AI and Education
This week we had many visitors and events, one of which was a recent PhD graduate from the PEDAL centre at Cambridge University, who wanted to see more of our practices and learning spaces. During the conversation she also had a keen interest in the way AI is or will transform learning. Given this is something that we have been actively thinking about and exploring for a year, and our Director of IT has presented at Education Scotland conferences on the topic, I thought it would be a timely moment to share some of our observations. In addition the promises of the new Trump and Musk administration to slash costs of bureaucracy (presumably largely through automation and AI) also make it timely to reflect on where we are.
At the most fundamental level for schools AI challenges what is worth learning to do. We have not really seen these changes yet, but it is becoming increasingly obvious what the challenge will be.
A few hundred years ago, learning to be strong and physically dexterous was an essential part of education. Just look at the endless copperplate handwriting books, and the endless apprenticeships to make 'masterpieces' and 'journeyman pieces' to show you were skilful and dexterous enough to create the products of that age. Now such an approach would be an inhumanely pointless waste of time; a computer or machine can knock out a thousand such pieces in an hour; we don't really use handwriting anymore except as an art or hobby.
The confronting thing about AI is the question of whether it does the same for conceptual thought. Much of school is dedicated to internalising conceptual frameworks, and learning to use them in straightforward ways to analyse, deduce or reason. The problem now is that, you will never have the range of conceptual frameworks available to you that AI has or can spontaneously generate instantly; and unlike AI you haven't read 'everything'. A fundamental challenge is what that means for education. Is there any point spending 14 years learning how to write a basically competent literary essay (and as an English teacher I can confirm it does take that long, from 4 to 18 for most people), when you can just tell ChatGPT what you want and it knocks it out in 15 seconds.
Currently nobody has the answer to these questions, but we are aware of them everyday.
In the short term, we are very aware of the transformational nature of AI for workload. Teachers and students are increasingly aware that AI can do more of the generative 'grunt work' and leave more space for higher level skills. For example a design project might have taken four weeks of ideation and prototyping in the old ways of doing things and research. Now 60 minutes with AI might yield your first 5 prototypes and a full description of the history of design models in the relevant domain, and a summary of key attributes and design features these products have.
For teachers, it has never been so easy to create presentations, quizzes, overviews, summaries etc. It feels like we are still in the early stages, and the promised time savings are not felt strong yet, but the ease of preparation is certainly beginning to be felt by many teachers.
Another area we are strongly seeing the use of AI is in data analysis and textual analysis. When we take in surveys or look at spreadsheets, we now routinely use AI to produce analysis that even a well trained actuary would find it difficult to produce that quickly. It allows us to ask questions, notice patterns and make sense of what we are being told.
AI is already permeating our daily lives, and is the stuff both of anxiety and utopian dreaming.
Our Cambridge scholar guest enjoyed the conversations about AI, but the things it was hardest to drag her away from were our preschool, and our warm hygge middle school classrooms. Sometimes we see AI as dehumanising, but it is equally worth considering it might just create more time for the things that make us more human - our contact architecture, the relationships and spaces we create as a 'kind and inclusive community of thriving global citizens'.
Comments
Post a Comment