Raphael Thys
Artificial intelligence (AI) is taking the world by storm. We explore the possibilities and limitations of AI tools like ChatGPT in education, and how you might come across them in your studies.
In November 2022, artificial intelligence (AI) tool ChatGPT ("Chat Generative Pre-Trained Transformer") was released by OpenAI, and excitement across mainstream and social media soon followed. But, not everyone was receptive to its release, with The NSW Department of Education and other districts moving swiftly to ban its use in their schools.
So, what does this mean for the use of ChatGPT and other AI tools at University? We uncover what ChatGPT is, what its limitations are, and how the University will balance the use of it and other AI tools while maintaining academic integrity.
What is ChatGPT?
ChatGPT generates convincingly human-sounding text and engages in realistic conversation. It currently generates its responses based on a large language model (find out how these work with Educational Innovation's informative video) trained on huge amounts of text from books and the internet. Similar to autocomplete, it predicts the next words in a response based on previous ones using the probability of words appearing after other words that it learned from its training.
For example, if you ask ChatGPT about the ways to limit the spread of airborne pathogens, it will promptly respond with paragraphs of text, which it uniquely generates for you, referencing masks, hygiene, and ventilation. It does this not because it understands your question, but because it has learnt that certain words like 'mask' and 'cleaning' and 'indoor spaces' are likely to appear together, after other words like 'limit the spread' and 'airborne pathogen'.
At the moment, ChatGPT is free for you to try. However, it does require you to sign up, agree to its privacy policy, and acknowledge that your use will form part of the AI's training.
What are ChatGPT's limitations?
While the content ChatGPT produces seems impressive on the surface, the reality is not all responses are worded well or are even correct. This becomes apparent if you ask ChatGPT highly conceptual questions or pose difficult or complex calculations.
On ChatGPT's FAQ page, OpenAI confirms that the tool's "outputs may be inaccurate, untruthful, and otherwise misleading at times".
Additionally, it's important to remember that the content ChatGPT produces reflects biases, opinions and views of the humans that informed the text it has been trained on, which may not align with your own values.
ChatGPT in the classroom
As AI technology and tools develop, you will see increasing use of them in your studies and assessments. While eventually AI will fundamentally shift the way we study, learn and work, don't expect any profound changes in the immediate future. For now, in your current studies, you can expect to experiment with, as well as discuss and critique AI's outputs.
Like any other resource you use in your studies and assessments, you should approach AI tools criticially, recognising their limitations in an honest and authentic manner.
In line with our academic honesty guidelines, you shouldn't use ChatGPT or other writing tools in an assessment unless this is explicitly allowed. Failure to acknowledge the use of AI tools in your studies could leave you vulnerable to accusations of cheating. If you are allowed to use these tools in your studies, you must acknowledge you have done so.
Expect to hear more from the University, your teachers and supervisors about how we're managing the opportunities and challenges AI present to higher education.
We're continuously working to safeguard academic integrity and the value of your degree, whilst helping to prepare you to be ethical leaders in a future where these technologies provide new and exciting possibilities.
As always, if you are unsure about whether or how to use a resource, ask your teacher, supervisor or tutor for guidance.
To find out more about what ChatGPT means for you and your studies, come along to one of our student-staff forums: