Scholarship in the age of AI

What does it mean to be a scholar in the age of AI?

It is a truth universally acknowledged that AI is transforming traditional scholarship.

Anthea Roberts, visiting professor at Harvard Law School and founder of AI tool Dragonfly Thinking, recently explained: “In my academic work, I have Gemini, GPT, and Claude open and in dialogue… I’m constantly having a conversation across the four of us.”

And yet, the role of AI in the learning process is far from straightforward.

Recent research on the impact of using AI for essay writing sent ripples through the educational world when it found that an extraordinary 85% of students who used ChatGPT to support essay writing were unable to recall a single sentence of their essay just one hour after submission.*

By contrast, 90 per cent of students who wrote their essays unaided were able to recall information about their work.

These results present a stark indication that use of AI can lead to cognitive offloading.

This has of course been happening for millennia – every time you write a to-do list, you are offloading information. Yet with AI it seems people are offloading the thinking process itself and will therefore ‘atrophy’ critical thinking skills.

Even more worrying still is the suggestion that students growing up in an AI age will not merely ‘atrophy’ these critical thinking skills – they will never develop them.

It comes amid discussion about the impact of AI in universities, with one Oxford academic suggesting that professors can no longer tell whether an essay has been written by AI – so much so that “students can get a degree without reading a book”.

Despite this bleak picture of the impact of AI on education, within the research there are signs of an exciting AI-augmented future.

Follow-up studies show students who were trained to use AI as an effective part of the learning process produced significantly stronger arguments, reported higher cognitive engagement, and actually demonstrated lower reliance on AI-generated content.**

Thus, teaching students how to use AI is imperative.

Many schools are seeking to ‘AI-proof’ their assessment frameworks. However, if students don’t understand the cognitive basis for the cautious adoption of AI – or the means by which they can use it effectively – such strategies are likely to have limited long-term impacts.

Firstly, and most obviously, students need a solid grounding in AI literacy – a strong understanding of how AI models work, their strengths and weaknesses, including their biases. Whilst the evidence on the impact of AI on learning is mixed, there will not be a workplace unaffected by these changes by the time our students enter the job market. Those best placed to thrive will be able to use AI to support them in completing real-world tasks.

Secondly, and perhaps less obviously, students will require a strong understanding of their own human mind – both to understand the process of cognitive offloading and therefore best guard against it, and to make informed decisions about how to amplify their cognitive potential through the careful, considered use of AI.

That’s why at Putney High School we have embedded AI literacy within our Science of Learning curriculum, which empowers students with an understanding of how their brains work, enabling them to become both efficient and effective learners.

Having developed an understanding of the cognitive science underpinning learning, students can understand the complex impact AI is having on this process and, building on the work of Gerlich and other nascent research in this area, are explicitly taught how to use AI to amplify rather than circumvent learning.

Armed with this understanding, students are given opportunities to use AI to tackle real-world problems through our innovative Artemis Project.

Designed following consultation with experts, parents, and universities, Artemis is crafted to support students to develop key competencies that will allow them to thrive in the modern world – to become true modern scholars.

Of course, these include digital fluency and data intelligence but also, crucially, the human-centric skills of communication, leadership, critical thinking, and creativity.

Whether it is working as a team to agree and build an investment portfolio and using AI to critique that portfolio against defined criteria or using AI to support in preparing a presentation to pitch a children’s book they have themselves written and illustrated, it is the very human skills and experiences that are the bedrock of students’ learning.

Our approach must be forward-thinking, acknowledging the reality of the AI age and developing young people who understand how to use AI to enhance learning rather than replace it.

* Nataliya Kosmyna et al. Your Brain on ChatGPT: Accumulation of Cognitive Debt when Using an AI Assistant for Essay Writing Task

** From Offloading to Engagement: An Experimental Study on Structured Prompting and Critical Reasoning with Generative AI by Michael Gerlich, Center for Strategic Corporate Foresight and Sustainability, SBS Swiss Business School, 8302 Kloten, Switzerland. Data 2025, 10(11), 172; https://doi.org/10.3390/data10110172