Toward a Minor Tech:RRA 500

But does it scale?

Roel Roscam Abbing

This is a terrible question common in technical circles to judge the merit of proposals and projects: can your idea expand in size to be relevant to many and, therefore, relevant at all? It is also often used as a way to put down alternative proposals, based on the implication that these proposals won’t scale and are therefore not worth pursuing further. Simultaneously, scalability is one of Silicon Valley’s core concerns as it enables the massive profits of social platforms.

Initially, I found myself avoiding the question of scalability, but due to recent developments I find myself compelled to consider it sincerely. Alternative digital infrastructures can engender different social relations than those of the scaled social platforms. However, if we are to build other systems that “mirror the world we want to see” and build actual prefigurative counter-powers (Keyes et al.) to platform capitalism, these alternatives, in one way or another, will need to operate at scale.

The negative externalities of scaled social platforms are becoming ever more evident, leading to an interest for non-scalability or other undoings of scale. This is expressed in the grassroots of computational culture (de Valk), as well as within human-computer interaction research literature (Larsen-Ledet et al.; Lampinen et al.). Over scalability, this literature suggests other metaphors such as proliferation as a way to consider the impact of a project.

The concerns against scalability are manifold. Anna Tsing demonstrates how scalability is a system's property “to expand without changing the nature of what it does”(Tsing, 2012, p. 8) and, as such, is fundamental to extractive capitalism. Consequently, scalability has the effect of erasing difference and local diversity, leaving ruins in its wake.

In response to Elon Musk’s purchase of Twitter in 2022, millions looked to Mastodon. This social network differentiates from Twitter in that it is a part of a network of thousands of smaller and interconnected sites known as the Fediverse, itself not run by any single entity. In the months after the purchase, this has proven to be a scalable system, but one that scales differently. Thousands of new and self-sovereign social networks were set up and through federation to become a part of a larger network. Thus, rather than scaling a single platform vertically, the process saw a network of networks scaling horizontally (Zulli et al.).

As someone who co-administers one of those small social networks, the months during Musk’s takeover made the necessity of scalability as a design property of software acutely felt. Our little space had to grow substantially within a short period of time. Not for growth or profit, but to be able to accommodate friends in need.

Through a different scalability, but scalability, nonetheless, millions managed to explore an alternative to the platform model by joining and trying, if only briefly, another model. Had the software and the model not been scalable at that moment of urgency, it would have been dismissed straight away. Instead, through scalability, the ideas and the model started to proliferate beyond the originary technical communities, after almost two decades of being around but being dismissed. Now that the terrible question is answered, we can start collectively posing more interesting ones.