Notification

×

Iklan ok

Artificial Intelligence (AI) and the Tyranny of the Data Commons

| Agustus 20, 2023 WIB



Opinion | In the late 20th and early 21st centuries, the concept of the sharing economy emerged, promising a grand vision of a digital commons where data would flow freely, leading to a new mode of production untethered from the exploitative mechanisms of capitalism. Enthusiasts believed that data, being non-rivalrous, would pave the way for a world where ownership was less relevant and where individuals could freely create and distribute goods. The optimism surrounding this idea was akin to the hope instilled by fantastical tales, much like "The Lord of the Rings."


However, this idealistic narrative of a data-driven utopia was, to put it bluntly, a deception. Unbeknownst to the data-generating masses, a new entity, akin to the insidious One Ring of Tolkien's universe, was being forged behind the scenes – artificial intelligence (AI). As AI technologies advanced, it became increasingly clear that the data we had generated on Big Tech platforms was being appropriated and repurposed to train AI models. Ironically, the very corporations responsible for this appropriation began expressing faux concern about the disruptive potential of AI, urging for regulation even as they amassed staggering profits.


The roots of this predicament can be traced back to the characterization of data as a non-rival resource, which underpinned the concept of the sharing economy. This term suggested that data could be used and consumed by many without diminishing its value, akin to a digital image of a cake that could be used on multiple websites without any loss in quality. However, this idealistic perspective obscured the labor and resources required for the creation and distribution of these digital goods.


The analogy of a digital cake illustrates this disparity. The digital image of a cake presupposes the existence of an actual cake or its artistic representation. The human effort, creativity, and resources invested in this process are often overlooked in the data economy. Beyond this, energy costs for data circulation and storage, environmental implications, and the often-exploitative human labor required for labeling images for AI model training are conveniently ignored. While data might seem non-rivalrous on the surface, the human endeavors behind it are inherently rivalrous goods.


This dichotomy has led to a dual standard in data valuation. Individual and corporate data are legally protected as private assets, while public data, often referred to as the data commons, has been declared free and unowned, ready to be harvested. This corporate accumulation of shared data is what the author Nick Couldry and the narrator term "data colonialism." The transformation of this appropriated data into AI training sets represents not a tragedy, but rather a tyranny of the commons – a sinister twist on the well-known concept.


Drawing on this narrative, it's important to distinguish between "tyranny" and the more commonly invoked "tragedy" when considering the fate of the commons. The latter is often associated with the overexploitation of shared resources, as depicted by ecologist Garrett Hardin's allegory of shepherds overgrazing a communal pasture. While Hardin's narrative suggests that privatization or state control are the only solutions, Nobel laureate Elinor Ostrom presented examples of effective community management of commons.


These notions of the power of commons greatly influenced the sharing economy's optimistic ethos, encouraging individuals to contribute their data under the premise of creating a harmonious digital ecosystem. However, the reality is starkly different. Behind the scenes, corporations have actively worked to privatize and monetize public data, culminating in the ongoing tyranny of data appropriation.


Platforms like Facebook, Twitter, and Reddit have demonstrated that the data generated by users is not under communal control but rather the property of profit-driven corporations. What was promised as a world of collective digital plenty has, in actuality, entailed the transformation of data into a tool used against users. This data-fueled AI is still in its nascent stage, yet its influence is already discernible, often to the detriment of society's most vulnerable.


While Big Tech corporations continue to advocate the idea of data as a non-rival good and tout their AI models as public goods, these actions must be seen as self-serving. Beneath the altruistic façade lies a strategy to evade legal repercussions, delay stringent regulations, and legitimize the privatization of the data commons – akin to how public spaces were created from lands dispossessed from Indigenous communities, serving as a cover-up for the initial act of dispossession.


To safeguard against a recurrence of this deceitful pattern, it is imperative to scrutinize the intentions and actions of these corporations. Vigilance is needed to ensure that the promise of a digital commons is not simply another smokescreen for data appropriation and profit accumulation. The story of the demise of the sharing economy is a cautionary tale that reminds us to question, resist, and advocate for accurate equity and transparency in the digital age.

×
Berita Terbaru Update