Library Guides
- Library
- Library Guides
- AI for T&L in Education
- Using AI for academic work
Using AI for writing research papers
Researchers have been using AI for decades for complex calculations, data analysis and modelling. AI writing aids have been around for decades (remember "Clippy", Microsoft's anthropomorphic paper clip?) Generative text AIs available via subscription have been available for the public since at least January 2021. However, it was the release of the freely available research version of ChatGPT from OpenAI that saw the explosion in the use of generative text AIs and has prompted the development of policies in the research environment that are specific to its use.
Publishers such as Elsevier, Springer Nature, Cambridge Press, Taylor & Francis and others all have AI use policies related to the many issues around integrity, authorship, originality and the methodology and citation of AI use.
Elsevier's AI use statement format may be one academics find useful if they choose to allow students to use AI assistance in their work:
"During the preparation of this work, the author(s) used [NAME TOOL / SERVICE] in order to [REASON]. After using this tool/service, the author(s) reviewed and edited the content as needed and take(s) full responsibility for the content of the publication."
However, it's worth noting that many journals explicitly do not allow the use of AI-generated images and multimedia unless specifically part of the research itself, due in large part to the rapidly evolving issues around copyright.
When it comes to citing the specific use of AI in a given work, citation frameworks have either updated their guidelines or, as APA Style did in April 2023, supplemented their frameworks with how-to articles regarding how AI use should be cited.
Using AI for conducting research
When it comes to the use of AIs for research, governing bodies and professional associations recommend balancing the efficiencies to be gained from using AI tools with the aforementioned considerations of ethics/transparency, IP and critical evaluations of the outputs.
The Australian Research Council (ARC) has a specific policy to address the use of Generative Artificial Intelligence in ARC grant programs. In that policy, the ARC acknowledges that the use of generative AIs can be beneficial but must be balanced with the inherent risks of using a third-party tool, such as security and risk of exposing IP in addition to the ethical issues related to the extent to which AI is used for authorship.
Developments are moving forward in the integration of AIs into academic databases to summarise papers and make connections between them. Elsevier is building custom-engineered generative AI into its Scopus abstract and citation database and began beta testing in August 2023. The subscription-based, qualitative citation analysis tool Scite.ai helps science researchers make connections between papers. Elicit is described as a tool to support the literature review workflow and provides context from the bodies of articles based on natural language inputs. However, even a generative tool trained on quality content can hallucinate (produce erroneous data), as Elsevier cautions in their disclaimers on the Scopus AI beta page.
In their submission to the inquiry into the use of generative AI in the Australian education system, Universities Australia urged researchers and research supervisors to strike a balance between harnessing the capabilities of AI to improve research efficiency, being mindful of copyright and privacy issues and the development of "the general capabilities and skill that foster their professional expertise, critical thinking, evaluation and intellectual curiosity".
TEQSA has published 10 tips for using generative AI in research (PDF, 85 KB) that provides high level guidance for researchers that you may wish to share with your students.
Using AI for content and activity creation
If you are new to using AI to support your work, we'd recommend you first look at the section on AI Skills and Literacies under the topic of Preparing Students for an AI Future for guidance on working with AI and how to prompt AIs to get the output you want.
AI can be a useful tool for educators to plan and manage their courses as well as to assist with content vetting and creation and to check your assessments to see if and how well AI can complete the work. New AI tools and new abilities in existing AI tools are developing at a rapid pace. Use your trusted professional networks to seek advice on useful AI tools.
Below are some examples of how AI can support you in your work.
Depending on the level of what you are planning, you'll first need to assemble the relevant information you'll need to provide a context for your planning documents.
At the course level, this could include the course learning outomes, TEQSA level, number of weeks, number of assessments, pedagogical approach, cohort size, number of teachers, whether there are pre-requisites or this course is a pre-requisite, whether the course is a core university or core program course, what skills and competencies you feel you can expect from your learners, if it is an elective, how it is being delivered (online, face to face, blended), if it is a course to prepare students for a practicum, the timeline for the course, timings of key activities and assessments, feedback mechanisms, the key skills you'd like to scaffold students up for and over what time period, any soft skills that aren't specifically mentioned in the course learning outcomes, any hardware or software the students are likely to have access to, any key concepts they must master at which points in order to succeed, any academic skills needed, diversity and access issues you want considered, etc.
Class planning could also include a description of the space and delivery mode, the topics and subtopics being covered, the number of students, timings, types of activities you want to include, pedagogical approaches, etc.
Use this information to craft your prompts and to hone your results.
Remember, you can also direct the AI tool you are using to generate multiple versions of the same request, then focus on the one you like and work to refine that output.
Collaborating with AI on content and using it to outsource time-consuming tasks can help boost your productivity and create customised, relevant content for your courses.
As well, search is starting to incorporate AI and vice versa. For instance, the free Bing search engine provides for AI chat (click on Chat after going to the Bing site) and OpenAI's subscription version of ChatGPT 4 provides for search via Bing. Your prompting skills will be important to ensure the content you get back is relevant, timely, credible and suitable to your context.
For publicly available video and audio sources with transcripts, you can use AI to summarise the main points to determine whether or not you wish to watch them all the way through for use in your course. Simply copy and paste the transcripts into an AI tool and ask the tool to summarise them. If you like what you see, copy the summary as well as the video or audio source name and URL into a spreadsheet for later review.
Provide all the relevant context of the assessment as well as the length, format, and learning objectives. Once you are happy with the result, copy and paste it into a Word document directly from the chat window. Then, ask the AI to regenerate the result as HTML. Copy and paste the HTML into the code view of the text editing area of your course site.
These two activities can take up a lot of time for a busy academic. Depending on your subject area and the likelihood of content related to it existing on the public internet to which you'd have access, generative AIs can provide you with content to edit and hone to suit your needs.
This video from the Wharton School's Eli and Lilach Mollick describes a base prompt that allows you to collaborate with AI on these content types as well as how to use AI to create quiz questions and other content types.
Providing a text explanation for images, charts and diagrams meant to convey information helps students with vision impairments as well as those who are non-native English speakers or those unfamiliar with how to read certain types of charts and diagrams. Attach an image to an AI such as Bing and ask for an explanation. Edit and use the explanation in your course, for your presentation notes or paper, always remembering citation requirements and ethical use guidelines.
If your students want MCQ drill activities to help them to self-test their knowledge of foundational knowledge such as that found on public internet sites, you can prompt a generative AI with the year level, discipline and topic of your course module, then ask it to generate questions and answers.
These are just a few of the many ways in which academics can use AI to support their work. Remember that collaboration with and review from competent peers is as important as keeping competent in the loop during the design phase.