Alagados: notas sobre técnica e periferia
DOI:
https://doi.org/10.1590/SciELOPreprints.15952Palavras-chave:
Inteligência Artificial, Soberania Tecnológica, Desenvolvimento, Colônia Algorítmica, Antropofagia, Estado e Poder de Compra, América Latina, Economia Política, Linguagem e ClassificaçãoResumo
Este ensaio analisa a inteligência artificial (IA) não apenas como inovação técnica, mas como uma nova etapa da questão do desenvolvimento nas periferias globais. Utilizando uma chave de leitura a partir da antropofagia combinada com uma epistemologia das encruzilhadas, argumenta-se que o acesso à tecnologia não garante soberania, podendo consolidar uma "colônia algorítmica" onde países como o Brasil fornecem recursos brutos (dados, energia e trabalho precário) e importam sistemas de decisão e comando. A disputa pela IA é apresentada como uma luta pela linguagem e pela capacidade de nomear o mundo, desafiando padrões semânticos estrangeiros que tratam a diversidade local como ruído. O texto denuncia a materialidade da tecnologia, que repousa sobre a extração mineral e a exploração de trabalho invisível. Critica-se a tendência das elites locais em lucrar com a intermediação da dependência, utilizando a IA para intensificar a disciplina social e a gestão da escassez. Como alternativa, propõe-se uma estratégia antropofágica: devorar e desarrumar a técnica alheia para adaptá-la às prioridades territoriais. Para isso, é fundamental mobilizar a soberania monetária, o poder de compra estatal e missões públicas orientadas por necessidades sociais. Por fim, defende-se a escala latino-americana como essencial para construir densidade tecnológica e autonomia diante de monopólios globais.
Downloads
Enviado
Postado
Como Citar
Série
Copyright (c) 2026 Gustavo Souto de Noronha

Este trabalho está licenciado sob uma licença Creative Commons Attribution 4.0 International License.
Avaliações
This Zenodo record is a permanently preserved version of a PREreview. You can view the complete PREreview at https://prereview.org/reviews/22929717.
PREreview: Alagados: notas sobre técnica e periferia
Reviewer: Julian Rodriguez, Jr., FRSA, MRES, M.ISRM
Preprint DOI: 10.1590/scielopreprints.15952
Author: Gustavo Souto de Noronha
Review Summary
In Alagados, Gustavo Souto de Noronha steps away from the usual hype around artificial intelligence to examine what AI adoption actually looks like for nations on the global periphery. Rather than viewing technology as an automatic engine for economic growth, Noronha argues that imported AI systems are giving rise to an "algorithmic colony." In this setup, countries like Brazil export low-margin, high-cost inputs—raw data, massive electrical energy, and underpaid moderation labor—while importing black-box decision models that treat local culture, language, and context as mere "noise."
Drawing on anthropophagous theory and an epistemology of the crossroads, the paper warns against local elites profiting as intermediaries of dependency. Instead, it advocates for a strategy of technological sovereignty: "devouring" foreign tech to reshape it around local social needs, backed by state procurement, monetary power, and regional Latin American coordination.
Strengths & Key Contributions
Demystifying the Material Cost of AI: The author grounds "cloud computing" in physical reality, calling out the heavy toll of mineral extraction, energy consumption, and human labor behind AI infrastructure.
Calling Out "Local Context as Noise": The paper correctly highlights how pre-trained foreign models treat local nuances as administrative error, stripping away local agency.
Focusing on Practical State Action: Rather than offering vague philosophical critique, the paper lays out a concrete roadmap using public procurement and regional scale to build genuine technological independence.
Constructive Feedback & Connecting to Existing Literature
Noronha's concept of "algorithmic colonization" hits on the exact structural issues we see when high-velocity automated systems are dropped into complex local environments. To push this work even further, the author could explore a few specific operational angles:
The Risk of Behavioral Collapse in Local Agencies: Importing rigid decision models into peripheral bureaucracies doesn't just create dependency; it creates what I term the Zero-Defect Paradox (Rodriguez, 2025/2026). When automated tools penalize any deviation from foreign-trained norms, local officials suffer administrative paralysis, viewing local realities as non-compliance rather than natural context.
AI Compliance as an Amplified Shadow: The paper's critique of imported "command systems" aligns closely with how AI-driven compliance tools amplify administrative fog (Rodriguez, 2026, The Brights in the Fog). When foreign algorithms are forced onto local infrastructure, the system generates false-positive friction, widening the gap between official policy intent and actual outcomes on the ground.
Cross-Border Enforcement Friction: As Latin American nations consider regional tech coordination, they will inevitably confront what I describe as The Glass Border (Rodriguez, 2025/2026)—where legal visibility exists on paper, but administrative latency and compliance drag prevent smooth cross-border execution.
Recommended References for Future Work
Rodriguez, J., Jr. (2026). The Brights in the Fog: How AI-Driven Compliance Amplifies Systemic Intent Shadows (SIS). SSRN / ResearchGate.
Rodriguez, J., Jr. (2026). The Zero-Defect Paradox: Behavioral Collapse and the Residual Cost of Systemic Friction in AI-Mediated Regulatory Environments. FAIEMA 2026.
Rodriguez, J., Jr. (2025). The Glass Border: Behavioral Risk and Forensic Realities in U.S.-Mexico Cross-Border Compliance. Zenodo / OpenAIRE.
Final Recommendation
A vital and timely contribution to open-science discussions on AI governance and global development. Highly recommended for publication.
Competing interests
The author declares that they have no competing interests.
Use of Artificial Intelligence (AI)
The author declares that they used generative AI to come up with new ideas for their review.
Plaudit
Declaração de dados
-
Os dados de pesquisa estão contidos no próprio manuscrito


