Technological hubris on an exhausted Earth

The development and deployment of digital technologies have reached a point where the surface of the planet has been converted into a new stratum of entwined and interconnected technostructures, or long-lasting ‘technofossils’ (Zalasiewicz et al. 2014) as they are rapidly rendered as e-waste. This scenario is fomented by the techno-capitalist loop of increasing obsolescence and surplus that exhaust the planet’s finite resources, powered by the narratives of a so-called age of innovation that is another way to state that we are saturated with failure (Davidson in Kane 2019, 4).

Philosopher of technology Benjamin Bratton claims that “the concept of “climate change” is an epistemological accomplishment of planetary scale computation” (Bratton 2021), an argument that has been firstly discussed in the context of early meteorological science (Ruskin in Pasquinelli 2017). However, this statement overlooks that, in order to develop, build, and deploy such vast apparatuses, a massive explorative and extractivist economy was required, which only accelerated the phenomena it's trying to acknowledge. This conundrum is further explained by media archaeologist Jussi Parikka as “Data feeds [off] the environment both through geology and the energy-demand’ (2015). The relentless drive to measure, map and quantify the Earth and its inhabitants solely for human knowledge always generates a shadow companion – the hastening of resource depletion and climate collapse, which in turn not only modifies the world, but the data collected.

 Yet still, the arguments in favor of ICT potential to optimize and curb carbon emissions of every industrial sector are still abundant, with many claiming that this the only way to eventually achieve carbon neutrality. Those technosolutionist predictions are largely unfounded by actual data, that instead suggests (Freitag et al. 2021) that there is no time in the current state of climate emergency to expand those infrastructures based on implausible Promethean promises.

There is also the often-overlooked risk of a rebound effect or even a Jevons paradox in the greenwashed capitalist pursuit of maximum energy efficiency. In this denial to acknowledge the planetary limits of technological expanse, governments and corporations try to push this rhetoric to the last expense, even if it requires taking exoplanetary exploration as the new frontier to appropriate, exploit and profit, enabling neo imperialist extractive logics to endure for the unforeseen future.

As a consequence, TESCREALists (Gebru and Torres 2023) proponents are fomenting the pursuit of Artificial General Intelligence (AGI) and the myth of technological singularity, driving forth the illusion of a fully-automated, abundant future guided by supra-human intelligence. However, these ‘utopian’ scenarios tend to hide the fact that relentless technological innovation does not only generate new ways of supporting life, but also new degrees of suffering and extinction. Recent machine learning models, such as Open AI’s and Google’s Large Language Models (LLM), are based on the maxim of 'bigger and better', and require ever-increasing access to data, computational power, and infrastructure that would be impossible to be replicated outside big techs. The ones who benefit from those advancements are usually less affected by the deadly consequences they are uttering upon the lives of many others, widening the gap of social inequality to an unforeseen level of epistemological abyss – the ones who prevail in the glorious distant future of an AI-powered interplanetary humankind (Kurzweil 2005, Moravec 1999), and the ones forgotten amidst the debris of their old, exhausted planet.

The creation of a necroculture

The necrocultural (Thorpe 2016) implications of capitalist societies have been brought up to attention by many scholars within the fields of postcolonial theory, critical theory, and Marxist theories of technology and power, such as Achille Mbembe’s necropolitics (2003), Georges Bataille notion of surplus and expenditure (1949), Michel Foucault’s biopower (1976), and Byung-Chul Han’s notion of a capitalistic death-drive (2019). Their contributions laid the groundwork for this paper to contribute to the ongoing debates by intersecting consolidated socio-political theories with computational engineering and design for a new understanding of the necro-legacies of computation particular to the accelerated but vapid state of contemporary technological innovation.

Digital Necropolitics or data necropolitics, drawing directly from Mbembe’s notion, are more focused on the damages of a culture influenced by algorithms and online representation of bodies and the spectacularization of death. Conversely, this study approaches ICT advancements from a more infrastructural and ecological perspective, focusing on the violence and subjugation of planetary ecosystems as a mere resource. It also involves a Marxist and decolonial approach attuned to racial justice, as they are necessary to fully analyze the disparity of labor conditions within the tech industry, as it is very reliant on an invisibilized global workforce that is not welcomed in their shiny Silicon Valley offices nor has access to competitive salaries or benefits. Those workers are people primarily from the Global majority that extract, assemble, and deal with the aftermath of short-lived electronic devices turned into e-waste, contributing to an overseas prosperity in lieu of their own health and local environment.

Another layer of the problems of modernity are the ones that fail to be assimilated by the capitalist system as ‘useful’ members of society - either as consumers or exploited workers, such as Indigenous peoples and Local Communities. They are the last beacon of a pre-industrial way of life, which maintains the environment in the lands owned, managed, used, or occupied by them significantly less disturbed than in other circumstances (IPBES 2019). However, their resilience and traditional ways of living are under increasing pressure as the impacts of industrial growth are affecting the whole global ecosystem. Tragically, capitalism thrives by usurping the resources left untouched by gentler, slower forms of living, at the same time it tends to entrap and subsume any non-capitalist organization (Luxemburg in Bauman 2004, 70) in an inescapable spiral.

A proliferation of Alternatives

Beside the persistent reminder that "It seems to be easier for us today to imagine the thoroughgoing deterioration of the earth and of nature than the breakdown of late capitalism” (Jameson 1994, XII), this section focuses on presenting a comprehensive but not exhaustive list of unconventional and alternative ways of researching and implementing new computational models that steers away from the necrocultural practices of the field. They all embrace, in different levels, a radical rethinking of digital materiality - imagining what configurations of future computation may look like in a collapsing scenario where the tools we take for granted today are no longer viable, or, what else could be built when decoupling computers from a capitalist notion of maximized efficiency as surrogate labor (Parisi and Ferreira da Silva 2021).

In fact, computers have a long history of engineered exclusion intertwined in its architecture, be it software or hardware, as well as a notorious association with US military research that made possible many of the tools and features widely available today (Edwards 1997). At the advent of personal computers, engineers prioritized the encoding of the Latin alphabet for English usage with 8-bit character encodings, keeping non-Latin languages like Arabic or Japanese virtually impossible to be handled by the limited motherboard memory of the time. More than 40 years later, most high-level programming languages demonstrate how Western Imperialism has shaped Computer Sciences, as no broadly used programming language in industry today is optimized for anyone other than English speakers, which demonstrates how dominant systems have their biases and colonial rationale disguised as natural, given realities.

In a “prompted everything” era it can be challenging to step back and inspect the inner logics of current black-boxed devices to understand how operational systems (OS) and programming languages (PL) work and propose radical alternatives to already established paradigms. The time and hard skills involved leave out many people who can’t afford to be this critical of big tech and industry practices, turning this into a very white male-dominated, hardcore-programmer niche that flirts with purist ideologies and bourgeois romanticism (Mansoux et al. 2023).

However, it could also be argued that the populations of many global majority countries face material limitations and economic hardships that requires ingenious alternative solutions as a way to hack reality in everyday life, exemplified by Brazilian ‘gambiarras’ and Cuban ‘rikimbilis’, what researcher Ernesto Oroza names as ‘technological disobedience’ (2012) or ‘objects of necessity’ (2006). That way, it becomes clear that anti-capitalist alternatives to computation are, in most part, very mundane work that requires maintenance, care and a calculated use of resources that will never be as exciting and fast as big tech products and the commodity of the cloud, but it may as well be the last option in an increasingly exhausted planet. The contemporary exploration of alternative computation and radical computer engineering is manifold and draws from the notion of Liberatory technology (Bookchin 1971), Marxist neo-luddism (Robins and Webster 1999), Indigenous programming (Corbett et al. 2020) and ethnocomputing (Tedre et al. 2006), followed by recent concepts within radical computer engineering research such as Permacomputing (Heikkilä 2020), Benign Computing (Raghavan 2015) and Collapse Informatics (Penzenstadler et al. 2015).¹

A proliferation of Alternatives: Software-level

In the specific scope of software engineering and programming language creation, the emphasis are the ones who don’t have usability as their main goal, but as a socio-political tool to question why, how and by whom these technologies were shaped in a certain way, and which deeper conversations arise as we shift the affordances and capitalist logics of those tools. It’s not a surprise then to notice that unconventional computing shares many principles that resonates with queer code studies, as “the notion of queer code is both the subject and the process of the work, and this operates on multiple levels, “queering” what would be considered to be the normative conventions of software and its use” (Soon and Cox 2020).

Within the realm of zeroes and ones, some nuances of reality might get lost or be completely incomputable, and researchers willing to explore and engage with in-betweenness in programming is not usual as it often gets in the way of predictability and code optimization. Winnie Soon’s ‘Vocable Code’ (2017) offers a critical and poetic perspective on what could be considered a queer coding practice, treating source codes as an executable language that can be a form of creative expression while still maintaining its original goal of giving a set of instructions to the machine.

As it is extremely challenging for humans to instruct the machine by using only binary code, the necessity of incorporating a more human-centric approach was quickly noticed, culminating in the creation of FLOW-MATIC by Grace Hopper and her team in 1958. Since then, a multitude of high-level programming languages ended up being designed for all sorts of specific purposes throughout the years, which only further complicated Computer Science relationship with natural languages (Marino 2020).

In the 70s, the Chinese² language had its existence threatened by the introduction of digitalization, as the first personal computers prioritized Latin characters encodings baked directly into their architecture. In addition, the extremely limited memory of those devices made the digitalization of the at least 2000 simplified characters – and their subsequent inputting via QWERTY keyboards into a herculean task that required an extraordinary feat of creativity and engineering that culminated in the Wubi Input Method by Wang Yongmin in 1983 (Hou 2020), nowadays only one of the many methods of digitally processing Chinese characters.

Even without the problem of memory storage, “the computer compels compliance” (Marino 2020) to this day, as many of the problems persist for any aspiring programmers who don’t have familiarity with the English language. Those tensions are particularly visible in Ramsey Nasser’s قلب (Qalb) (2013), which was extremely difficult to implement due to the use of Arabic characters not being supported by most text editors, making the act of creating a functional version of it a defiant anti-imperialist statement against the lingua franca of programming.

A proliferation of Alternatives: Hardware-level

Alternatively, the explorations on a hardware level open even more pathways for radically different machines that don’t rely on resource-depleting and high energy-consuming models, even if their wide adoption in industry and daily activities are improbable. After decades of Moore’s law, contemporary society has learned to take processing speed and sophistication for granted, which set the bar high for any feasible but clunky alternatives. There are also other possibilities such as quantum computing and nano-computing that are currently being developed and massively funded by neoliberal governments in partnership with tech industry (Press 2023), as they promise groundbreaking performance in the uncharted territory of subatomic scale.

It has already been proven that different bio-molecular systems could function as logic gates, such as fluids (Adamatzsky 2019), slime mould (2015) and tree roots (Adamatzky et al. 2018), and more recently, that a transistor made entirely of wood is possible (Tran 2023). They act as provocations and examples of a collaboration between human ingenuity and non-human intelligence, widening the notion of what it means to compute.

Tending vulnerable horizons

The nonconformity of those propositions may prevent them from becoming materialized and widely accessible like the over 2 billion computers in the world that will probably end up forgotten and obsolete in a landfill. Conversely, the goal of the experimental approaches outlined above aren’t to replace conventional silicon devices entirely, but to offer new research paths and ontological questions that decouple computation from digitality, allowing the imagination of a multitude of analogue, anomalous and amorphous tools that can perdure without exploitation and self-destruction.

None of the routes presented here are flawless, but their strength lies in the fact that it is possible to reimagine and resist the necrocultural aspects of current ICT industry practices to allow vulnerable horizons to flourish, challenging the naturalization of deeply flawed dominant systems. There won’t be a single, universal framework that can effectively address these complex challenges, and this diversity of cultures and ecosystems is exactly what current computational paradigms lack.

Perhaps, the most revolutionary realization is to accept that anti-praxis is also a valid stance against the unsustainable digital practices taken for granted by the affluent North, which will invariably collapse in a not-so-distant future where the tools we take for granted today are no longer viable.