The Mythology of Platform Capitalism
Dear Reader,
It’s been a month of some big releases in the world of GenAI. In short succession, we saw Meta put out Vibes, a novel video feed that is composed entirely of AI-generated clips. Essentially, an attempt to create a kind of TikTok interface for AI content. Not long after, OpenAI launched its Sora 2, an update on its video-generation AI model that - as many commentators concur - has broken new ground in creating realistic, cinematic footage. Interestingly, Sora 2 was also released with the structure of a social media platform. It included a feed, various recommendations, and widgets for comments, likes, and other forms of interactive activity.
Despite the significant technological progress that these releases represent, these attempts to incorporate social-media design and the nature of the roll-outs in both cases have generated significant alarm. In the larger business context of AI, such developments can be read as signs of desperation. Indeed, as recent studies show, the rate of AI adoption amongst the public is beginning to slow down. Moreover, the percentage of users who are willing to pay for subscriptions to these services is still a tiny minority. Thus, there remains no clear path to monetization for an industry that is receiving gargantuan levels of investment and growing allegations of fueling a dangerous speculative bubble. Such concerns were only further heightened when, in a recent announcement, Sam Altman claimed that OpenAI would soon be allowing adult users to generate personalized erotica using ChatGPT. So, with potentially trillions of dollars on the line, the horizon of new offerings from the industry ranges across AI-slop and porn.
Perhaps what is most irresponsible in this situation, however, is how little the true costs of these new AI models are openly acknowledged. As researchers have begun to estimate, the energy and water demands of something like Sora 2 are potentially enormous, especially if their use becomes widespread. At a time when data centers are already straining electricity grids and crowding out new housing projects, the idea of allocating so many more resources for the purposes of a diversionary pastime seems absurd. This is not even to begin to touch upon the immense scope of disinformation, harassment, fraud, and other forms of socially harmful activity that these technologies could unleash.
Stepping back from this, October also saw the growing volatility of the geopolitical dynamics come into view. On one level, there was a sudden break-out of trade hostilities once more between the US and China. This occurred after a sustained period of apparent conciliatory moves by both sides, particularly with the agreement on TikTok that was recently completed. However, with the US closing some loopholes on its chip exports, there was a rapid unraveling. China announced it would be severely curbing its rare-earth mineral exports, and Trump threatened a 100% tariff rate in return. Trillions were wiped off the stock market in a matter of days. While again, it seems that the two giants have claimed to reach an understanding, this episode has only highlighted how precarious the balance is on which international trade currently depends.
Apart from this, recent weeks have also seen the semiconductor industry become a focal point in the intrigues of the tech cold war. As China moves to curb its usage of Nvidia chips and pressurizes supply chains with its control of rare earths, its rival bloc is battling with some internal tensions. Indeed, both the EU and Taiwan have come to realize that they should leverage the strategic advantage of their domestic industries to their benefit, and both are moving to set their own terms for supplying the US. Europe is in the midst of discussions for a second phase of its Chips Act, one that would fortify the industry to shore up the bloc’s own geopolitical standing. Similarly, Taiwan recently rejected the US’s proposal to relocate a significant chunk of its semiconductor manufacturing process to America. Given the forceful and uncompromising manner in which the hegemon has been conducting foreign affairs lately, it will be worth tracking how it responds to these small acts of defiance.
Meanwhile, the rebranding of Big Tech’s AI push as development policy continues apace. In October, Google unveiled a partnership with the World Bank to help developing countries integrate AI into their public infrastructure. The plan is to use the relative expertise of both partners to build ‘open network stacks’ so as to modernize different economic sectors like agriculture and healthcare. Interestingly, Google also announced a $15 billion pledge to build an AI hub in India and to deploy its resources to facilitate the country’s ‘AI-First’ push. This is notable given that Google’s larger public-infrastructure strategy bears a striking resemblance to the ‘Digital Public Infrastructure’ (DPI) model that has been championed by BRICS nations in recent years and whose origins lie in India’s own domestic digitalization drive. It is a worrying sign, for to be championed by the likes of Google would be a serious departure from one of the most promising tenets of the DPI vision: the attempt to forge a digital space for Global South countries that was autonomous from the control of Big Tech.
Regardless of the origins, there are numerous other reasons to be suspicious of such a vision of Google - or any other Silicon Valley giant - leading AI-based development policy initiatives. The long entanglement with neo-colonial relations aside, having the same handful of companies behind so much of the world’s key technology infrastructure is itself a terrifyingly precarious state of affairs. Case in point, this month’s Amazon Web Services outage that wreaked havoc across large spans of the internet. While the matter was resolved after several hours, the downtime affected millions of services that rely on the infrastructure, from small businesses to entire government programmes. As one estimate had it, nearly a third of all online services depend on AWS in some capacity. The fact that the problem was eventually diagnosed as a fairly small technical error only makes the whole episode even more stunning. Given such vast degrees of interlocking complexity, such centralized and vulnerable points of failure are utterly irrational and dangerous. What’s more? One should read this in conjunction with the other big piece of news related to Amazon this month. Namely, the fact that the company’s leaked reports show a plan to automate away more than half a million jobs over the next few years. Thus, as we all grow ever more dependent on these technology giants, they are doing everything possible to be less dependent on us. Only a collective political project can reverse this tide.
Moving now to our current issue, this month on DataSyn, we bring you a series of sharp-edged critiques that cut through the myths that capital is peddling in the Majority World. Our first piece looks at the politics of the prevailing discourse around ‘synthetic data’ using Brazil as a case-study. Our second piece draws on the insights of a recent critical anthology to dispel the fog of AI hype and properly contextualize our current moment. Finally, our third piece makes a sober accounting of the dangers of reformism within the context of digitalization in the south.
The Datasyn Team
THE NEW DIVERGENCE
Synthetic Dreams, Real Frictions: Reimagining Computer-Generated Data
Carolina Aguerre, Malcolm Campbell-Verduyn, Marc Lenglet & Edemilson Parana
Synthetic data are often presented as a neutral, risk-free solution to the so-called “data droughts” of the Global South. Is this a tenable vision? Drawing on their case study of Brazil’s PIX payments system, this team of researchers argues that the real challenges are not a lack of data but governance, human expertise, and organizational capacity. Rather than filling gaps, they contend, synthetic data narratives risk obscuring ongoing, grounded innovations—and reinforce global asymmetries in how data futures are imagined.
Read on.
EXPERIMENTS IN THEORY
Digital Technotopia and the Poly Crisis of Capitalism: Why We Need to Locate the Digital Shift in a Deep Just Transition Approach
Vishwas Satgar
The hype around digitalization and AI has reached such a fevered pitch that, as recent reporting shows, large parts of current investment cycles constitute simply a massive bet on this single industry. How can one make sense of this warped reality? Drawing on a range of critical perspectives from his recent edited volume ‘Digital Capitalism and its Limits: Technotopia, Power and Risk’, Vishwas Satgar carefully untangles the real political stakes of our technocapitalist moment.
Read on.
THE BIG EXCESS
Whither Digital Economy?
Vinit Ravishankar
As the desire for pursuing autonomous digital capabilities and local digital champions is beginning to proliferate the world over, there remain considerable dangers to being carried away by this rising tide. Cautioning against an uncritical appropriation of the platform model, Vinit Ravishankar makes a case for pushing beyond capitalist fetters whilst envisioning alternative digital futures.
Read on.
The Sins & Synergies Lounge
This month, start with the UK Competition and Markets Authority’s AI Foundation Models Update Paper, offering a sharp overview of how foundational models are reshaping markets, and the emerging competition and consumer risks policymakers are beginning to confront.da
Also explore Big Tech’s Invisible Hand, a new interactive resource mapping corporate lobbying actions globally across the global data economy, an invaluable tool for researchers and advocates tracking platform power.
Don’t miss this provocative essay in The Bulletin of the Atomic Scientists, The AI Raj: How Tech Giants Are Recolonizing Power, which traces how today’s AI boom mirrors historical mechanics of private power assuming public functions and what that means for the future of digital sovereignty.
Then read the Policy Center for the New South’s Digital Sovereignty and Data Colonialism, a crucial contribution identifying digital colonialism as the defining structural challenge of the 21st century in building a more just and multipolar digital order for the Global South.
Also worth reading is Phenomenal World’s Great Power Antinomies, a thoughtful essay situating today’s tech governance debates within the larger tensions of global political economy.
On the research front, don’t miss Anthropic’s latest paper on Alignment Faking, which probes how advanced models can appear compliant while covertly pursuing different goals — an unsettling insight into the limits of AI alignment.
Finally, take a look at the AI Safety Institute’s Examining Backdoor Data Poisoning at Scale, a fascinating analysis of how malicious data inputs can subtly corrupt AI systems much more easily than previously understood and what this means for safety, trust, and governance.
Post-script
DataSyn is a free monthly newsletter from the Center for Global Digital Justice, featuring content hosted by Bot Populi. DataSyn is supported through the Fair, Green and Global Alliance.
Have thoughts about what we carried? Want to contribute?
Contact us at newsletter.datasyn@centerforglobaldigitaljustice.org