In this article, the author, UTM Lecturer Mr. Joe Lam, argues that it is important for scholars to integrate artificial intelligence knowledge in their lectures and coursework. This integration is essential to prepare students for an evolving future job market where proficiency in AI will be increasingly valuable.
Artificial Intelligence (AI) transforms various aspects of the workplace by automating tasks such as summarising extensive reports and writing code. While there are concerns about AI challenging human capabilities and possibly leading to job losses, the integration of AI at work is also seen as a revolutionary development (World Economic Forum, 2023).
As noted in Harvard Business Review (2023), history shows that technological revolutions often create new job opportunities for those who adapt. Similarly, an MIT Technology Review (2024) article emphasised that learning how to utilise innovative technologies opens doors to new career paths. By introducing AI to regular educational timetable, we could improve students’ employability and competitiveness globally.
At the forefront of AI technologies are Large Language Models (LLM), which have become widely accessible since 2023. An LLM is a type of artificial intelligence model that is pre-trained (trained before release) on a vast amount of data, and it can understand human-like natural languages. LLM is also a type of generative AI (Gen AI) because of its capability to generate contents. Prominent examples include ChatGPT, Claude, and Llama.
This semester, efforts were made to integrate Gen AI into group project presentations with my Year 2 students at UTM. Each student group was given an option to participate in this experimental method. As part of the requirements, each group must conceptualize and develop a chatbot designated as a virtual group member. Groups must select a base chatbot model from a variety of available Large Language Models (LLMs) and enhance its output with the information related to their projects.
Since most LLMs are pre-trained, the knowledge of each LLM has a limited time range, usually known as the “cutoff date” when the pre-train ends. For instance, the cutoff date of ChatGPT-3.5 is January 2022. Technically speaking, ChatGPT-3.5 does not include any knowledge that occurred beyond this date. In addition, privately owned information will not be included too.
Nevertheless, fine-tuning processes such as the Retrieval-Augmented Generation method, or RAG, could enhance pre-trained LLMs. RAG allows students to insert additional files about their projects into the chatbot model so that the chatbot is equipped with updated knowledge base. This fine-tuning process involved adding project details, group data, and other relevant information to the chatbot, thereby turning each chatbot into a “real” group member who can answer relevant questions for each group accordingly.
While group members prepared the chatbots, they remained involved during the presentation. Human members must intervene immediately if the virtual member makes mistakes or exhibits hallucination, a phenomenon where LLMs generate incorrect or nonsensical outputs. By using fine-tuning methods such as RAG and managing AI errors, students gained first-hand experience in harnessing the power and understanding the limits of generative AI. Despite the speed and efficiency of LLMs, human cognitive abilities remain unique and irreplaceable in the overall process.
Anonymous surveys were distributed as part of this initiative, and 58 completed questionnaires were collected. Feedback from students was positive. One student commented: “This is a completely new presentation method, so I think it’s very interesting.”
More than 90 percent of students believed that this new presentation method helped them understand the coursework better. Meanwhile, when asked about the usefulness of chatbot creation experience, more than 80 percent expressed interest in future exploring this technology for studies and future jobs.
As we continue to explore the potential and boundaries of AI, we should remember that AI technologies are still in their initial stages of development. By staying informed and engaged with the advances in AI, students can prepare themselves to thrive in the future workplace and seize new opportunities.
Author: Mr. Lam, Hou Peng (Joe), UTM Lecturer
Editor: UTM Public Relations Team
References:
World Economic Forum. (2023, August 14). AI: 3 ways artificial intelligence is changing the future of work.
https://www.weforum.org/agenda/2023/08/ai-artificial-intelligence-changing-the-future-of-work-jobs/
Harvard Business Review. (2023, April 4). AI Won’t Replace Humans — But Humans With AI Will Replace Humans Without AI.
https://hbr.org/2023/08/ai-wont-replace-humans-but-humans-with-ai-will-replace-humans-without-ai
MIT Technology Review. (2024, January 27). People are worried that AI will take everyone’s jobs. We’ve been here before.
https://www.technologyreview.com/2024/01/27/1087041/technological-unemployment-elon-musk-jobs-ai/
