Introduction

Over the past year, news about Artificial Intelligence (AI) has abounded. Information about breakthroughs and new applications have become commonplace, and we have been thrust into a world where AI-enabled technologies are starting to change how we work and live. For better or for worse, we have ushered in the era of AI.

Many are asking what the implications of this might be for the education sector. Will it affect teaching and learning positively or negatively? How can AI-enabled technologies personalize education – and will this be educationally beneficial? What organizations are already working on AI in education and how, if at all, is this work regulated? An air of uncertainty pervades the sector in terms of the benefits and risks of harnessing AI in education.

At OER Africa, we have written extensively on how Open Educational Resources (OER) might improve aspects of education, including access, relevance, and quality. Likewise, the intersection of OER and AI necessitates greater exploration, particularly given the opportunities that it offers to scale access to high quality education.

In this article, we consider three ways that AI might change how we use OER.

How AI could change our engagement with OER

1. OER Content Development

AI tools can be used to develop new OER using natural language processing and machine learning capabilities. They might be able to help educators create interactive learning materials, assessments, and learning simulations, expanding the range of available OER and enabling novel teaching practices.

One of the most popular platforms that demonstrates such capabilities is ChatGPT (or Chat Generative Pre-Trained Transformer), an AI chatbot developed by OpenAI. Since its release in November 2022, is has amassed over 100 million users.

There is no doubt that generative tools like ChatGPT hold great potential to save time and effort for OER creators. With a few well-crafted prompts, ChatGPT can: generate thousands of words on a subject; create dozens of sample questions that could be included in an open textbook for learners to be able to self-evaluate their own learning; create lesson plans and assignments; and develop question prompts that can be used as asynchronous discussion prompts in discussion forums.[1]  However, it is crucial that OER creators use their own expertise to check that what is generated by ChatGPT is indeed correct.

For educators working with OER, copyright ownership of AI-generated works is important to determine as, by definition, OER are materials unencumbered by legal restrictions that may prevent the reuse, sharing, redistribution, and adaptation of copyrightable works. While some are using the rise of generative AI to question the validity of copyright itself, the question of who owns the copyright when a work is created by AI is a very murky area, both legally and ethically.[2,3]  When we asked ChatGPT whether the content it produces is openly licensed, it had the following to say:

As of my last update in September 2021, the content generated by ChatGPT and similar AI language models is not openly licensed. AI language models, including ChatGPT based on the GPT-3.5 architecture, are proprietary technologies developed by organizations like OpenAI.

…Therefore, when using content generated by ChatGPT or any other AI language model, it is essential to review the terms of service, usage policies, and any specific guidelines provided by the organization that owns the AI model to ensure compliance with their requirements.

It's worth noting that the field of AI and its legal and ethical implications are continually evolving, and there might be changes or developments in the licensing and usage of AI-generated content beyond my last update. I recommend checking with the organization that provides the AI service for the most current and accurate information regarding the licensing and usage of their AI-generated content.[4]

We recommend that users of these technologies stay abreast of these kinds of debates, read terms of service of the organizations that create these technologies, pay close attention to licensing conditions, and state clearly when they have used AI tools to generate intellectual property.

2. Personalized Learning

Some AI algorithms can develop tailored recommendations for OER based on a learner's performance, learning preferences, and development areas. This implies that learners can use OER that meet their requirements, making the learning pathway more engaging and effective.

For example, Siyavula is a South African organization that provides personalized and adaptive learning platforms. Siyavula has produced book titles from Grades 4-12. These are high quality OER that are aligned with the South African curriculum for mathematics, physics and chemistry. Learners can now also access Siyavula’s adaptive learning software, which adjusts the difficulty levels of exercises through machine learning to cater to each learner’s individual needs.[5]

3. Translation and Localization

AI can enable translation and localization of openly licensed content. Software with machine translation capabilities, such as Google Translate, can translate OER into different languages, facilitating knowledge sharing. It is always recommended that users state when they have used these kinds of tools for translation purposes.

The United Nations Educational, Scientific and Cultural Organization (UNESCO) has been exploring the use of AI for the translation and localization of educational content. It has collaborated with partners to develop machine translation systems and tools that bridge language gaps in accessing OER. For example, the Global Digital Library (GDL) collects existing high-quality openly licensed reading resources and makes them available on the web, mobile and for print. The platform also supports translation and localization of GDL-resources to more than 300 languages. UNESCO partnered with the GDL team under the auspices of the Norwegian Programme for Capacity Development in Higher Education and Research for Development (known as NORAD) and the Global Book Alliance to launch the GDL in Asia. Reading materials in 41 Asian languages, including seven Nepali languages were launched.[6]

Despite the benefits of integrating AI into OER, there are several potential challenges and concerns. For example, the issue of data privacy has received a lot of attention recently, as the use of AI algorithms often entails the collection and analysis of user data. Ensuring that such data is stored securely and used responsibly is critical to maintaining the trust and privacy of both learners and educators.[7]

A second challenge is the potential for AI to exacerbate existing inequalities in education. As AI-powered OER become more widespread, there is a risk that those who cannot access such resources or platforms may be left behind due to unstable internet connections for example. There may also be inherent biases in the data that is used to train AI models, such as a lack of data from Sub-Saharan African countries. Thus, introducing measures to ensure that AI-driven educational tools are accessible to learners regardless of their geography or socioeconomic contexts is key to promoting educational equity.[8]

Conclusion

Regardless of one’s outlook on the impact that AI could have on society over time, its integration into most spheres of our lives in some shape or form is progressing fast. With regard to OER, AI offers exciting opportunities to augment the production, dissemination, and access to quality educational resources. However, rolling out such capabilities means that we need to consider potential shortfalls, including that we might inadvertently inhibit access to such platforms for those who face educational barriers.

Some further reading on this topic:


Related articles in OER Africa’s archive


[1] Lalonde, C. (2023). ChatGPT and Open Education. BC Campus. Retrieved from: https://bccampus.ca/2023/03/06/chatgpt-and-open-education/

[2] Lalonde, C. (2023). ChatGPT and Open Education. BC Campus. Retrieved from: https://bccampus.ca/2023/03/06/chatgpt-and-open-education/

[3] See article here

[4] Conversation with ChatGPT on 24 July, 2023. OpenAI's ChatGPT, based on the GPT-3.5 architecture.

[5] See article here

[6] See article here

[7] Frackiewicz, M. (2023). AI in Robotic Open Educational Resources. Retrieved from https://ts2.space/en/ai-in-robotic-open-educational-resources/#:~:text=AI%2Dpowered%20chatbots%2C%20for%20instance,thinking%20and%20problem%2Dsolving%20skills.

[8] Frackiewicz, M. (2023). AI in Robotic Open Educational Resources. Retrieved from https://ts2.space/en/ai-in-robotic-open-educational-resources/#:~:text=AI%2Dpowered%20chatbots%2C%20for%20instance,thinking%20and%20problem%2Dsolving%20skills

Authors
Mohini Baijnath