New technology typically inspires both utopian dreams and dystopian fears. AI is a salient example of this optimism-pessimism dichotomy. 

This week, the New Zealand Initiative released a new report, exploring the implications of AI for education. Welcome to the Machine analyses the risks utopian thinking about AI poses to education. But the report does not take a dystopian view.  It also explores promising potential for teachers to use AI productively. 

Some AI enthusiasts have argued that AI can free young people from the need to commit knowledge to memory. Instead, they believe, students can focus on critical thinking and creativity. This is superficially plausible. Notwithstanding the odd hallucination, AI can produce a factual summary on just about any topic.  

Unfortunately, the idea that AI can liberate students from the hard yards of learning basic knowledge is as misguided as it is appealing. The things that the techno-utopians would like them to focus on – critical thinking and creativity – depend on knowledge. 

When we commit knowledge to long-term memory, we build cognitive schemas. Our schemas collectively provide us with a working model of the world. We draw upon that model constantly when we think.  

When we hear a claim, we test it against our knowledge schemas, using reason, to see if it is likely to be true. The more we know, the more likely we are to be able to sift truth from falsehood. Knowledge, then, is the raw material for thinking critically. 

Creativity also depends on cognitive schemas. When we think creatively, we introduce an element of randomness into our schemas. Much of the time the results will be no good. But occasionally, we come up with a gem of an idea. Our knowledge schemas enable us to distinguish our good ideas from our bad ones. The more knowledge we have, the more likely it is that our creative endeavours will be successful. 

Rather than using AI as a substitute for learning knowledge, its greatest promise in education is as a virtual teaching assistant. Rather than giving students ideas, it might coach them to improve their own, by asking probing questions. It might provide formative feedback, helping to move students towards their learning goals. Like any teaching assistant, though, AI must be supervised by teachers. 

Used well, AI will provide powerful support for students and teachers alike. But we must not allow personally held knowledge to be supplanted by overreliance on technology.

Dr Michael Johnston's report, Welcome to the Machine: Opportunities and risks of generative AI in education, was published on 17 June.