By Professor Ari Seligmann, Jo Hook, Carmen Sapsed, and Andrew Junor
Posted Friday 30 May, 2025

Many of us are contemplating how we can guide students to incorporate diverse intelligences into the production and expression of disciplinary knowledge, whilst still demonstrating the required forms of learning. 

The work-in-progress presented here is the result of monitoring, evaluating and reflecting on some AI integration efforts in our Arts faculty. We initiated a dialogue in a Faculty workshop, and developed our colleagues’ insights into a set of resources. We are currently using the evolving resources to engage Arts colleagues in conversations about writing with AI and gathering examples of practice. We wish to share the resources here not as a solution, but as a thought experiment to solicit more input and broaden the discussions across disciplines.

Reimagining the essay

The essay is an enduring staple in arts and humanities disciplines as a valuable method of developing students’ research, critical thinking, critical analysis and argumentation skills, while students hone a distinct and individual academic voice. The essay writing process also encompasses ‘writing as thinking’, requiring deep engagement with unit material and making space for intellectual growth  (Hounsell, 2005). As AI capacity continues to grow, and AI is increasingly integrated into our daily tools, such as Microsoft Word and Google Docs, coupled with the problems with trying to detect AI in submitted outputs, it is impossible to ‘design AI out’ of essay tasks. What is the future of the essay in an age where digital and human intelligences coexist? How can we encourage our students to develop their writing practice rather than just outsource the writing and thinking process to the growing range of digital and AI tools?

What is the future of the essay in an age where digital and human intelligences coexist?

Reflecting on these questions leads us to a further purpose of the essay as a tool for assessment of learning (Hounsell, 2005). A recent post on the LSE HE blog, We tried to kill the essay – now let’s resurrect it, reminds us that we often treat the essay as a product transacted for a grade. The post, which builds on the Manifesto for the essay in the age of AI, offers a thoughtful discussion on how to save the essay as an assessment task – by reframing it for students as a space for independent thinking and personal sovereignty. AI can participate in this process as a collaborator that challenges students at each stage in the writing/thinking process.

Khalifa & Albadawy (2024), among others, have identified the multiple domains of the academic writing process where AI may, with human oversight, contribute to academic functions. They suggest idea development and research design, content development and structuring, literature review and synthesis, data management and analysis, editing review and publishing support and communication outreach and ethical compliance as six broad categories that AI could productively contribute to. Similarly, in the context of student essay writing, we considered the key stages of the writing process to be analysing the task, formulating a response, searching for information, analysing sources, planning, drafting the essay, editing and proofreading. 

By breaking down the essay writing process into parts, and identifying the function of each, we can evaluate the range of analog and digital collaborations that might occur at each stage to help a student meet the learning goals of the task. 

By breaking down the essay writing process into parts, and identifying the function of each, we can evaluate the range of analog and digital collaborations that might occur at each stage to help a student meet the learning goals of the task. 

Students already use search engines and library databases, reference management and mind mapping tools, word processing, and spelling and grammar checkers in the various steps of writing production. Similarly, AI-powered tools have increasing roles to play. We mapped out the stages of the essay writing process and where we might consider including AI tools and other intelligences in the process in the attached “steps and workflows” diagram.

For example, consider the similarities and differences between asking a teacher to read an essay and provide early feedback and asking for the same kind of feedback from increasingly powerful AI large language models (LLM). Both genAI and teacher feedback have different, yet possibly complementary, functions. A recent, cross-institutional survey of 7000 students’ experiences (including > 4000 Monash students) of AI and teacher feedback revealed that 50% of students surveyed have used genAI for feedback. Students responded that while teacher feedback was more reliable, contextualised and nuanced, genAI feedback was seen to be more easily and immediately accessible, objective and positive (for more on this, see Henderson et al., (2025)). 

Recipes for writing with AI: three approaches

We recognise that writing is a complex, iterative process which includes thinking, doing, reflecting and revising, sometimes individually and at other times in dialogue with other intelligences. As contemporary digital tools evolve, they continue to change how we generate and engage with writing. There are a range of ways to ‘cook up’ assessment tasks, and students might collaborate with different intelligences at each stage of the writing process. Therefore, we must clearly articulate the learning goals of our assessment tasks, and guide our students’ judicious, responsible and effective collaboration with different intelligences at each stage.

Comparable to steps in a recipe, we have formulated a simple set of guidelines for three approaches to writing with AI. The differences lie in how the writing approach starts and how the process unfolds; whether it’s ‘draft – assess inputs – revise’, or ‘optioneer – evaluate – synthesise’, or a combination of these. The ‘recipes’ identify the ‘ingredients’, or the various tools and possible roles that AI might play in the process. All recipes are licensed for sharing and non-commercial reuse under Creative Commons (CC-BY-NC-SA 4.0).

The first recipe [#1a] has the human starting with an opening gambit. The educator can guide students to start with a draft of portions of their text and then collaborate with AI tools, such as an LLM, to generate, analyse and implement (where appropriate) feedback on their drafts in an iterative process. Following responsible use principles, students are guided to clearly acknowledge AI contributions as part of the final submission. This approach can provide students with a range of individualised feedback points and learning throughout the writing process. 

The second recipe [#1b] follows a similar process but recognises that in our multicultural, multilingual context, English is not the only language that a student may feel comfortable thinking through, so how might they might be aided by AI translation and editing to help generate clear disciplinary appropriate expressions of their ideas in English? We propose integrating translation as an identifiable part of the process to free students to express their ideas without the constraints of language proficiency, and to support them as they develop discipline-specific academic writing skills in English. 

The third recipe [#2] is based on optioneering, guiding the students to prompt an AI to generate several versions of possible responses to an assessment task – then students compare, contrast and/or synthesise portions of these responses. This approach can seed students’ thinking and help them to frame the direction that they wish to take the task. Learning occurs through engagement and considering options while crafting texts. Across the recipes, students develop skills as authors, editors and co-authors.

illuminating the learning process

If you know what aspects of learning you want to observe in your students’ writing process, then you can calibrate the assessment task to meet these goals. 

We want our students to value the writing process, and avoid having them ‘just ask AI’ to do it for them. In contrast to outsourcing the essay product, the recipes we have developed reflect more robust and thickened processes with opportunities for multiple forms of engagement with learning. This approach also demands a shift in focus from the evidence of learning solely sitting with the product, to a recognition that learning is occurring through the engagement, evaluation and iterative formulations during the various stages of the writing process.

We also want our students to value working with AI and recognise the benefit to their learning of integrating inputs of multiple intelligences in carefully considered ways. We believe it is possible to guide our students to collaborate with other intelligences and use new tools whilst learning. Figuring out innovative ways of doing this requires careful consideration of how assessment design can help students to structure, support and demonstrate their multiple engagements with learning. 

So, although the recipe steps may look deceptively simple, each recipe has a preamble of key considerations to help with the planning of written assessments. 

As we thought about the various ways in which our educators might approach integrating AI into their written assessment tasks, a number of common issues emerged for assessment design and instructions to students, so we included them in a preamble to set the stage, provoking considerations of the roles, processes, tools and collaborations involved in academic writing. The considerations include: 

• Which stage of the assessment task are we focusing on? 
• Which tools should students be guided to use, and is there equitable access to these tools? 
• How much guidance or autonomy do your students need in their collaboration with the AI? 
• How do you want your students to document and evidence their collaborations with other intelligences? 
• How should students interrogate and evaluate the AI output? 
• How will AI engagement be graded? 
• What level of AI literacy do your students have, and do they have the foundational disciplinary knowledge and judgement skills to confidently and independently evaluate AI-generated outcomes and information? 
• What scaffolding do students need, given the place of the unit in the degree program? 

Over to you…

While we are valiantly trying to expand conversations and thinking about academic writing with AI, we recognise that change is hard and discussions and guidance are a way in as we all navigate these transformations in contemporary higher education. We hope the recipe steps can help you to think about the various steps in writing, and where AI may play roles in the process, whilst preserving the integrity of the learning that the task is designed to promote. We invite you to try the recipes in formulating written assessments and share your feedback on these evolving resources.

References

Chung, J., Henderson, M., Pepperell, N., Slade, C., Liang, Y. (2024). Student perspectives on AI in Higher Education: Student Survey. Student Perspectives on AI in Higher Education Project. https://doi.org/10.26180/27915930

Hounsell, D. (2005). Contrasting conceptions of essay-writing. In R. Marton, D. Hounsell, & N. Entwistle (Eds.), The experience of learning: Implications for teaching and studying in higher education (3rd ed., Internet version, pp. 106–125). University of Edinburgh, Centre for Teaching, Learning and Assessment.

Henderson, M., Bearman, M., Chung, J., Fawns, T., Buckingham Shum, S., Matthews, K. E., & de Mello Heredia, J. (2025). Comparing Generative AI and teacher feedback: student perceptions of usefulness and trustworthiness. Assessment & Evaluation in Higher Education, 1–16. https://doi.org/10.1080/02602938.2025.2502582

Khalifa, M., & Albadawy, M. (2024). Using artificial intelligence in academic writing and research: An essential productivity tool. Computer Methods and Programs in Biomedicine Update, 5, 100145. https://doi.org/10.1016/j.cmpbup.2024.100145

King’s College London. (2024). Manifesto for the essay in the age of AI. https://info.lse.ac.uk/staff/divisions/Eden-Centre/Assets-EC/Documents/AI-Manifesto-Sept-2024/24-0451-Manifesto-Ai-A5-v5-online.pdf

Syska, A. (2025, February 27). We tried to kill the essay – now let’s resurrect it – LSE Higher Education. LSE Higher Education – Enabling Dialogue and Sharing Different Perspectives in a Changing HE Landscape. https://blogs.lse.ac.uk/highereducation/2025/02/27/we-tried-to-kill-the-essay-now-lets-resurrect-it/