Generative AI’s Impact on Skills may lift or crumble academic integrity – Monash University

Increased awareness of Generative AI’s advantages has also posed a threat to skills development across the globe as observed by Monash University. Adapting to innovation is essential to academic institutions yet academic integrity is being challenged as generative AI such as ChatGPT and similar applications become more convenient to learners and educators.

In these early stages of the introduction of generative AI, acknowledging the challenges and rolling out an effective approach to how AI tools are used in students’ skills development and educators’ output assessment could better cultivate good education and a stronger foundation in the academe.

The threats of AI

Artificial Intelligence has the ability to produce content that’s superficially plausible but entirely incorrect. These outputs can thus leave users with a deeply mistaken picture of the world. Contrary to appearances, ChatGPT is not “trying” to assert facts about the world. Instead, it is (successfully) performing a different task – that of generating superficially plausible or convincing responses to a prompt.

Students can submit AI-generated school work without writing a single word themselves – or necessarily understanding the material they’re supposed to be tested on. The reliance on these AI tools will result in the erosion of important skills. If the writing process is outsourced to artificial intelligence, it’s a skill students might lose, not to mention comprehension and critical thinking skills are essential to remain competitive in the workforce. On the other hand, Generative AI will also become capable of tasks that many educators think of as central to their own profession. A non-profit education organization Khan Academy recently announced it will be launching a GPT-4 powered AI assistant, Khanmigo, which can play the role of a tutor to students and can also assist teachers with tasks such as creating lesson plans.

The Philippine government acknowledged that Filipino learners lack 21st-century skills such as critical thinking, problem-solving, and communication skills among others. According to the Global Skills Report 2023 by Coursera, participating Filipino learners got 17 percent in communication skills, 7 percent in data science, especially in data analysis skills, and 7 percent in mathematical skills.[1] The report shows that Filipinos lag behind their neighboring countries in terms of skills in business communication, mathematics, and data science, thus signaling a stronger call for academies and the government to invest and roll out effective approaches to AI-infused challenges.

Addressing the Challenges

ChatGPT is a valuable tool and worries regarding misinformation are valid. However, these issues can be best addressed by educating students on how to utilize these tools, recognize their limitations, and fact-check their output. Universities should be teaching students how to read critically, assess or corroborate facts, and discriminate between good and bad arguments through developing specific assessment tasks where students generate, analyze, and criticize AI outputs.

Designing assessments that engage with, and leverage, students' interests could motivate learners to remain engaged such that they don’t see value in outsourcing the pursuit of their own knowledge to AI. This strategy may undercut some students’ desire to turn to AI, but it’s unlikely to fully dissolve the temptation to save time by relying on an AI particularly as it would not lower and might even raise the final grade.

Another strategy is to design AI-resistant written assessments where students are required to demonstrate their own understanding (independently of the written work they’ve submitted) or combine written essays with the presentation and discussion of their work during class time. This is modeled after the format of a viva presentation or thesis defense where students are required to understand and be able to communicate the ideas they’ve defended in their essay – regardless of whether they outsource any of the writing to an algorithm.

In an experiment by Monash University, they found that ChatGPT can generate convincing responses about major works in various disciplines. However, it fares very poorly when asked about the cutting edge of scholarly debate and is prone to “hallucinate” sources that don’t exist. Monash University recommends that educators, consider how to adapt their assessment regime in light of ChatGPT to ask themselves the following questions:

The immediate challenge for educators is to determine what an AI-literate skill set looks like, and how to evaluate whether students have these skills. If the academe can get this right, perhaps teaching and assessing the core skills of research, reflection, argument, and critical thinking that have traditionally sat at the heart of the university would lift academic integrity.

Monash University continues to adapt to the required innovations in order to maintain its educators’ and students' global competitiveness. Read more about Monash University on their website.