Redefining Education Amid AI Learning and Knowledge Theft Controversy

The fact that generative artificial intelligence (AI) only produces answers within the scope of a user's existing knowledge poses a fundamental question about what constitutes true learning and education. This issue is closely linked to the 'knowledge theft' controversy, which suggests that AI may infringe upon the intellectual property rights of original authors in its process of creating new outputs based on learning data. Such debates demand deep contemplation on how our educational systems should be transformed.

Humans experience learning at various points in life, including eating, walking, speaking, and reading. A 10-month-old child decided to take their first steps independently, forging a connection in their brain between the concept of movement and the actual act of moving. The author has witnessed many 'lightbulb moments' in university classrooms for over two decades, and while learning sometimes occurs in structured classroom environments, it can also happen in the kitchen or backyard.

The advent of generative AI tools has raised many questions about the process and purpose of education. Generative AI-based learning systems offer the possibility of acquiring knowledge quickly and easily. Arizona State University's (ASU) 'AI learning advisor' service, 'Atomic,' fulfills these expectations. This beta system promotes to its monthly subscribers that they can 'learn only what they want,' creating customized non-credit learning modules in just five minutes using ASU's own online course content. When a user asks about their 'main goals or objectives,' the AI designs a personalized learning path based on these inputs.

However, issues have emerged regarding AI learning systems unauthorizedly collecting digital materials without instructors' consent. Videos, slides, and assignments used in ASU courses were found to have been scraped by AI, which then browses keywords within ASU's online course materials and provides short summaries. Concerns have been raised that information could be pieced together out of its original context or that biased knowledge might be delivered through this process. This is akin to AI only responding to questions the inquirer already knows the answers to, making it difficult for AI to fully replace deep human understanding and critical thinking processes.