Building off the work of computer scientists, social scientists, and legal scholars, I use the term Big Computing to denote the entire ensemble of technical processes, organizational changes, and institutional strategies that threaten the governing power of democratic institutions. Big Computing entails more than “the cloud,” though its ascendance coincides and thrives from a growing ecosystem that depends on massive computing power. Above all, Big Computing is a political positioning of tech firms that leave states—and ultimately users—in a bind.
The technical changes making resistance more difficult are important to unpack. According to Kostova, Gürses, and Troncoso, in the last twenty years the software industry has birthed “software as services.” This transitions software production from a sequentially order manner of production to a dynamically updated model of software development. Whereas the sequential model—referred to as the waterfall method in industry circles—equates with a linear process of designing, coding, testing, implementing, and going to market the dynamic method—referred to as the agile model—dispenses with that linear order. Virtual or cloud storage makes it easier to chunk out steps and improve these modularized parts at any time during the production cycle.
The “agile revolution” extends beyond software services into what the industry refers to “infrastructure as service” and “platform as service.” Both imply a move away from a traditional, more static, if not linear technological environment. Infrastructure as a service replaces an in-house web server with virtual hardware. The platform as service replaces back-end management like back-up and recovery for developers trying to make applications. In all three services—software, infrastructure, and platform—computing capacity is key to flexibility.
That dynamism has implications for configurations of power, not only within tech firms, but outside of firms too. From the perspective of software as service, and inside the firm, certain kinds of work within the software production process risks being diminished. For example, privacy engineers could previously count on tweaking a product before it went to market, whereas now, the perpetual updating of software weakens the ability of well-intentioned engineers to work towards an end goal of holistic protection for the user.
More generally, dynamism means modular teams intimately know and work within their own workstream. A lack of linearity in the development model means being knowledgeable about whole—the end-product—is challenging. In terms of resistance, the nature of modularized work complicates how workers see themselves in relation to choices made and directions taken by the executive class, as well as how they seem themselves in relation to other, more junior employees. That can have implications for how pushback or protest unfold or are sustained.
Across the industry, the agile model privileges those firms which furnish the computational fundament that make dynamically updating software services possible. Whether it’s a development platform, virtualized servers, or operating systems, these firms can shift what happens downstream and choose winners and shape winners and losers in the overall ecosystem. That means the dominance of computational service providers like Microsoft, Alphabet, Apple, and Amazon is all but guaranteed.
In addition to “agile” innovations, there have been so-called innovations in financial strategies that make it difficult for states to reign in Big Computing and for users to resist tech power. For more than a decade, the tech industry has followed a path of financialization. That’s essentially a code name for making money from money. In the case of technology companies, financialization means tech companies accrue wealth by investing in financial markets, not just by monetizing personal data.
Changes in financial strategies give computational providers outsize power, both politically and economically. Technology companies have prioritized working around corporate tax, keeping cash reserves low, writing off debt, or writing off charitable contributions, all of which contribute to disinvestment in public infrastructure. Recently, for example, Microsoft, Amazon, and Facebook have delved in housing finances, doling out loans to state agencies in the United States to finance the development of affordable housing. In essence, tech companies keep their court with shareholders, not citizens or consumers. When viewed alongside these companies’ immense lobbying power—for example, Amazon now outspends Big Oil and Big Tobacco in Washington, DC, Big Computing bodes poorly for those individuals and groups seeking tech accountability.
It's unclear whether an appeal to individual or collective rights will go far in this era of Big Computing. The agile model of production has also complicated consumption models. End users are no longer mere purchasers of a software product. Reminiscent of the discourse on prosumers, user generated content, and platforms, consumers of software services are also co-producers, or perhaps more accurately, co-engineers whose product usage is reported back to the tech firm and factors into tweaks, improvements, or updates. For users concerned about extraction, surveillance, or privacy, user rights can matter when our behavior becomes part of the process of technological development.
If bureaucracy was the organizational-institutional innovation that made modern day liberal democracy, the computational ecosystem is the current-day equivalent. It is making whatever societal system comes after liberal democracy. At the moment, there is little clarity or movement by individuals and groups to recalibrate configurations of power. This may change in the future with mass efforts to disengage from software development feedback loops, but also with regulations that treat tech not as technology, but as money—the medium that binds us all and that as such is too important to leave industry to its own devices.