Given the rapid progress that generative artificial intelligence (GAI) has made over the past two years, the realm of education – and particularly the higher education sector – now finds itself at a critical juncture. The implementation of GAI by learners presents promising opportunities for personalised instruction and increased pedagogical effectiveness; however, it also gives rise to substantial ethical and pragmatic concerns. In particular, the function of homework assignments is called into question by AI. It is abundantly clear that the academic community is not yet prepared for the AI ‘revolution’ in this respect. Many parents and educators remain uninformed about the true potential of these tools (especially as the media frequently oversimplifies the topic) and statistical data on the practical implementation of generative AI among individuals aged nine to 25 is lacking. Consequently, it becomes challenging to make quantitative predictions regarding future developments. can read, summarise, pose questions and even instigate a debate based on any text. This is somehow the reality that we are faced with nowadays. With astonishing accuracy, GPT4 can break down a text, offering an analysis to rival those of our most diligent students. And the most disconcerting part? It’s become almost impossible to distinguish an AI-generated analysis from authentic human reflection. G ARTIFICIAL INTELLIGENCE How AI might hamper homework Is AI capable of eliminating homework? Certainly not; with AI comes the possibility that students will cheat. This is because they will attempt to optimise their grades by minimising the amount of time devoted to work, particularly if they consider the assignment to be devoid of significance. This pursuit is motivated by the desire for immediate gain: earning a passing grade with little effort. Given that 90 per cent of students aged from 12-17 have access to a computer at home and the average age of first-time mobile phone owners is nine years and nine months, this trend is likely to begin early. In addition to the requisite hardware, AI tools are readily available and frequently provided at no cost for initial utilisation. With such unrestricted availability and an established background of academic dishonesty, I hold the intentionally sanguine belief that, in the end, everyone will have access to assignment assistance as a form of ‘social justice’ that is exceedingly comprehensive. In fact, this is the goal of the US-based, non-profit educational organisation the Khan Academy: rather than requiring sophisticated computing, a basic mobile phone is sufficient to access generative AI of a quality good enough to provide solutions to homework problems. This is a fascinating concept, so long as information regarding the feasibility of utilising AI for assignments is properly disseminated, with a particular focus on its limitations given that copy-pasting continues to be especially hazardous due to the inherent biases and hallucinations of these tools. Moreover, children already incorporate AI tools such as Canva into their daily routine (it is widely used from middle school onwards) and they will continue to do so in their endeavours; interacting and learning in collaboration with AI will soon be the norm. This will prompt children to pose this question to their educators: why are they being given assignments that appear superfluous in light of artificial intelligence? The fundamental rationale behind allocating such work will be called into doubt. A catastrophe for comprehension Apart from the issue of homework, another AI-related topic in the education sector is the matter of engaging students with reading material. Whether it’s deciphering reports, condensing chapters or debating articles, the core of these tasks is to immerse students in text, inviting them to interact with words and ideas. However, AI is disrupting this rewarding activity. It devours PDFs, swallows entire books and spits out summaries with disconcerting ease. The temptation is real: why wouldn’t a student ask AI to do the heavy lifting? Sure, AI may stumble and oversimplify, but these quick and convenient summaries can shape, or even replace, a student’s thinking. The danger is that you end up with a classroom filled with AI echoes, depriving discussions of their richness and spontaneity. Reading and text comprehension, fundamental pillars of education, are undergoing a silent revolution. Historically, assignments were designed to assess students’ ability to absorb, analyse and reflect on content. But today, with the advent of large language models (LLMs), the authenticity of this understanding is being challenged. Imagine a world where ChatGPT4, Google Gemini or Claude
Ambition | MAY 2024 | 25
Made with FlippingBook - Share PDF online