, and that remembering simultaneously concerns the generation of knowledge and its loss. It is through states of decomposition and recomposition of information that previously inconceivable future meanings emerge that surpass the intentions and interpretations of those who deposited and stored archives. Those like us.

If we fail to do so, errors and mutations will in any case alter them over time, leaving room for unexpected developments and emerging intelligences that transcend any apparent purpose for humans without rendering this knowledge irrelevant. As our bodies disappear and our data become seemingly meaningless detritus, our histories will be vaguely remembered by bacteria, viruses, and machines.

Memory Pacts

In her visit to Chile earlier this year, Marina Otero shared these reflections and concerns. That day she talked with architect Nicolás Díaz and different attendees at the In.FAIR meeting at the Lo Contador Campus of the Catholic University. Among colleagues, students and curious people, a conversation was opened about how to let go of data. Here are the main points that emerged.

1. Data

“Between 2012 and 2022, the National Library of Chile received one million two hundred thousand books occupying 15,600 shelves. In terms of digital information, 2020 as a society we produce 3.5 quintillion bytes per day, that is, a number with eighteen zeros. We would need 900,000 disks, four rooms full of disks to store what is produced in 24 hours,” says Nicolás Díaz.

Díaz comments that throughout history there have been museums, archives and libraries that have tried to preserve humanity’s information. These buildings have a clear purpose: the national library stores the heritage written by Chileans. However, when it comes to data centers, something different happens: it is never considered a public institution, for reasons of sovereignty.

It is interesting, says Marina Otero, that there is no clear definition of what a data center is. This allows us to rethink its spaces, relationships and functions. Despite their apparent banality, data centers are fundamental to current and future activities. Climate catastrophe forces us to reinvent how data is managed, produced, circulated and stored if the industry is to deliver on the promises of AI, the Internet of Things and the metaverse without causing planetary collapse. Curiously, just when we seem to have reached the limit of our shared world and are aware of the dangers of equating progress with infinite growth, we take refuge in the digital world where the promise of infinite possibilities is still present. It is a world that is imagined as a limitless space: the cloud, where everything is possible.

The production and storage of digital information is reaching an unprecedented volume. In his report “How Much Is Not Enough?”, Gratner refers to a Zone of Potential Insufficiency and estimates that the demand for data storage could skyrocket by as much as 50% between 2020 and 2030. This information overproduction is bound to create the possibility of alternative futures in science, health, city organization, computing, space exploration. But by outstripping current storage possibilities, its energy impact and endangers the only world we inhabit.

Governments, under pressure from local communities, environmental movements and rising energy prices, have begun to impose controls. Several data center hubs, including countries such as Singapore and the Netherlands, have implemented temporary bans on the construction of data centers due to their excessive consumption, which negatively affects nearby residents. These and other episodes demonstrate the importance of demanding new paradigms for data production, consumption and storage, eco-social models for digital infrastructures. Companies, universities and governments are collaborating to develop alternative models: heat and water reuse systems, underwater or space-based data centers; data storage systems in holograms, fluorescent dyes, crystals and DNA; macro-hubs giving way to decentralized computing and micro-hubs; quantum computing. Most of the new models, however, assume the inevitability of ever-increasing data production. And, in turn, they fuel the extraction of more resources, the laying of new undersea cables, the building of more data centers, the production of more batteries, often to the detriment of local communities around the world.

To devise other models we need criteria and consensus on how to classify and store data, for how long, for what or for whom. Is it really necessary for us to store so many videos of kittens? To me they seem like indispensable documents of our time, but for many other people they may certainly be dispensable. So who makes the final decision, and by what process? Museums, archives, heritage commissions, can offer us examples of processes in which decisions are made about what to keep and what not to keep.

In addition to certain rituals, I am particularly interested in establishing alternative institutional practices for data management, which instead of relying on large corporations, rely on self-organized collectivities. Let’s think of a community that manages its data at the neighborhood level, and that is connected to other similar communities. I would love to work on a pilot project that considers information as a good distributed throughout the city and that each one of us is responsible for taking care of it, for giving it space in our daily lives through ways of doing things and their respective social structures.

2. The limits

I am also in favor of working with lower resolutions, argues Marina Otero. Why not think of speed and resolution as political, aesthetic and ecological weapons? What if we consider working in another way?

3. Mourning

“How do we want to relate to this expanding identity, how do we speculate about other ways of relating to what is beyond the interface? -Martín Tironi, sociologist and director of Núcleo Fair, intervenes. And he adds: “The duel of data goes beyond the relationship with data. It also relates to certain infrastructures of capitalism, the notion of progress and modernization, but which are no longer coherent with our reality.”

In view of these reflections, Marina Otero argues, I believe that we should not be catastrophists or conformists, but we should bear in mind that there will be processes, territories and ways of life that will disappear. And we need to develop practices of memorization and forgetting, as well as rituals that help us to process this mourning. It is necessary to abandon habits, processes, products that do not respond to the ecosocial challenges we face. Making these decisions can be difficult, but it is necessary to recognize their strategic importance.

Grief is something very human, but what happens if we place it in another context, in other ecologies in which humans play a fundamental role? We can speculate that many of these servers and media that store information, such as artificial DNA, will continue to exist even beyond the presence of the human species on the planet. That detritus of information, an amalgam composed of organic and inorganic elements, of data that has already lost the meaning we gave it when we conceived it, will become something on which bacteria, viruses or artificial intelligences will build a new reality, another kind of knowledge. They will be, perhaps, the ones that will face the duel for us humans. Our memories will become a working material for those entities that will survive us. I think it’s nice to think of it that way: it makes us even smaller and renders our obsession with storage banal.

4. Future

Chile is one of the most interesting countries to analyze the future of digital industries. It is a strategic territory due to the number of cable connections with other territories, such as the Pacific, the United States and Australia. It is also important the perception of security and stability of the country from abroad, something fundamental for large companies that are considering establishing infrastructures here. However, having certain conditions and resources has never been a guarantee, and it is something that is well known in Latin America, to achieve total autonomy from foreign powers. This is also the case among regions in Europe. Europe has recently intensified its efforts to achieve energy independence. For this it is necessary to open mines of minerals and elements considered strategic, such as lithium. Lithium mines are being planned and approved, with much resistance from local communities, in Portugal, Spain, Serbia and the Czech Republic. These are the places of sacrifice, while it is other countries that concentrate the battery factories where the resource acquires added value. These extractivist and neocolonialist dynamics usually occur between the global south and north, one enduring the catastrophic ecosocial effects of extractivism and the other benefiting from it.

And that leads to the questions: who should mourn the data? Individuals? How much responsibility do we distribute for that excess? Are there excesses that are more harmful than others and have a more significant impact?

Ultimately, we are all harmed by industries and practices that are based on resource depletion. Only by being aware of these processes, and the implications of an unrealistic conception of digital infrastructure and resources as infinite, will we also be able to hold those with greater decision-making power accountable. And as long as we see ourselves as individuals and not as collectivities, it will be difficult to have a strong position vis-à-vis these powers.