Artificial Intelligence (AI) has permeated various aspects of our lives, but its influence on education is particularly profound and multifaceted. In a recent conversation, Sharon Tewksbury-Bloom sat down with Maha Bali, a professor of practice at the Center for Learning and Teaching at the American University in Cairo, to delve into how AI is transforming education and the broader implications of these changes. This blog post encapsulates their insightful discussion.
A Growing Presence: AI’s Ambiguous Impact
Maha Bali introduced herself as a veteran in the field of AI and education. With a background in computer science and 20 years of experience in educational support, she has witnessed the gradual integration of AI into educational practices.
Maha Bali: “AI is not a new phenomenon. Twenty years ago, AI was already being used in medical diagnoses, like breast cancer detection. What’s different now is the prevalence and accessibility of AI tools such as ChatGPT, which was made free and accessible to virtually everyone with an internet connection.”
However, Bali emphasized that this demand has led to misconceptions and hype, especially in education.
Misconceptions and Realities of AI in Education
Bali outlined several critical misconceptions about AI. Firstly, the assumption that everyone had access to AI platforms like ChatGPT is false—many regions, including Egypt and Saudi Arabia, were initially excluded.
Maha Bali: “The term ‘artificial intelligence’ itself is misleading, as it suggests some form of sentience or human-like intelligence, which it lacks. This misconception leads users to anthropomorphize AI, saying “please” and “thank you” to machines, further distorting its role and capabilities.”
Bali underscored the importance of recognizing AI as a statistical tool, not a sentient entity.
The Ethical Quandary: Bias and Data Integrity
Sharon Tewksbury-Bloom: One central concern is the bias inherent in AI. Bali provided a stark example of implicit bias by asking various AI platforms about terrorism, which predominantly returned answers linking terrorism to Islamic groups.
Maha Bali shared that this bias stems from the overwhelming Western-centric data sets these AIs are trained on. Consequently, AI can unintentionally reproduce systemic inequalities and biases present in the data. The lack of transparency about where AI sources its information exacerbates this issue, making it hard to trace or verify information accuracy.
A Call for Critical AI Literacy
So, what can educators do? Bali suggests that teachers need to understand how AI works to identify when and how it can be used effectively and ethically in their contexts.
Maha Bali gave an example of one of the funniest but useful exercises is using AI to create prompts that help design AI-resistant assignments. This involves using AI tools to develop assignments that students can’t easily complete using AI, pushing them to engage in deeper, critical thinking.
Practical Recommendations and Resources
Sharon Tewksbury-Bloom: With practical experience in creating custom AI models for specific tasks, Bali explored the potential benefits and limitations. She discussed using tools to summarize large documents, which can be helpful but often misses nuanced details.
Maha Bali: There are resources like AI for Education that provide prompt ideas for teachers. However, Bali reiterates the importance of using AI critically and being aware of its limitations. She is particularly enthusiastic about AI’s potential to support accessibility for people with disabilities, despite the challenges and limitations she has observed.
The Cultural and Linguistic Gaps
Sharon Tewksbury-Bloom: The discussion also highlighted how AI tools trained primarily in English struggle with non-English languages and cultures, resulting in outputs that can feel foreign or inaccurate to non-Western users.
Maha Bali: AI responses often mirror U.S. or U.K. cultural norms, marginalizing other cultures. The disparity is especially evident in countries like Egypt, where AI can misinterpret or inaccurately represent cultural specifics.
Looking Forward: Responsible and Ethical AI Use
Sharon Tewksbury-Bloom: Bali’s final thoughts centered on encouraging responsible and ethical use of AI. She emphasized the need for AI literacy, not just among educators and students but across all sectors affected by its rapid adoption.
Maha Bali: While AI’s capabilities are impressive, critical engagement and continual evaluation of its ethical implications are crucial. Understanding its biases, recognizing its limitations, and leveraging its strengths responsibly can help harness AI’s potential while mitigating its risks.
Continuing the Conversation
To stay updated on Maha Bali’s work and thoughts on AI and education, you can follow her blog. Her insights are invaluable for anyone interested in the intersection of AI, education, and social justice.
The dialogue between Sharon Tewksbury-Bloom and Maha Bali offers a nuanced look at the promises and perils of AI in education. As AI continues to evolve, conversations like these are essential to navigate its complex landscape responsibly.