Skip to main content Skip to navigation

Learning to Prompt in the Classroom to Understand AI Limits: A pilot study

Project Overview

The document explores the integration of generative AI, particularly Large Language Models (LLMs) like ChatGPT, into educational frameworks, emphasizing a pilot study in a high school designed to improve students' comprehension of AI's potential and constraints. Through engaging with ChatGPT, students exhibited reduced negative perceptions of AI, enhanced interaction quality, and developed critical thinking skills regarding AI technologies. The study underscores the importance of AI literacy interventions in fostering a more positive and informed approach to AI among learners. Nevertheless, it also highlights significant challenges, including ethical considerations and the necessity for effective prompting strategies to optimize the use of AI in educational contexts. Overall, the findings suggest that with appropriate guidance and resources, generative AI can play a transformative role in enhancing educational experiences while encouraging critical engagement with emerging technologies.

Key Applications

ChatGPT in classroom activities

Context: High school setting with students aged 16 to 18

Implementation: A two-hour workshop where students learned about AI and interacted with ChatGPT, employing prompting strategies to produce educational content.

Outcomes: Improved interaction quality with the LLM, reduced negative sentiments about AI, and a better grasp of AI's limitations.

Challenges: Ethical concerns regarding academic integrity, such as plagiarism, and the need for effective prompting strategies.

Implementation Barriers

Ethical

Concerns about academic integrity and the potential for students to misuse AI-generated content.

Proposed Solutions: Incorporate discussions about the ethical use of AI and promote proper attribution and engagement with AI tools.

Technical

Technical issues and accessibility challenges in using ChatGPT during the educational intervention.

Proposed Solutions: Ensure reliable access to AI tools and provide technical support for students during workshops.

Perceptual

Misunderstandings about the capabilities and limitations of generative AI like ChatGPT, leading to unrealistic expectations.

Proposed Solutions: Implement comprehensive AI literacy programs to educate students about the realistic capabilities and constraints of AI.

Project Team

Emily Theophilou

Researcher

Cansu Koyuturk

Researcher

Mona Yavari

Researcher

Sathya Bursic

Researcher

Gregor Donabauer

Researcher

Alessia Telari

Researcher

Alessia Testa

Researcher

Raffaele Boiano

Researcher

Davinia Hernandez-Leo

Researcher

Martin Ruskov

Researcher

Davide Taibi

Researcher

Alessandro Gabbiadini

Researcher

Dimitri Ognibene

Researcher

Contact Information

For information about the paper, please contact the authors.

Authors: Emily Theophilou, Cansu Koyuturk, Mona Yavari, Sathya Bursic, Gregor Donabauer, Alessia Telari, Alessia Testa, Raffaele Boiano, Davinia Hernandez-Leo, Martin Ruskov, Davide Taibi, Alessandro Gabbiadini, Dimitri Ognibene

Source Publication: View Original PaperLink opens in a new window

Project Contact: Dr. Jianhua Yang

LLM Model Version: gpt-4o-mini-2024-07-18

Analysis Provider: Openai

Let us know you agree to cookies