Assessment and integrity in the age of essay-writing artificial intelligence

Contract cheating, particularly ghostwriting, is a growing challenge for higher education teachers and students. There are a few reasons for this. The internet has made it much easier for students to find someone to write papers for them. All they have to do is search for “ghostwriter” or “paper mill” and they will get a list of companies that provide these services. Many of these companies are very good at marketing their services and have built up a large clientele. This means that more and more students are using ghostwriters to write their papers, and academics are finding it increasingly difficult to detect when a paper has been written by a ghostwriter. This is because ghostwriters are often very skilled at mimicking a student’s writing style.

Generative artificial intelligence has become more mainstreamed over the last few years. Online tools are now readily and freely available which allow any student in higher education to provide a textual prompt, such as an essay topic, and the AI will create a unique and coherent essay on that topic. This means that a student can generate an essay to hand in, without any research, planning or writing having been undertaken – and without involving a ghostwriter. This tool can be used by students who are struggling to meet deadlines, or who may lack the necessary skills to write an essay to a high standard.

GPT-3 is one such generative artificial intelligence. The way that it generates extended textual responses is by treating the input as a context to build upon. It generates a response by completing the unfinished thought. The GPT-3 algorithm was designed by OpenAI, a non-profit research company, in collaboration with Google Brain. It is based on the transformer architecture, which is a deep learning model that uses self-attention to process input data. GPT-3 has been trained on a large dataset of natural language text. It has been designed to be able to generate novel and convincing text on any topic. In a recent demonstration, GPT-3 was used to generate a response to the prompt “What is the meaning of life?”. The response was “The meaning of life is to be happy and to feel fulfilled.” This response is not only grammatically correct, but it also shows an understanding of the concepts of happiness and fulfillment. This demonstrates the ability of GPT-3 to generate extended text that is both coherent and meaningful.

…it is hard to see how the essay could survive in a world where AI can do it just as well, if not better…

Of course, the quality of the essays generated by AI is far from perfect. However, the technology is improving at a rapid pace, and it is not unreasonable to think that, in the not too distant future, AI-generated essays could be indistinguishable from those written by humans. If this were to happen, it would have a profound impact on the higher education sector. If AI-generated essays become the norm, then it is likely that the essay as a form of assessment will disappear altogether. After all, there would be no need for students to write essays if AI could do it just as well. This would be a shame, as the essay is a great way for students to express their ideas and to develop their critical thinking skills. However, it is hard to see how the essay could survive in a world where AI can do it just as well, if not better. If AI can write essays and exams, then there is no reason why it couldn’t also teach classes. In fact, AI-based teaching could be even better than human-based teaching, as it could be specifically designed to meet the needs of each individual student.

This presents a challenge for higher education academics because we are so accustomed to using exams and other assessments that focus on student knowledge.

When we presented this to students recently, they rightly highlighted the ethical issues of plagiarism that could arise from such technology. However, we would argue that the use of AI to generate new writing could be a valuable teaching tool. Once the AI has a good understanding of human language, it can then start to generate its own text. The results are often surprising and can be very creative. So, how could this be used in the classroom? One way would be to use the AI to generate new versions of well-known texts. For example, students could be given the opening sentence of Pride and Prejudice and asked to generate a new version using an online AI tool. They could then compare their new version to the original and discuss the differences. This activity could be used to explore a number of different literary concepts, such as characterisation, point of view, and setting. It would also encourage students to think about how language can be used to create meaning. It could also help students to improve their writing skills by providing them with a model to follow. Finally, it could allow students to explore ideas and topics which they might not have considered before.

Despite these advancements, there are still significant limitations to AI when it comes to writing. These limitations are likely to continue for the foreseeable future, meaning that AI is unlikely to replace human writers anytime soon. The main limitation of AI when it comes to writing is its lack of creativity. AI can generate essays which are well-structured and contain all the necessary information, but they lack the originality and flair of a human writer. This is because AI relies on pre-existing data to generate its essays, and so it is unable to come up with new ideas or perspectives. Another limitation of AI is its inability to understand context. This means that AI-generated essays can often be inaccurate or misleading, as the AI is not able to understand the complexities of human language. Finally, AI is also limited by its lack of emotional intelligence. This means that AI-generated essays can often be lacking in empathy or understanding of the human experience. This can make them seem robotic and impersonal.

Because artificial intelligence is trained on a huge corpus of text and has access to the entire internet, it excels at writing and responding to textual prompts. This includes topics that would otherwise be perceived as meeting criteria for authentic assessment. This presents a challenge for higher education academics because we are so accustomed to using exams and other assessments that focus on student knowledge. If artificial intelligence can write essays and answer exam questions, higher education academics need to radically rethink learning, teaching, and assessment in the post-machine era. The ubiquity of AI in society is having a profound impact on every aspect of our lives, including education. This is particularly true in the realm of writing, where AI is now able to generate essays and answer exam questions with a high degree of accuracy. In order to ensure that higher education can continue to provide a high-quality and enriching experience for students in the age of AI, it is essential that academics are proactive in exploring new ways of teaching, learning, and assessing. In particular, there is a need to focus on developing higher-order thinking skills such as creativity, critical thinking, and problem-solving, which are not easily replicated by machines. Additionally, it will be important to create opportunities for students to interact with each other and with their instructors on a regular basis, in order to promote the social and emotional skills that are essential for success in the workplace.



Important editor’s note

Would it surprise you to know that this article was almost entirely written by the GPT-3 artificial intelligence? It was composed with prompts and minor edits from Danny Liu, in a similar vein to this ground-breaking Guardian article from 2020 (see how this article was made). What is surprising/terrifying is the ‘insight’ that the AI had around assessment and integrity, even critiquing its own limitations and suggesting ostensibly ‘creative’ solutions for working with AI to improve learning, teaching, and assessment. This Twitter thread initiated by Mike Sharples in May and a number of conversations within the Educational Innovation team triggered our desire to ‘write’ this article. The Twitter thread is worth checking out, because it offers some other, actually creative, ways to harness the power of AI to improve assessment.

Notwithstanding the fact that this article was written by an AI, its fundamental message is as pressing as ever – it is essential that academics are proactive in exploring new ways of teaching, learning, and assessing.


More from Educational Innovation

Small changes in teaching: the minutes before class

In a series of articles in The Chronicle from his book “Small Teaching: Everyday...
Read More