The current global supply chains of microchips largely overlap with the plantation and production sites of cocaine in the twentieth century. This is neither conjecture, nor coincidental. The places where much of the production of microchips currently takes place - the United States, the Netherlands, Taiwan, and Japan – are the same places that were involved in monopolising worldwide cocaine production for medicinal use by the late 19th century.
Monocrops and Digitalisms
The deep entanglement between microchips and cocaine all starts from a seedling. Before cocaine was a cash crop, it was one of the world’s oldest cultivated harvests. A subtropical shrub mostly found in South America, its leaves have been consumed for millennia for religious, social, and medicinal purposes, most notably against altitude sickness. Throughout the 1800s, as the result of the colonisation of Latin America and globalisation, coca leaves started to make their way into commercial products. A brand of Bordeaux wine, and the drink now known as Coca-Cola experimented by adding it to their recipes.[1] By the 1850s, the leaf trade is booming. German company Merck imports it as an anesthetic. Freud pens ‘On Coca’ (1884). In the same year, Peruvian pharmacist Alfredo Bignon discovers a process that extracts cocaine paste from the leaves. European colonial powers, the Netherlands, France, and Great Britain sent seedlings around the world, from Trinidad to Tanzania, to Sri Lanka (then Ceylon) and and India, where it fares better. The crop was to compete with tea and sugar, established colonial money makers. United States pharmacists champion the drug as a legal ‘wonder cure’. Rampant, unregulated commercialization followed. By 1910, three distinct cocaine supply chains started to emerge: the United States, the Netherlands, and Japan, each with their imperial extensions – Peru, Java, and Taiwan, respectively.
The United States had the largest role in the industrialisation and commercialisation of cocaine, importing coca leaf, extracting cocaine, and selling it in forms ranging from injectable solutions to coca cigarettes and coca-flavoured beverages.[1] This position is soon overtaken, however, by The Netherlands, whose effort in the cocaine business, stem from coca with high alkaloid content grown on plantations in Java. This approach proves successful only after the founding of 'de Nederlandsche Cocaine Fabriek' (the Netherlands Cocaine Factory - NCF) in 1900. The NCF produced a majority of the world's legal supply of cocaine until to 1960[2]. World War I rapidly increases the demand for cocaine’s anesthetic functions and made the Netherlands one of the major cocaine producers in Europe, alongside early adopter Merck and Boehringer Mannheim.[3] As a result, the Dutch have been called the 'drug lords of the interbellum', playing a prominent role in the global narcotics industry through World War II, after which the whole industry was regulated out of legality. The infrastructure of cocaine as a medical product remains scantily visible in this history — the first site of the NCF occupied a canal-side location in the middle of Amsterdam, which currently houses a general practice doctor (which happens to be my family doctor). Medicine may have changed, but urban designations remain.[4]
Contemporary supply chains for the production of microchips mimics the geographical footprint of late-nineteenth and early-twentieth century cocaine production. Many of the world’s microchip machines are produced by machines created by ASML in Eindhoven, who by their own estimate make as much as 90% of the world’s photolithography machines, onto which microchips are printed.[9] While ASML is an extensive logistics company, sourcing tens of thousands of parts from third parties internationally for each of their machines, and ASML machines are sold globally, ASML’s arguably most notable trade relationship is with the Hsinchu foundries in Taiwan, a company known in English as a 4-letter acronym that bears similarities to Taiwan’s cocaine past: TSMC. TSMC’s foundries physically produce 60% of the world’s microchip output and 90% of all the world’s most advanced microchips.[10] Japan is another leader in the tech field: Canon and Nikon are two companies that still, like ASML, make photolitography machines, accounting for some of the remaining market share. Japan also produces a range of products needed in the microchip supply chain, such as chemical mixes, like ‘photoresist’, and silicon wafers.
The United States, for its part, has taken charge of much of the capital, technology, policy, and demand that keeps this global supply chain intact. All microchips produced globally fall under the United States’ extraterritorial governance through a rule called the ‘The Foreign Direct Product Rule’ (FDPR). FDPR allows the US to control foreign-made products if those products are the ‘direct product’ of US-origin technology or software — which applies to every single microchip manufactured today, since the manufacturing process relies heavily on specialised machine parts and automated design software that originates from the United States, even if no part of its production takes place on its soil.[11]. These restrictions are now actively reshaping the postcolonial landscapes across which microchips are now produced.
Furthermore, most US economic growth has become dependent on AI, with AI-related investments accounting for up to 90% of total U.S. GDP growth at points of 2025.[12] The wholesale expansion of AI is dependent on the steadily increasing capacities of microchips, and their continued production. These smart stones account for the vast majority of semiconductor income and profits.[13] During a gold rush, the only person turning a profit is the shovel salesman; and so, at this point, microchip production companies like ASML are some of the only companies actually generating a return on investment on the AI boom beyond the realm of speculation or projection.[14] The oversupply of generative AI to consumers has led to a global geopolitical addiction to the stuff. And AI, infrastructurally, is railing microchips.
TCDC Coca Plantation, undated. From A Promising Tropical Medicinal Plant: Taiwan as the Production Hub of Japan’s Coca Empire
The Semiconductor Plantacionocene
The supply chain of microchip manufacturing is founded on the infrastructure of the plantation. Planetary change driven by humankind unevenly extracts human labor and natural resources for the continuation of capitalism - a concept that has been coined the ‘plantacionocene’[15]. There is a historical continuity between colonial plantation economies and tech manufacturing, and Taiwan sits at the centre of this entanglement.
The history of Taiwan is the history of colonial plantation. The plantation industry got its start during the Dutch colonisation of the island by the VOC, and reached its peak during the Japanese regime. Throughout the centuries, Taiwan was once a major provider of the world’s tea, sugar[16], camphor[17], and bananas[18]. Despite diversifying its outputs from agriculture to technology, the production of tropical fruits, vegetable and orchid cultivars still make Taiwan a world leader in agricultural innovation.[19] In Taiwan, the Japanese colonial regime (1895 - 1945) established an island-wide plantation infrastructure and a labour force, in the second decade of its occupation. Japan sought to rival other empires by transplanting tropical industries to their colonial sphere. Except crops that were already long-established such as tea and sugar, cocaine proved to be successful enough to turn a profit, albeit for a short time.
Japanese’ efforts in the cocaine trade learned from the limitations of opium monopolies. Global (and particularly British) opium revenues started to decline in the 1920s, as several states restricted its trade. As a result, states were reluctant to monopolise coca, instead licensing specific companies as the sole operators authorised to undertake cocaine production. Japan’s Governor-General’s Office appointed TCDC and Hoshi Pharmaceutical as the only companies allowed to produce coca leaves in the Japanese empire, of which the former established both plantations and factories in colonial Taiwan, and the latter shipped leaves to Tokyo for further production.[20]
These Japanese cocaine businesses in Taiwan adapted capital, equipment, and managerial resources from preexisting plantations to suit their new medical purpose. Techniques for harvesting and drying the coca leaves were adapted from Taiwan’s long-established tea industry: tea driers were used for drying coca leaves, and pruning techniques were copied from literature on tea trees. Equipment from sugar factories was adapted for the processing of cocoa leaves: the Ensuiko Sugar Company factory in Xinying Village processed all of TCDC’s coca leaves into powder before building a designated factory for the process on the other side of the village. ‘Extraction vessels, replenishment vessels, stirred reactors, condensers, stills, boilers, compressors, and engines’[21] are readily available in sugar factories and necessary for the production of cocaine. Workers could work in both industries in different seasons of the year: the harvest seasons of tea and coca do not overlap.
Towards the end of World War 2, Taiwan produced some 80% of the world’s cocaine.[22] As the Republic of China took over Taiwan in 1945, the assets of Hoshi Pharmaceutical a were liquidated by the Taiwan Provincial Government, while TCDC fell into the hands of the Taiwan Medical Supplies Company. Arguments between the Ministry of Interior intending to demolish the industry, with the Ministry of Health arguing for its retention, seem to have been cut short by two large bush fires in the period 1946-1947.[23]
Despite Japan’s cocaine business meeting its ashy end, the plantationocene in Taiwan persists in different forms, and with different outputs. The 1960s saw the development of electronics factories in Taiwan, moved there as outsourcing efforts from Western companies intensified. Land reforms opened up labour as a resource for manufacturing.[24] The Taiwanese government, like many developing countries, offered cheap labor, tax incentives and a physical environment to welcome Western tech industries to offset their manufacturing costs. It set up one of the world’s first ‘export processing zones’. At this time, Taiwan was no longer a colony, yet had no clear path towards national sovereignty. Electronics industries on the island throughout the 1960s-80s were focused on assembly, not manufacturing: chips made abroad were tested and packaged. Taiwan faced stiff, zero-sum game competition from Hong Kong, Singapore, and Malaysia, in an outsourcing race to the bottom of the global wage table. As a way to differentiate themselves on the market, small tech entrepreneurs in Taiwan then embraced another path: Original Equipment Manufacturing (OEM). Within this model, small and medium-scale enterprises produce specific electronics parts for final assembly.
The decision to embrace the OEM path was a strategic choice made by the Taiwanese government. As South Korea and Japan invested strongly into research & development, Taiwan lacked access to the technologies and family-run conglomerates that would secure profit margins akin to those in the West. What Taiwan excelled in, however, was a measurement standardised during the plantation era: yield. Taiwan’s geography lends itself to small plantations necessitating hyper-intensive cultivation methods. [25] On cocaine plantations, this was measured as the cocaine extracted from dried leaves, a tiny percentage of the crops grown on the plantations. In the OEM industries, where yield refers to the percentage of usable, non-defective products created from total inputs. Historian Honghong Tinn argues that Taiwanese electronics factory female workers and male engineers’ tinkering skills enabled the successful miniaturisation and mass production of electronics.[26] The high yield produced by Taiwanese OEM factories made the island an attractive place to set up shop.
In 1968, an executive from Texas Instruments visited the island to help set up one such of the company’s factory. Although it was his first visit to Taiwan, he is now perhaps the person most associated with it: Morris Chang.[27] Some twenty years later, Morris Chang established his own OEM in Taiwan, known in the West as TSMC (Taiwan Semiconductor Manufacturing Company). TSMC established a novelty model in the electronics manufacturing business: the pure-play foundry model. In a pure-play semiconductor foundry, microchips are manufactured for external clients, without selling competing products.
The Taiwanese government believed that this invention could lift Taiwanese chip manufacturing out of its role as a site for low-profit margin assembly plants. In 1987, they offered Chang 48 percent of startup capital for TSMC on the condition that Chang could find a foreign chipmaker to provide advanced technology. Three years earlier, Philips had founded chipmaker ASML. The relationship between TSMC and ASML remains the most important relationship in microchip manufacturing today.
The foundry model allows for further specialisation in microchip manufacturing: with access to a good foundry, chip designers can go ‘fabless’, outsourcing the actual making of the chips to a single plant. This revolutionised the industry – among fabless producers today are Apple, Nvidia, Qualcomm, and AMD – and led to a proliferation in the types of applications of microchips. It also relentlessly consolidated production in a way that became increasingly capital-intensive to keep up with the latest developments.
With TSMC, Taiwan returned to the known model of government-sanctioned monopolies. It was a position familiar from the cocaine trade, when Hoshi and TCDC were exclusively designated to grow leaves on the island. Semiconductor dominance is therefore not a miracle of modernity, but the latest iteration of a much longer history of colonial craft labour and infrastructure. In Taiwan, agricultural lands and former plantations were re-zoned into science parks.[28] The demand for their products still comes from the West, the technology from the United States, the machines from the Netherlands, and much of the machine parts and chemical mixes from Japan. The plantation owner may have changed name and moved domicile, but the supply chains are intact. Lines of coke became lines of code.
Deskilling, Automating, or Sloppifying Labour
The reasons that microchip production mirrors that of the previous cocaine trade are not just geographical; they are also a matter of shifts in the way capitalism encloses labour. Computing was developed as a historical response to the abolition of slave labour in plantations. Meredith Whittaker traces the machinery and labour practices that were used on plantations directly to their successors in the form of programmable computers[29]. Charles Babbage may be primarily remembered as the inventor of the computer through the Difference Engine and the Analytical Engine, but he spent a more considerable effort during his life as a labour theorist.
Charles Babbage, 1840, The general plan of Mr. Babbage's Great Calculating Engine. Arrangement and rack, details, etc., Science Museum Group Collection, London.
In 1830, Babbage published On the Economy of Machinery and Manufactures, a book on the disciplining of labour. Whittaker writes that both Babbage’s theories as well as his experiments in conceptualising engines can be read as working towards a singular objective: in Whittaker’s words, “the discipline of purportedly “free” labor such that it could continue to produce for the British empire.” His anxieties are motivated by maintaining Britain’s dominant economic position on the world stage as the prospect of slavery abolition stood to refashion the world economy. The technologies for control that Babbage proposed in the Economy of Machinery, such as surveillance and rationalization, all have roots on the plantation. However, Babbage made a particular intervention in labour theory: the “Babbage principle[30]” argues that dividing a complex task into simpler component parts – which he designated as “low skilled” – justifies paying the people who perform each part less. The automation promised by the engines that became our computers must therefore be seen originally as a labour disciplining measure, and not a technological innovation that brought about efficiency for its own sake. The classification of ‘free’ labour through the mobilisation of technologies, such as surveillance of work time and the rationalization of yield per worker, was built on the unfree labour of enslavement and indentured plantation labour. This reclassification enabled a new era of industrialisation.
Taiwan’s transition from cocaine colony to global high-tech manufacturing headquarters similarly came from novel ways of deploying labour on the island. The post-colonial transformation originated in the application of skills honed on plantations to the free market. Historian Honghong Tinn rewrites the technological history of the island from the starting point that historians of technology have overemphasised singular ‘innovators’ while neglecting the broader labor forces involved in their production and implementation. Underpaid women used handcrafted skills honed in the home to produce magnetic core memory units, transistors, and microchips. Their ‘nimble fingers’ were highlighted by UN experts and local journalists alike.[31] Tinn ultimately argues that women’s participation in the production of electronics in Taiwan was above all a way to ‘standardize bodily and technical practice’.[32] This has a direct precedent: Taiwanese plantations. During the Japanese colonial period, sugarcane cultivations led to a phenomenon dubbed ‘the unbinding of Taiwan’ - the sudden demise of foot binding - as bound women were not able to work in physically demanding, low-wage jobs. Fifty years later, women workers and their embodied memory of craft skills propelled the Taiwanese economy forward. To this date, more than 75 percent of microchip manufacturers in TSMC are female, while less than 15% have engineering or managerial positions.
Technologies do not give way to a liberation from labour, but arise from the need to enclose labour in novel ways. What emerges across these histories is that technology is summoned into being by the problem of how to keep workers producing . Read this way, computing is not a neutral tool that happened to find industrial applications; . The persistence of this logic suggests that the history of computation should be told as a labour history, one in which enclosure, not efficiency, is the engine driving each new technological leap. Much discourse around microchips exceptionalises technological development, yet technology never quite seems to deliver on its promises. This is because what it promises remains utopian, while what it achieves is labour discipline.
The tendency of microchips to enclose labour has been accelerated in the current integration of generative AI into the workforce. AI is blamed for mass layoffs, even if the culprit may actually be bad business decisions. AI hallucinates and disperses accountability for mistakes. AI disposses workers of their data and extracts value from it.[33] Most of the jobs threatened by AI are held by women.[34] But in fact, the promises of AI mostly necessitate ‘skill shifts’: re/up/soft-skilling. All of this derives from a form of value that was first made quantifiable in relative terms by Babbage as the foundation of computing itself. Just as Babbage defined ‘skill’ circularly - as a quantity determined by wages which are determined by the demand for the workers skill - so is the value shift needed for AI seemingly judged against itself: the functions that can be done by AI shall be, and workers are only to fill in the blank spaces left by the tool. The plantation, the Analytical Engine, the integrated circuit, and the large language model are separated by more than a century, yet are all responses to the question of how to extract the discipline of labour from a workforce that is ‘free’.
High on its Own Supply Chain
The historical demise of legal cocaine could provide us insight into the future of AI. Coca plants were deeply embedded in local tribal cultures in South America. As a commodity, cocaine powder is a product of colonial extraction around which an era of the plantacionocene had oriented itself. Its criminalization involved controlling its access and marking it as something that could be ‘abused’ by individuals.
In early 2026, in his keynote speech at Computex – one of the world's largest ICT and technology trade shows held annually in Taiwan – Jensen Huang listed logos of Nvidia’s Taiwanese partners and its local supply chain. Taiwanese companies privy to Nvidia’s supply chain are eager to get their surge on the stock market: Taiwan has recently overtaken India to become the world’s fifth-largest exchange. Huang’s white lines on a black background can be read as something of an AI-high for this stock market surge.
Decolonisation in Taiwan has looked like freedom from one regime to be enlisted into the economics of another. Taiwan’s desire for sovereignty is bound with the world’s most powerful microchip cartel. However, no industrial miracle happens without sacrifice. High profit turnover in both production revenue and stock investment manufactures the consent of to work around the clock. Those suffering symptoms of present-day microchip plantation extractivism find their remedies in running clubs and psychiatric clinics[35]. ‘The supply chain of work stress’ involves at least 170,000 households; the demand for mental health provisions has grown with more than 50% in the Hsinchu region[36].
Nvidia CEO Jensen Huang at Computex, 2026, I-Hwa Cheng / AFP via Getty Images
Socio-political practices around the illegalisation of cocaine were key to gatekeeping its supply chains, eventually driving its global production into total illicitness. The parallels between this and the subjection of microchip exports to FDPR – especially export controls to China, under the auspices that microchips are a key to AI – is undeniable. The Chinese restrictions on microchips, and Taiwan’s position as world leader in the technology, without any formal sovereignty, creates an acute tension around its position. Protected as it is by its ‘silicon shield’ – the national economic strategy of making high-end microchips as a way to ensure the islands independence – is foundational to the global economy.
The implementation of AI in the workforce causes a cognitive decline just as long-term cocaine use brings on the loss of grey matter.[37] Microchips, as the main material building block for an AI-centered world economy, encourages the dispossession of social processes under quasi-feudalistic property rights - a new plantationocene. Between the growing illicitness of microchip export controls and growing moral questions around AI, the silver lining of the AI-led planetary crisis might be the re-equipment of personal tinkering skills. Our nimble fingers might be what gives us more agency over our tech in the future.
Lua Vollaard is a contemporary art curator and writer based in Amsterdam. Her current research focuses on the 'Art History of the Microchip', tracing the origins of photolithography through its artistic outputs.
Ying-Tzu Lin is a guest researcher at INC and postdoctoral researcher at Utrecht University. Taiwanese based in Amsterdam. She links computing with gardening.
Notes:
[1] Paul Gootenberg, Between Coca and Cocaine: A Century or More of U.S.-Peruvian Drug Paradoxes, 1860–1980, Working Paper no. 251 (Washington, DC: Woodrow Wilson International Center for Scholars, Latin American Program, 2001), P4.
[2] Ibid, P4.
[3] Micha Peters, “De Nederlandse Cocaïnefabriek,” Andere Tijden, 9 Feb. 2017, https://anderetijden.nl/artikel/107/De-Nederlandse-Cocainefabriek; and Hans H. Bosman, The History of the Nederlandsche Cocaïne Fabriek and Its Successors: As Manufacturers of Narcotic Drugs, Analysed from an International Perspective (PhD diss., Maastricht University, 2012)
[4] Toine Pieters, “Java Coca and the Dutch Narcotics Industry: An Almost Forgotten 20th C. History of Drugs Story,” Points: The Blog of the Alcohol and Drugs History Society, 10 Dec. 2012, https://www.pointshistory.org/post/java-coca-and-the-dutch-narcotics-industry-an-almost-forgotten-20th-c-history-of-drugs-story.
[5] “Eerste Schinkelstraat 30,” Amsterdam op de kaart, updated Jan. 2024, https://amsterdamopdekaart.nl/1850-1940/Eerste_Schinkelstraat/30.
[6] Arjo Roersch van der Hoogte and Toine Pieters, “From Javanese Coca to Java Coca: An Exemplary Product of Dutch Colonial Agro-Industrialism, 1880–1920,” Technology and Culture 54, no. 1 (2013): 90–116, https://www.researchgate.net/publication/295624195_From_Javanese_Coca_to_Java_Coca_An_Exemplary_Product_of_Dutch_Colonial_Agro-Industrialism_1880-1920
[7] 「帝國的溫室：產業、營養、認同與殖民時期的蔬菜園藝」呂紹理教授 — 科技部人文沙龍系列講座 第041場. YouTube, uploaded by 國科會人文沙龍計畫, https://www.youtube.com/watch?v=Zc1JtWAVuz4.
[8] Paul Gootenberg, Andean Cocaine: The Making of a Global Drug (Chapel Hill: University of North Carolina Press, 2005), and Shao-li Lu, “A Promising Tropical Medicinal Plant: Taiwan as the Production Hub of Japan’s Coca Empire”, Berichte zur Wissenschaftsgeschichte 47, no. 4 (2024), pp. 352–381, (pp. 353–354, 365).
[9] “Financial Results,” ASML, https://www.asml.com/en/investors/financial-results.
[10] “Who Makes the World’s Chips? Top 5 Semiconductor-Producing Countries Ranked,” WaferProcess, https://waferprocess.com/artificial-intelligence/who-makes-the-worlds-chips-top-5-semiconductor-producing-countries-ranked/.
[11] https://learnexportcompliance.com/insights/understanding-the-foreign-direct-product-rule
[12] Joe Rennison, “Stocks Overcame a Long List of Worries to Gain in 2025. A.I. Helped a Lot.,” The New York Times, 31 Dec. 2025, https://www.nytimes.com/2025/12/31/business/stock-market-2025-artificial-intelligence-bubble.html.
[13] Jeroen Kusters, Deb Bhattacharjee, Jordan Bish, Jan Thomas Nicholas, and Karthik Ramachandran, “2025 Global Semiconductor Industry Outlook,” Deloitte Insights, Deloitte Center for Technology, Media & Telecommunications, 4 Feb. 2025, https://www.deloitte.com/us/en/insights/industry/technology/technology-media-telecom-outlooks/semiconductor-industry-outlook.html
[14] “Is AI Profitable Yet?” isaiprofitable.com, https://isaiprofitable.com/
[15] Maan Barua, Rebeca Ibáñez Martín, and Marthe Achtnich, “Introduction: Plantationocene,” Theorizing the Contemporary, Fieldsights, Society for Cultural Anthropology, 24 Jan. 2023, https://www.culanth.org/fieldsights/introduction-plantationocene
[16] Steven Crook, “Taiwan’s Sugar Industry,” Taiwan Business TOPICS, American Chamber of Commerce in Taiwan, 8 July 2020, https://topics.amcham.com.tw/2020/07/taiwans-sugar-industry/
[17] “Camphor Monopoly in Formosa Island; Pioneers There Have an Adventurous Time to Avoid Head-Hunting Savages,” The New York Times, 23 May 1920, https://www.nytimes.com/1920/05/23/archives/camphor-monopoly-in-formosa-island-pioneers-there-have-an.html
[18] 太報, "臺灣歷史中的三大經濟作物," TaiSounds (太報), 30 November 2017, https://www.taisounds.com/news/content/132/42077
[19] Wen-Li Lee, K.D. Chiou, K.S. Chang, "Tropical fruit breeding in Taiwan: technology and cultivars," Acta Horticulturae, no. 1205, International Society for Horticultural Science, 2016, pp. 57 -588, https://www.ishs.org/ishs-article/1205_72
[20] Lu, “A Promising Tropical Medicinal Plant,” P. 358, P. 376.
[21] Lu, “A Promising Tropical Medicinal Plant,” P. 360.
[22] United Nations, The Supply of Raw Materials for the Manufacture of Cocaine (Working paper E/CN.7/W.58, Commission on Narcotic Drugs, United Nations, 1949), https://digitallibrary.un.org/record/1641763/files/E_CN-7_W-58-EN.pdf
[23] Lu, “A Promising Tropical Medicinal Plant,” P. 374.
[24] Lydia Kung, Factory Women in Taiwan (New York: Columbia University Press, 1994).
[25] Chih-ming Ka, “Agrarian Development, Family Farms and Sugar Capital in Colonial Taiwan, 1895–1945,” The Journal of Peasant Studies 18, no. 2 (1991): page number, https://doi.org/10.1080/03066159108438450
[26] Honghong Tinn, Island Tinkerers, Chapter 7.
[27] Chris Miller, Chip War: The Fight for the World’s Most Critical Technology (New York: Scribner, 2022), P. 239.
[28] Steven Crook, “Taiwan’s Sugar Industry,” Taiwan Business TOPICS, American Chamber of Commerce in Taiwan, 8 July 2020, https://topics.amcham.com.tw/2020/07/taiwans-sugar-industry/.
[29] Meredith Whittaker, “Origin Stories: Plantations, Computers, and Industrial Control,” Logic(s) Magazine, no. 19, 17 May 2023, https://logicmag.io/supa-dupa-skies/origin-stories-plantations-computers-and-industrial-control.
[30] Harry Braverman, Labor and Monopoly Capital: The Degradation of Work in the Twentieth Century (New York: Monthly Review Press, 1974).
[31]Honghong Tinn, Island Tinkerers: Innovation and Transformation in the Making of Taiwan’s Computing Industry (Cambridge, MA: MIT Press, 2024), P. 249.
[32] Honghong Tinn, Island Tinkerers, 2024, P. 244.
[33] Eliza McCullough, “Questioning the AI Revolution,” Leverhulme Centre for the Future of Intelligence (LCFI) Blog, University of Cambridge, https://www.lcfi.ac.uk/news-events/blog/post/questioning-the-ai-revolution
[34] Ben Casselman, “Forget Coders. The Real A.I. Threat Is in the Back Office,” The New York Times, 10 June 2026, https://www.nytimes.com/2026/06/10/business/economy/back-office-workers-ai.html
[35] National Science and Technology Council (Taiwan), 「東方矽谷：竹科高科技產業的地緣歷史與政經」工作坊, https://www.nstc.gov.tw/…/307e2c3c-7649-4f69-94b4-ca5b4a45ccb6.
[36] 陳虹瑾, “【竹科媽媽心內話4】無塵室裡憋尿憋出腎臟炎 女工程師曾多次流產：「辭職是因為我生不出小孩」” Mirror Media (鏡週刊), 7 Apr. 2025, https://www.mirrormedia.mg/story/20250407pol006
[37] Nataliya Kosmyna et al., “Your Brain on ChatGPT: Accumulation of Cognitive Debt When Using an AI Assistant for Essay Writing Task,” arXiv, arXiv:2506.08872, 10 June 2025, https://arxiv.org/pdf/2506.08872v1. This paper has attracted significant public debate, with supporters highlighting its findings on reduced cognitive engagement while critics have questioned aspects of its experimental design and the extent to which its conclusions can be generalized.