Parakram Pyakurel, Associate Professor at NMITE, briefly outlines how AI could be better utilised by students and instructors. This post also explores guidelines for developing AI policies.
Artificial Intelligence (AI) is having a significant impact on the education sector as we grapple with how to effectively respond to this rapidly evolving technology. Below are some ways that AI, particularly Generative AI, may be productively utilised by students, teachers and academic institutions.
For students
Students should first and foremost understand that AI produced contents can sometimes be inaccurate and misleading. AI tools are trained on huge datasets and the outputs they produce reflect any inaccuracies and biases that may have been present in the training datasets. Even when training datasets are accurate, AI generated responses can sometimes be factually incorrect or irrelevant due to Large Language Model (LLM) hallucinations. Therefore, it may be worthwhile to crosscheck the accuracy of AI responses from other reliable sources.
Generative AI tools can be good at simplifying complex topics without using too much jargon, which can be very helpful for students trying to learn complex subjects. Attempting to first learn a complex topic from AI and then moving on to academic textbooks or sources could greatly help students learning journeys. The quality of AI response depends on the quality of input queries or statements (prompts), therefore, AI tools also provide students with opportunities to develop their skills of asking precise, succinct and clear questions.
Students can utilise AI to summarise long texts and gather information from various sources. This could be a good starting point to help break down long and complex topics into manageable learning chunks that are less difficult to understand. AI tools can also provide step-by-step explanations on topics that students are struggling to understand. Likewise, students could benefit from AI generated personalised learning support based on individual strengths and weaknesses.
For instructors
The UK government is supporting AI companies to train their tools to generate accurate and high quality academic contents. Given rapid improvements in AI tools, teachers can utilise AI to help prepare study materials such as notes, presentations, handouts, case studies, quizzes, exercises and so on. AI powered interactive learning tools such as chatbots and virtual assistants can also support students outside regular office hours. As AI integrates in education, careful and balanced approach is needed. Overreliance on AI can hinder students from adequately developing soft skills and other essential skills like critical thinking, problem-solving, and cognitive abilities. Easy access to information can impede training of cognitive faculties for knowledge construction. For example, overuse of AI platforms to easily access information can hamper development of independent thoughts. To ensure that students do not just fully rely on AI, they should be encouraged to critique AI responses, and more dialectic methods (e.g. debates, problem-based learning etc.) may have to be embedded in education delivery.
There is not enough research on the professional development required for teachers to effectively and appropriately utilise AI. Upskilling and development needs of teachers and any gaps they may have in AI literacy should be addressed.
As cheating using AI has become a big issue, education institutions need to explore innovative ways to assess students. Assessments could be designed to imitate real life challenges in order to discourage students from submitting solely AI generated responses. Assessments that require applying personal experiences, context specific responses and reflections are also less prone to cheating using AI.
AI policy
Academic institutions should have adaptive AI policies because of the rapidly evolving nature of this technology. Privacy and data protection issues increase as AI gets more embedded into digital infrastructure. AI tools could compromise privacy of students and staff, demanding a careful balance between innovation and privacy.
For students, institutional policies should enable equal access to AI technologies to mitigate potential digital divide. An example could be giving all students access to AI tools like Grammarly or ChatGPT. Students should also receive clear guidelines on what counts as appropriate use of AI for both learning and assessments. For teachers, policies should promote innovative AI use along with support for training and development for effective AI integration in teaching and learning. Assessment policies should diversify assessments and enable alternative assessments in cases where cheating is suspected.
Just like with any promising technology, there are risks, opportunities and hypes related to AI. Tempered discussions are needed as AI tools are complex with both advantages and disadvantages when integrated in education systems.