The debate around AI in education starts in the wrong place. Institutions ask whether to ban it, whether accreditation bodies are responding fast enough, or how to police its use. The question that matters is simpler: has learning actually taken place? A polished document or a research outline produced in seconds tells you what the tool can do. It tells you very little about what the student has understood.
Creative education has navigated technological shifts before. The typewriter changed how students produced work. Search engines changed how they researched. AI is another step in that evolution. Pretending it does not exist is pointless because students are already using it. The real test is whether they have enough foundational knowledge to judge what the tool gives them.
AI is not magic
One of the biggest misunderstandings about AI is that people treat it as something mysterious. It is a calculation. It takes what has been fed into it and produces an approximation. In creative fields, the work is not only about structure or surface. It is about judgment, context, lived experience, and emotional intent.
A student who understands music can listen to an AI-generated composition and hear where it fails. They can hear what is generic, unresolved, or disconnected. A student without that foundation may accept the output simply because it looks complete. The same applies to written work. A student can submit something polished, but if they cannot explain the argument, defend the sources, or show how the ideas connect, the document reveals little about their learning. For professionals working in creative disciplines, resources on AI for Creatives explore how tools intersect with craft, but the core principle remains: the human must know enough to evaluate the machine's output.
Foundations still matter
Used properly, AI can guide early thinking, provide structure, and remove some of the administrative burden from creative work. It cannot replace the hard part of education, which is learning how to think through a field with enough depth to make informed decisions.
This matters acutely in contexts where cultural knowledge is not well documented online. In South Africa, much of the country's music history, mythology, and creative traditions still sit with people, communities, practitioners, and memory. When AI is asked about local creative traditions, the answers are often superficial. If the knowledge is not documented, AI can only guess from what it has. If education abandons foundations for tools, it weakens what students need most. In music, sound, and production, students still need to understand scales, notation, acoustics, sound behaviour, performance, language, and creative process. Without those foundations, AI cannot elevate their work. It can only make weak work look more finished.
What assessment should measure
Education is not just content transfer. It is socialisation, collaboration, debate, feedback, networking, and learning how to operate within a field. Assessment should ask whether the student can think, adapt, explain, and apply knowledge in context. Institutions cannot abandon the concepts that govern their industries or prevent students from encountering the innovations that shape them. As tools become part of AI for Education, the question is whether institutions use them to deepen learning or to disguise the absence of it.
"Has learning actually taken place? That is the test," the academy's academic leadership said. "AI can produce a polished document or a research outline in seconds, but speed and appearance are not evidence of understanding."
Why this matters for Creatives
The line between using AI as a tool and letting it substitute for craft is thin. The practical test is straightforward: if you can use the tool, explain the process, challenge the output, and improve it through craft, learning has happened. If you cannot, the machine did not make you creative. "It merely coloured in the numbers." For working creatives, the implication is clear. Your value is not in producing finished-looking work quickly. It is in the judgment that knows when the output is generic, the ear that hears what is unresolved, and the experience that tells you what to change. AI can accelerate process, but the foundations that make that acceleration meaningful are built through study, practice, and the kind of deep engagement no tool can shortcut.
Your membership also unlocks:
