ChatGPT is an advanced language model powered by AI and developed by OpenAI, a US-based company. It has been trained on a diverse range of internet text, enabling it to generate responses that are similar to those of a human being to a wide array of questions and prompts. Since its introduction in late 2022, it has become a widely discussed topic, not only amongst educators and educational organisations, but also in academic circles. This is largely due to concerns about the potential for misuse by students and researchers and the risks it may pose to academic and research integrity.
ChatGPT can be a highly valuable tool for individuals seeking to enhance their learning in a legitimate manner. It offers comprehensive guidance and instruction on a wide range of topics including general writing, language correction and editing, text summarisation, coding, job applications and interviews, essay writing, and idea generation, among others. To demonstrate its capabilities, here is an example of a text generated by ChatGPT. When I asked the chatbot to write a poem about “my mom”, it generated a beautiful four-verse poem in less than five seconds, as seen below.
My mother, oh so kind and true A love that always shines through She’s always there with a gentle touch And a warm and loving smile so much Her heart is filled with endless grace And her laughter lights up any space She guides me through life's stormy weather And makes everything feel just a little bit better She's my rock, my inspiration A constant source of motivation Her strength and wisdom know no bounds And her love for me knows no bounds So here’s to my dear mother, with love and pride For being there for me, by my side Forever grateful for all she's done She’s the best mother under the sun!
I may not be an English teacher, but I find the poem to be elegantly composed. It effectively conveys the love and sacrifices a mother would make for her child, and the expressions and emotions depicted are quite authentic. Simply amazing!
To further test ChatGPT’s capabilities, I asked it to critique the poem it had just written for me. The response, seen below, was remarkably human-like.
The poem you wrote is a touching tribute to your mother and is filled with affection and admiration. The language is simple and easy to understand, making it accessible to a wide audience. The rhyme scheme adds a musical quality to the poem and helps to reinforce the message. One aspect that could be improved is to make the imagery more vivid and descriptive. For example, you could add more sensory details to bring the poem to life, such as describing the sound of your mother's laughter or the way her smile lights up a room. Overall, it’s a sweet and heartfelt tribute to your mother and will certainly bring a smile to her face.
The critique generated by ChatGPT was based on a simple prompt: “critique the poem that you just wrote”. Yet, in certain aspects, it surpasses what one might expect from a first-year literature student. With prompt engineering – including additional and specific information in the prompt, the response can be significantly improved. This capability of ChatGPT poses some threats to academic integrity in take-home assessments if misused by students, which has been a key concern raised by teachers and academic leaders recently.
Indeed, ChatGTP is not the only language model in this space, nor the last. Tech giants will always strive to improve their technological innovations and existing tools. Google has already unveiled Bard – a similar language model to ChatGPT, promising advanced features to produce more sophisticated responses with simpler prompts. This highlights the ongoing and never-ending race in the field of language models and impacts it may have on teaching and learning.
The implications of this technology are evident. Students in K-12 schools in several Australian states, including New South Wales and Queensland, have been prohibited from using ChatGPT. Within the higher education sector, there have been significant debates on how universities can mitigate the negative consequences of AI-powered language models like ChatGPT. There has been contention over whether Australian universities should abolish take-home assignments and revert to in-class evaluations. The Australian Tertiary Education Quality and Standards Agency (TEQSA) has urged universities to ensure that their academic integrity policies are robust enough to address the implications of AI. In the realm of research, the inclusion of ChatGPT as a co-author has already been observed and presents some complex issues, as it raises questions about the accountability for any inaccuracies in research content generated by the chatbot. Some academic journals have acted promptly and altered their publication guidelines, prohibiting authors from listing ChatGPT as a co-author and requiring them to declare any assistance received from AI-based tools in the creation of manuscripts.
The recent discussions and apprehensions surrounding ChatGPT are well-founded. However, the key issue at hand is whether forbidding students from using it is a viable solution. My answer is a resounding no. Such a reaction would be short-sighted and fail to address the bigger picture. Technology companies will continue to push the boundaries of innovation, and as a result, future AI-powered language models will only become more sophisticated. Banning students from using ChatGPT (and similar tools) would be akin to denying them accessing to resources like Wikipedia or the spelling and grammar checker in Microsoft Word. Students will find ways to circumvent the ban, making enforcement a significant challenge for teachers and educational institutions.
So, what is the answer? In my opinion, a more reasonable approach would be to redesign assessment tasks to promote deeper thinking. While ChatGPT and other AI-powered language models know almost anything and everything on the internet, they do not have the ability to think (at least for now). Therefore, they really cannot do what the educational philosopher and psychologist John Dewey, called reflective thinking – “active, persistent and careful consideration of any belief or supposed form of knowledge in the light of the grounds that support it.” As educators, our goal is to encourage students to think critically, rather than rely solely on memorisation. Why ask students to memorise the names of the 15 largest rainforests in the world or the birth year of a historical figure, when they can easily find this information online? With the ubiquitous nature of the internet and technology, such rote knowledge is becoming less valued. Instead, we should rethink the design of our assessments to encourage authentic and critical thinking, and focus on the process, not just the final product. Developing a well-written essay requires significant cognitive heavy lifting which can be evaluated through authentic assessments. This approach would also reduce the desires and likelihood of students seeking unauthorised academic assistance. So, our aim should be to utilise technology as a support tool and use AI in a way that enhances deep thinking and reflection rather than prohibiting students from using it.