#i Beyond Citations | Do AI export control rules help the US?
Missing the curated notes in the usual Technopolitik edition? What used to go out for last three months as a curated note in Technopolitik has migrated to this new Technopolitik section (akin to a sub-newsletter) called Beyond Citations.
Curated by Lokendra Sharma, Beyond Citations grounds current tech developments in foundational scholarship. It essentially dwells on deeper issues and answers that academic outputs — journal articles, book chapters, books, etc — have to offer about tech developments.
This inaugural edition makes the argument that AI export controls do not help India but they do not make the US any better.
Export controls are back in fashion three decades after the Cold War ended. And yet again, India finds itself on the other side of the table.
One of the defining features of the Cold War was that the superpowers were trying to stem horizontal nuclear proliferation while permitting vertical expansion as they amassed thousands of nuclear weapons. Any deviatory behaviour by a developing state wishing to acquire a nuclear weapon was portrayed as a threat to humankind like none other. Even among nuclear-have-nots, India’s nuclear ambitions were particularly ridiculed — every effort was made by the nuclear-haves to stymie India’s nuclear programme. Thus when India conducted the so-called Peaceful Nuclear Explosion (‘Smiling Buddha’) in 1974, the nuclear-haves got together and established the Nuclear Suppliers Group (NSG) in the 1970s. The NSG was a classic multilateral export control regime (MECR) designed to tightly control global nuclear tech and material flows. Other MECRs established during the Cold War included Australia Group (AG) in 1985 (to tackle chemical and biological weapons) and Missile Technology Control Regime (MTCR) in 1987. While the Wassenaar Arrangement (WA) came about in 1996, it had its roots in the Coordinating Committee for Multilateral Export Control which was established in 1949.
It took decades for India, which was left out of these exclusive clubs, to either become part of them (MTCR, AG, WA) or to live in peace with them (NSG).
But this peace ended when the US recently announced the AI export control rules. Formally called ‘Framework for Artificial Intelligence Diffusion’ they are anything but for diffusion. The rules aim to restrict the flow of AI tech and cutting-edge chips globally so as to maintain the US lead. The US and 18 close allies get unfettered access in tier 1, while China and Russia face the might of US export controls in tier 3. Majority of the countries are in tier 2 that gets some access to AI tech and chips but with a lot of strings attached. In a move that undoes some of the lingering goodwill from the US role in getting India a NSG waiver in 2008, India has been placed in tier 2. Bringing India back to the have-nots side of export control certainly does not help either Indian interests or the US-India tech relationship.
But does it help the US after all? If this week’s headlines are to be believed, then it does not. On 27 January, the AI chipmaker Nvidia lost about USD 600 billion in what is being called the ‘the biggest drop for any company on a single day in U.S. history.’ But why? Because a Chinese company DeepSeek has built an open-source large language model in just two months with just USD 6 million. To further make matters worse, they did it without cutting-edge chips. DeepSeek’s success demonstrated three things: first, you do not need billions of dollars to build a competitive AI model; second, cutting-edge chips are great, but one can work without them; third, open-source is a viable way to upend the proprietary dominant market. This has raised all kinds of uncomfortable questions about the inflated valuation (and prowess) of US AI companies as well as the utility of the newly announced export control rules. What good would export control rules do if a tier 3 state is ahead of you?
Market fluctuations aside, do export control really help the US when seen from a historical point of view? Many believe we are currently going through a Fourth Industrial Revolution defined by advances in AI among other cutting-edge technologies. Industrial revolutions in the past have been associated with power transitions. Would China’s lead in AI lead to another?
Jeffrey Ding has some answers for us in his following seminal 2024 book:
Ding, J. (2024). Technology and the Rise of Great Powers: How Diffusion Shapes Economic Competition. Princeton University Press. https://press.princeton.edu/books/paperback/9780691260341/technology-and-the-rise-of-great-powers
Ding studies in great detail what causes the rise and fall of great powers. More specifically, he investigates what leads to a great power transition during Industrial Revolutions. His argument, backed by a mixed-methods approach, is that it is the diffusion of general-purpose technologies (GPTs) and not cutting-edge innovation in leading sectors that matters in Industrial Revolutions. Thus, during the First Industrial Revolution, Britain’s ascent as the globally dominant power was cemented by diffusion of mechanical skills across its industries. In the Second Industrial Revolution, ‘the extension of interchangeable manufacturing techniques across many American industries functioned as the key GPT trajectory that fueled the rise of the United States.’ In the Third Industrial Revolution (the ‘information’ revolution), the US successfully averted a great power transition threatened by Japan’s innovations in consumer electronics. How? Partly by fostering computerisation across the US economy.
In the Fourth Industrial Revolution, Ding identifies AI as the leading GPT candidate that could play a role in US-China great power transition. If Ding’s thesis is to be believed, then the US should invest to expand AI skills base and diffuse AI to different sectors of the economy, including in small and medium-sized enterprises. The AI export control rules brought about the US focus too much on cutting-edge innovation — which, according to Ding, has historically neither helped a great power maintain its status or a challenger looking to upend the global order.