Neil Gordon, John Dixon, and Zhibao Mian from the University of Hull
Introduction
As educators in computer science, we constantly update our curriculum and assessments to match the rapid advancements in our field. We recognize the potential of computing technologies in higher education, and the advent of generative Artificial Intelligence (AI) promises a significant transformation in teaching and learning, though it also introduces new risks. Despite concerns leading some to restrict AI in academic settings, such measures are temporary, as both technology and pedagogy change limits e.g. utilising the fact that ChatGPT originally was only trained up until 2021. Instead, we should prepare students to leverage AI effectively, focusing on the human aspect of converting information into knowledge while maintaining quality and innovation.
Generative AI and Ethical Practice
We must also reflect on our own use of AI in tasks such as course development, content creation, grading, grant writing, and communication. Our approach to student guidance emphasizes ethical use of AI, recognizing its integration into everyday tools. This shift necessitates long-term strategies for acknowledging AI assistance in academic work.
Future Directions
Generative AI heralds a profound shift – quantum leap perhaps – in education. Traditional assessment methods, often reliant on deliverables like essays and projects, may no longer suffice as AI tools become more prevalent. We propose a shift toward evaluating students’ competencies in using AI as a routine tool and partner, focusing on the process and human-centric outcomes rather than just final products.
Current Approaches in Computer Science
In computer science, students have used AI tools like CoPilot for coding assistance for several years, making it challenging to distinguish between human and AI-generated work. Our research, funded by the Council of Professor and Heads of Computing (CPHC), aims to provide effective guidance on AI use and evaluate its impact on problem-solving and comprehension.
Moving Forward
Legislating or formalizing AI use is a temporary fix. As AI tools become standard in various tasks, we must focus on desired outcomes and their evaluation. Transparency in AI use is crucial, avoiding misuse and understanding AI’s capabilities and limitations. This involves rethinking assessments, emphasizing process scrutiny, self-reflection, and competency-based outcomes, potentially moving from small unit assessments to broader, synoptic evaluations across longer periods.
Understanding Student Perspectives
Understanding why students use AI and its emotional impact is essential. AI might be seen as a quick fix for poor comprehension due to disengagement, low-quality teaching, or limited staff availability, potentially causing feelings of inadequacy. Educators should guide students to use AI constructively, enhancing learning rather than undermining it.
Conclusions
We are entering an era where AI tools are easily accessible and capable of producing convincing solutions to traditional academic challenges. This shift should renew focus on critical thinking and reflective analysis. With AI enabling personalized education, we can transition from traditional teaching and assessment methods to more interactive and direct tutorial-style education, reminiscent of early university teaching models.
Note: this work was abridged and revised with the assistance of ChatGPT.
X (formerly Twitter): Follow @n_a_gordon for more information about Generative AI
Image: Attributed to “the good, the bad and the ugly faces of Artificial Intelligence” Designer https://www.bing.com/image. Content credentials Generated with AI ∙ 22 April 2024 at 11:57 am
