[Paper Review] ChatGPT is all you need to decolonize sub-Saharan Vocational Education

Isidora Tourni, Georgios Grigorakis+3 more|arXiv (Cornell University)|Apr 11, 2023

Text Readability and Simplification4 citations

TL;DR

This paper proposes that sub-Saharan African countries should prioritize Technical and Vocational Education and Training (TVET) over traditional academia by integrating fine-tuned Large Language Models (LLMs) like ChatGPT to decolonize education, enabling culturally relevant, scalable, and affordable skill development that accelerates socioeconomic mobility and empowers local knowledge systems.

ABSTRACT

The advances of Generative AI models with interactive capabilities over the past few years offer unique opportunities for socioeconomic mobility. Their potential for scalability, accessibility, affordability, personalizing and convenience sets a first-class opportunity for poverty-stricken countries to adapt and modernize their educational order. As a result, this position paper makes the case for an educational policy framework that would succeed in this transformation by prioritizing vocational and technical training over academic education in sub-Saharan African countries. We highlight substantial applications of Large Language Models, tailor-made to their respective cultural background(s) and needs, that would reinforce their systemic decolonization. Lastly, we provide specific historical examples of diverse states successfully implementing such policies in the elementary steps of their socioeconomic transformation, in order to corroborate our proposal to sub-Saharan African countries to follow their lead.

Motivation & Objective

• To address the persistent educational inequity in sub-Saharan Africa by shifting focus from elite academic institutions to practical, accessible TVET systems.
• To counter the legacy of colonial educational models that prioritize Western academic standards over local needs and Indigenous Knowledge (IK).
• To leverage the scalability, personalization, and low-cost deployment of LLMs to democratize access to technical and vocational training.
• To demonstrate how AI-driven education can accelerate socioeconomic development by aligning training with real-world, community-based needs.
• To advocate for a policy framework that embeds LLMs in TVET to foster systemic decolonization and sustainable development in post-colonial contexts.

Proposed method

• Propose a policy framework prioritizing TVET over traditional higher education in sub-Saharan African countries.
• Advocate for fine-tuning LLMs on region-specific cultural, linguistic, and vocational content to ensure relevance and reduce bias.
• Integrate multimodal generative AI (e.g., DALL·E, Stable Diffusion, Jukebox) with LLMs to create immersive, interactive learning experiences.
• Use LLMs to train local personnel in emergency response, first-aid, and technical trades through contextual, feedback-driven, personalized instruction.
• Position LLMs as tools to embed Indigenous Knowledge (IK) into curricula, countering the marginalization of local epistemologies.
• Emphasize the need for human supervision and infrastructure (internet, electricity) to ensure reliability and equity in deployment.

Experimental results

Research questions

• RQ1 How can Large Language Models (LLMs) be leveraged to decolonize vocational education in sub-Saharan Africa?
• RQ2 Why is a TVET-focused educational model more suitable for post-colonial African nations than traditional academic institutions?
• RQ3 In what ways can LLMs integrate Indigenous Knowledge (IK) into vocational curricula more effectively than traditional search engines?
• RQ4 What are the risks of deploying generic LLMs like ChatGPT in African educational contexts without cultural fine-tuning?
• RQ5 How can AI-driven TVET systems improve socioeconomic mobility and reduce dependency on imported expertise and colonial-era educational models?

Key findings

• Sub-Saharan Africa accounts for only 1.06% of global AI publications, highlighting systemic underrepresentation in AI and education.
• Traditional academic institutions in post-colonial African states often suffer from underfunding, lack of academic freedom, and gender imbalance, undermining their effectiveness.
• LLMs can be fine-tuned to incorporate Indigenous Knowledge (IK), offering a more inclusive and contextually relevant alternative to generic, functionally oriented search engines.
• TVET systems supported by LLMs enable faster, more practical skill acquisition with immediate real-world application, especially in emergency response and technical trades.
• The integration of multimodal AI (e.g., text, image, audio) with LLMs can create holistic, adaptive learning experiences that surpass traditional educational tools.
• Despite their promise, LLMs risk reinforcing colonial ideological hegemony if not trained on African cultural and economic realities, necessitating careful fine-tuning and human oversight.