By Dr. Sanjay Goel, Director, Institute of Engineering and Technology, JK Lakshmipat University, Jaipur
AI is transforming education by offering customised learning, coaching, and content-generation support to students. AI platforms like ChatGPT can help students improve their skills by giving instant feedback and creating personalised practice problems. AI can also help teachers in lesson planning, gamification, and automated assessment. However, AI tools in education present challenges such as personalising instruction while considering social and emotional aspects, avoiding bias and discrimination, lacking evaluation and standardisation, privacy and security concerns, ethical considerations, and prohibitive costs. The lack of meaningful use cases and examples of using AI in education can also hinder the effective use of such tools.
Engineering education has traditionally put a great emphasis on activities like problem-solving, laboratory work, assignments, projects, etc. With a focus on outcome-based education (OBE), these and other such activities have become even more important in engineering education. The National Academy of Engineers (NAE) suggested that the essence of engineering—the iterative process of designing, predicting performance, building, and testing—should be taught from the earliest stages of the curriculum. In deductive methods of teaching, these activities usually follow conventional lectures and self-study. However, in inductive teaching methods like project-based learning (PBL), many of these activities are not seen as the culmination of learning, but as the process through which learning takes place.
During these activities, engineering students engage in a variety of sub-activities and tasks like self-study, group study, ideation, brainstorming, field survey, literature survey, inspection, experimentation, data analysis, mathematical analysis, systems analysis, proposal making, conceptual modeling, mathematical modeling, simulation, high-level design, low-level design, prototyping, test planning, hardware implementation, programming, debugging, system integration, verification, validation, testing, report writing, documentation, presentations, reviews, etc.
Large Language Models in Engineering Education
Future engineers have to develop mastery in the use of modern technologies like AI for driving quality, innovation, and productivity in their work. Further, to create and support the fascinating world of engineering structures, machines, and systems, engineers have to engage with and generate a large volume of documents. Hence, large language model based generative AI platforms like ChatGPT and Bing Chat can be a great help for engineers and engineering students. Unlike ChatGPT, Bing Chat also leverages web search and cites its sources. Recently released GPT 4.0, a large-scale multi-model model can also analyse images. Future GPT versions may even analyse engineering drawings.
Platforms like ChatGPT and Bing Chat can help engineering students by playing supporting roles like search assistant, proofreader, reviewer, technical assistant, creative partner, coach, etc. Students can use these platforms to search, consolidate, organise, and re-organise ideas and facts. These platforms can provide templates and examples for different types of documents, offer feedback, recommend ways to improve the text, prepare reference lists and citations, find and synthesise relevant sources, help in qualitative analysis, assist in programming, and support revision, editing, and refinement. Through properly designed prompts, these platforms can assist students to perform a large number of engineering and learning activities, sub-activities, and tasks including design, mentioned in the above section on Outcome-based engineering education. The platforms can provide immediate feedback and generate practice problems to help students develop their skills. They can promote creativity by exposing students to diverse ideas, perspectives, and counter-arguments on a given issue, and can help generate ideas. In addition, the platforms can help students improve their critical thinking skills by reviewing the text through a variety of questions related to clarification, purpose, assumptions, evidence, relevance, interpretation, consequences, significance, accuracy, precision, consistency, completeness, and point of view. The ability of ChatGPT to summarise, shorten, elaborate, rephrase, and translate text can also help students improve their writing skills.
These platforms can help the engineering faculty in many ways. They can be used to serve in the roles of teaching assistant, administrative assistant, creative partner, external evaluator, peer reviewer for internal quality assurance, domain-specific advisor to enrich and contextualise the course content, etc. They can be used to support instruction design using either deductive or inductive approaches. They can assist teachers to create simplified, engaging, and diverse explanations. Faculty can use these platforms to generate examples, analogies, questions, and interconnections between different topics within and across courses for the enrichment of their own instruction. These platforms can generate practice tests, quizzes, and personalised assignments about a topic. With these platforms, faculty can review student submissions, and analyse student responses and feedback.
Conclusion
The ability to leverage AI tools will be a crucial competence for future engineers in all disciplines, enabling them to drive high quality, innovation, sustainability, and productivity in their work. AI is also transforming education, including engineering education. Throughout every stage of their work, engineers generate and utilise a vast quantity of documents. Hence, engineering students and faculty can use large-scale language models like ChatGPT and Bing Chat to improve their productivity, creativity, and quality. Both can use these to explore more, contextualize better and integrate diverse ideas. Intellectual challenges are must for enabling deeper learning. Hence, for designing the new generation of learning engagements and experiences through mindful usage of these tools, faculty must ensure that while the nature of a student’s cognitive work may change, the overall intellectual challenge is not diluted and the learning experience becomes more rewarding and joyful. At the same time, effective use of AI in education requires a total transformation in learning outcomes, pedagogy, and assessment methods as well as addressing some well-known AI challenges, e.g., bias, discrimination, privacy, security, ethics, and costs.