Sandwich makers – E-Learning Provocateur

The latest research paper to cause a stir among my peers is Your Brain on ChatGPT: Accumulation of Cognitive Debt when Using an AI Assistant for Essay Writing Task by a bunch of boffins from Massachusetts, mostly MIT.

It’s quite long (over 200 pages) but in the interests of time, I consider the premise of their argument to be articulated by this paragraph that refers to previous research published in BJET, which they go on to triangulate with methodologies such as electroencephalography:

One of the most prominent effects of using AI in writing is the shift in how students engage with the material. Generative AI can generate content on demand, offering students quick drafts based on minimal input. While this can be beneficial in terms of saving time and offering inspiration, it also impacts students’ ability to retain and recall information, a key aspect of learning. When students rely on AI to produce lengthy or complex essays, they may bypass the process of synthesizing information from memory, which can hinder their understanding and retention of the material. For instance, while ChatGPT significantly improved short-term task performance, such as essay scores, it did not lead to significant differences in knowledge gain or transfer [55]. This suggests that while AI tools can enhance productivity, they may also promote a form of ‘metacognitive laziness,’ where students offload cognitive and metacognitive responsibilities to the AI, potentially hindering their ability to self-regulate and engage deeply with the learning material [55]. AI tools that generate essays without prompting students to reflect or revise can make it easier for students to avoid the intellectual effort required to internalize key concepts, which is crucial for long-term learning and knowledge transfer [55].

In another words, while the likes of ChatGPT make writing quicker and easier, they eliminate the need for critical thinking and hence inhibit learning.

And this should come as no surprise to any digital native who’s only ever used a GPS navigation system and thus has no idea of how to get from A to B without it.

But what has surprised me is the amount of heat that the paper has attracted on social media. While I’m no stranger to calling out the limitations of academic research when I feel it’s warranted, I do feel some of the rocks being thrown at the Bay Staters are a tad unfair. But that’s a story for another day.

If we shift our attention from experimental design (and dare I suggest academic politicking) to real-world practice, then I think we can progress; not by preventing students from using modern technology to produce scholarly work, but by showing them how to do so in a way that promotes understanding and knowledge retention.

When I reflect on my own use of ChatGPT, I’m cognisant of the fact that I’m reluctant to let the robot do all the heavy lifting. For example, if I were to brainstorm a list of topics to be covered by a new curriculum, I’d draw up my own list first, then I’d prompt ChatGPT to generate a list, then I’d compare the pair – adding the topics I left out and consider important, and ignoring the ones I don’t.

Similarly, if I were to write an essay – as per the MIT study – I’d outline my key themes and structure first, then I’d prompt ChatGPT to write a draft, then I’d compare the pair.

My approach is what Kerry Harrison calls an AI sandwich, whereby the workflow is human-AI-human. She describes the steps as follows:

1. Before you start using AI tools, spend time thinking, exploring, researching and crafting your prompts. Use your amazing human brain.
2. Apply your AI writing tool of choice.
3. Take the output and fine-tune it. Add/refine prompts, apply your human creativity, flesh out an idea. Make it your own.

While Harrison feeds her thinking into her prompts, and I tend to withhold mine for cross-referencing purposes, our approach remains the same – we use AI to complement our intelligence, rather than replace it. So our outcomes are the products of combined forces, better than either force alone.

Which may be irrelevant to a great many people. Alas, not all students want to learn. But we should help those who do become sandwich makers.