🏫 Comparing ChatGPT, Gemini, and Copilot for Mechanical Engineering Education
This study evaluated three generative AI tools (ChatGPT, Gemini, Copilot) in undergraduate mechanical engineering education, focusing on their accuracy, student perceptions, and strategies for integration. The AI tools showed high accuracy with theoretical questions, but struggled significantly with numerical and complex problems, with Copilot performing best overall. Students generally perceived these tools positively but expressed concerns about frequent inaccuracies and difficulty verifying responses.
🧠 Using Gen AI with Metacognitive Scaffolding
Structured scaffolding can help students engage more deeply when using ChatGPT in a university course. Researchers in this study found significant improvements in deep learning approaches, cognitive complexity, and project outcomes for students who received scaffolding. Consider leading students through planning, monitoring, and reflection prompts to introduce using AI for learning.
🤖 AI Ethics: Growing Practice, Glaring Gaps
After reviewing 20 AI ethics auditing frameworks, the writers conclude that ethics audits often imitate financial auditing structures and fall short on stakeholder involvement, success metrics, and transparency. Who should be involved in these audits on the university campus? What should they look like?
💬 Word of the Day: Postplagiarism
Postplagiarism is a framework that reimagines academic integrity in the age of Gen AI. Student accept that human-AI hybridity is becoming normal, but are still grappling with what that means to originality and creativity. Faculty should focus on these grey areas.