Yeni Sömürgeciliğe Direnmek: Çin ve Amerikan Büyük Dil Modellerinin Karşılaştırmalı Analizi

Öz

Bu çalışma, büyük dil modellerinin yalnızca teknik araçlar değil, küresel bilgi düzeninde ideolojik işlevler üstlenen sosyo-teknik aygıtlar olduğunu ileri sürmektedir. Araştırma, ABD merkezli ChatGPT ile Çin merkezli DeepSeek modellerinin Küresel Güney ülkelerinin kalkınma engellerine ilişkin sorulara verdikleri yanıtları karşılaştırmalı biçimde incelemektedir. Çalışmanın amacı, bu modellerin metinsel çıktılarında sömürgecilik mirası, kalkınma söylemleri ve epistemik eşitsizliklerin nasıl temsil edildiğini ortaya koymaktır. Yöntem olarak tematik analiz ve eleştirel söylem çözümlemesi birlikte kullanılmış, modellerin yanıtları içeriksel çeşitlilik, eleştirel derinlik ve ideolojik yönelim açısından değerlendirilmiştir. Bulgular, mülkiyet yapılarının, veri ekolojilerinin ve politik-ekonomik konumlanmaların modellerin metinsel çıktılarında somutlaştığını göstermektedir. ChatGPT’nin kapalı fikri mülkiyet rejimi ve bulut tekellerine dayalı yapısı hegemonik yapıları yeniden üretirken, DeepSeek’in açık erişim ve düşük maliyetli hesaplama stratejisi yapay zekâ üretimine alternatif, maliyet odaklı bir yaklaşımı yansıtmaktadır. Araştırma, yapay zekânın Küresel Güney açısından yalnızca bir teknoloji meselesi değil, bilgi egemenliği, bağımlılık ve eşitsizlik ilişkileri bağlamında politik bir mesele olduğunu ortaya koymaktadır.

Anahtar Kelimeler

• Büyük dil modelleri
• Bulut-köleci sömürgecilik
• Küresel Güney
• Bilgi egemenliği

Resisting Neo-Colonialism: A Comparative Analysis of Chinese and American Large Language Models

Öz

This study argues that large language models are not merely technical tools but socio-technical apparatuses that perform ideological functions within the global knowledge order. It comparatively examines the responses of the U.S.-based ChatGPT and the China-based DeepSeek to questions concerning the developmental obstacles faced by Global South countries. The aim is to reveal how the legacies of colonialism, developmentalist discourses, and epistemic inequalities are represented in these models' outputs. A mixed methodological design combining thematic analysis and critical discourse analysis was employed to evaluate the models’ responses in terms of content diversity, critical depth, and ideological orientation. The findings show that ownership structures, data ecologies, and political–economic positioning are concretised in the models’ textual productions. While ChatGPT’s closed intellectual property regime and reliance on cloud monopolies reproduce hegemonic structures, DeepSeek’s open-access and low-cost computational strategy reflects an alternative, cost-reduction–driven approach to AI production. Overall, the study demonstrates that artificial intelligence represents not only a technological phenomenon but also a political issue-one that is deeply intertwined with questions of knowledge sovereignty, dependency, and global inequality in the context of the Global South.

Anahtar Kelimeler

• Large Language Models
• Cloude-slavery colonialism
• Global South
• knowledge sovereignty

Kaynakça

1. Adomi, E. E. (Ed.). (2011). Handbook of research on information communication technology policy: Trends, issues and advancements. Information Science Reference.
2. Alvarado, R. (2023). AI as an epistemic technology. Science and Engineering Ethics, 29(5), Article 32. https://doi.org/10.1007/s10676-023-09710-4
3. Assay, M. (2025, April 14). DeepSeek’s open source movement. InfoWorld. Retrieved April 23, 2025, from https://www.infoworld.com/article/3960764/deepseeks-open-source-movement.html
4. Başaran, F., & Geray, H. (2011). Reproducing dependency: How hegemonic discourses shape ICT policies in the periphery. E. E. Adomi (Ed.), Handbook of research on information communication technology policy: Trends, issues and advancements in (pp. 66-82). Information Science Reference.
5. Becchetti, L., & Solferino, N. (2025). Unveiling biases in AI: ChatGPT's political economy perspectives and human comparisons. arXiv. https://doi.org/10.48550/arXiv.2503.05234
6. Bloom, P. (2025, February 12). DeepSeek: How a small Chinese AI company is shaking up US tech heavyweights. The Conversation. Retrieved March 20, 2025, from https://theconversation.com/deepseek-how-a-small-chinese-ai-company-is-shaking-up-us-tech-heavyweights
7. Buolamwini, J., & Gebru, T. (2018). Gender shades: Intersectional accuracy disparities in commercial gender classification. Conference on fairness, accountability and transparency in (pp. 77-91). PMLR.
8. Buyl, M., Rogiers, A., Noels, S., Bied, G., Dominguez-Catena, I., Heiter, E., Grgić-Hlača, N., & De Bie, T. (2024). Large language models reflect the ideology of their creators. arXiv. https://doi.org/10.48550/arXiv.2410.18417

9. Char, O. (2025, January 28). DeepSeek just confirmed my suspicions about OpenAI. Medium. Retrieved April 23, 2025, from https://medium.com/write-a-catalyst/deepseek-just-confirmed-my-suspicions-about-openai-d74edd89dcb5
10. Coeckelbergh, M. (2023). Democracy, epistemic agency, and AI: Political epistemology in times of artificial intelligence. AI and Ethics, 3(4), 1341-1350. https://doi.org/10.1007/s43681-022-00237-w
11. Choudhary, T. (2024). Political bias in large language models: A comparative analysis of ChatGPT-4, Perplexity, Google Gemini, and Claude. IEEE Access. https://doi.org/10.1109/ACCESS.2024.3371601
12. Couldry, N., & Mejias, U. A. (2019). The costs of connection: How data is colonizing human life and appropriating it for capitalism. Stanford University Press.
13. Dal Yong, J. (2020). Facebook’s platform imperialism: The economics and geopolitics of social media. O. Boyd-Barrett & T. Mirrlees (Eds.), Media imperialism: Continuity and change in (pp. 187-198). Rowman & Littlefield.
14. Ercan, F. (1996). Gelişme yazını açısından modernizm, kapitalizm ve azgelişmişlik. Sarmal Yayınevi.
15. Etike, Ş. (2023a). ChatGPT ile baş etmek: Emek ve eşitlik odaklı bir çerçevenin gerekliliği. Emek Araştırma Dergisi (GEAD), 14(23), pp. 115-132.
16. Etike, Ş. (2023b). Yapay zekâyı insan ve kamusal yarar için çalıştırmak: ABD ve AB politika belgelerinin önerileri. Memleket Siyaset Yönetim, 18(39), 220-249.
17. Feng, S., Park, C. Y., Liu, Y., & Tsvetkov, Y. (2023). From pretraining data to language models to downstream tasks: Tracking the trails of political biases leading to unfair NLP models. arXiv. https://doi.org/10.48550/arXiv.2305.08283
18. Fisher, J., Feng, S., Aron, R., Richardson, T., Choi, Y., Fisher, D. W., Pan, J., Tsvetkov, Y., & Reinecke, K. (2024). Biased AI can influence political decision-making. arXiv. https://doi.org/10.48550/arXiv.2410.06415
19. Frackiewicz, M. (2023). ChatGPT and the risks of deepening political polarization and divides. TS2 Space Blog. https://forums.swift.org/t/if-vs-available-vs-if-available/40266.
20. Fulay, S., Brannon, W., Mohanty, S., Overney, C., Poole-Dayan, E., Roy, D., & Kabbara, J. (2024). On the relationship between truth and political bias in language models. arXiv. https://doi.org/10.48550/arXiv.2409.05283
21. Han, B.-C. (2024). The crisis of narration (D. Steuer, Trans.). Polity Press.
22. Hartmann, J., Schwenzow, J., & Witte, M. (2023). The political ideology of conversational AI: Converging evidence on ChatGPT's pro-environmental, left-libertarian orientation. arXiv. https://doi.org/10.48550/arXiv.2301.01768
23. Geray, H. (1994). Yeni iletişim teknolojileri: Toplumsal bir yaklaşım. Kılıç Ali Matbaası.
24. Geray, H. (1995). Küreselleşme ve masaüstü sömürgecilik. Mürekkep, (Winter-Spring), pp. 33-47.
25. Geray, H. (2005). İletişim ağları ve masaüstü sömürgecilik. H. Geray & F. Başaran (Eds.), İletişim ağlarının ekonomisi: Telekomünikasyon, kitle iletişimi, yazılım ve internet. Siyasal Kitabevi.
26. Geray, H. (2024). Cloud-slavery colonialism: Is there a route for resistance? In Dijital eşitsizlik ve veri sömürgeciliği sempozyumu 15-17 Mayıs 2024 bildiri kitapçığı (pp. 187-202).
27. Geray, H. (2025). Yeni sömürgecilik ve iletişim araştırmaları bağlamında Nobel Ödülü. Ö. Özer (Ed.), Görünenin ötesi: İletişimin ekonomi politik yazıları in (pp. 131-162). Literatürk.
28. Khan, M. (2012, May 24). Governance and growth: History, ideology and methods of proof. N. Akbar, B. Kwesi, H. Stein, & J. Stiglitz (Eds.), Good growth and governance in Africa: Rethinking development strategies in (pp. 51-79). Oxford University Press. https://doi.org/10.1093/acprof:oso/9780199698561.001.0001
29. Lerner, D. (1964). The passing of traditional society: Modernizing the Middle East. The Free Press.
30. Liu, T. (2025, January 28). DeepSeek: How a small Chinese AI company is shaking up US tech heavyweights. The Conversation. Retrieved April 23, 2025, from https://theconversation.com/deepseek-how-a-small-chinese-ai-company-is-shaking-up-us-tech-heavyweights
31. Lopatto, E. (2025, March 5). The questions ChatGPT shouldn’t answer. The Verge. Retrieved April 1, 2025, from https://www.theverge.com/openai/624209/chatgpt-ethics-specs-humanism
32. Manvi, R., Khanna, S., Burke, M., Lobell, D., & Ermon, S. (2024). Large language models are geographically biased. arXiv. https://doi.org/10.48550/arXiv.2402.02680
33. Mejias, U. A., & Couldry, N. (2024). Data grab: The new colonialism of big tech and how to fight back. Penguin Random House.
34. Mirlees, T. (2020). US empire and cultural imperialism: A reconceptualization and twentieth-century retrospective. O. Boyd-Barret & T. Mirrlees (Eds.), Media imperialism: Continuity and change in (pp. 45-59). Rowman & Littlefield.
35. Nebil, F. S. (2025, April 11). Would you prefer your artificial intelligence to be right-wing or left-wing? Turk-Internet.com. Retrieved May 20, 2025, from https://turk-internet.com/yapay-zekânizi-sag-goruslu-mu-sol-goruslu-mi-tercih-edersiniz/
36. Papadimitropoulos, E. (2021). Platform capitalism, platform cooperativism, and the commons. Rethinking Marxism, 33(2), pp. 246-262. https://doi.org/10.1080/08935696.2021.1893108
37. Schiller, H. (1970). Mass media and American empire. Augustus McKelley Publishing.
38. Tolan, S., Miron, M., Gómez, E., & Castillo, C. (2019). Why machine learning may lead to unfairness: Evidence from risk assessment for juvenile justice in Catalonia. In Proceedings of the seventeenth international conference on artificial intelligence and law (pp. 83-92).
39. Varoufakis, Y. (2023). Techno-feudalism: What killed capitalism. Penguin Random House.
40. Waugh, L. R., Catalano, T., Masaeed, K. A., Hong Do, T., & Renigar, P. G. (2015). Critical discourse analysis: Definition, approaches, relation to pragmatics, critique, and trends. A. Capone & J. L. Mey (Eds.), Interdisciplinary studies in pragmatics, culture and society in (pp. 71-135). Springer International Publishing.
41. Walker, C., & Timoneda, J. C. (2024). Identifying the sources of ideological bias in GPT models through linguistic variation in output. arXiv. https://doi.org/10.48550/arXiv.2409.06043
42. Westwood, S. J., Grimmer, J., & Hall, A. B. (2025). Measuring perceived slant in large language models through user evaluations. Standford Business School.
43. Zajko, M. (2021). Conservative AI and social inequality: conceptualizing alternatives to bias through social theory. AI & Society, 36(3), 1047-1056.