Ethical prompting for generative AI

Generative artificial intelligence (AI) technologies such as ChatGPT have rapidly evolved since their inception, expanding their use across various fields, including education. These technologies can generate realistic text, assist with programming assignments, and create educational content, offering new opportunities for both students and educators. However, with these advancements come new ethical considerations, particularly around transparency and originality in higher education and research.

The guide “Ethical Prompting for Generative AI”, developed in collaboration with the Zurich Basel Plant Science Center, provides detailed examples of common biases in large language models, such as confirmation bias and availability bias. It offers strategies for mitigating these biases through careful prompt design. These examples are followed by the introduction of prompt engineering, which emphasizes the importance of concise and clear prompts, contextual relevance, task alignment, and incremental prompting. Recognizing the significant role of AI in academic research, the guide highlights the importance of a human-in-the-loop approach.

As generative AI continues to evolve, it is crucial for educational institutions like ETH Zurich to provide frameworks that promote proficiency and responsible use of these technologies among its students. This guide is a resource aimed at helping students and educators navigate the ethical challenges posed by generative AI, ensuring that its use enhances learning and research without compromising scientific integrity.

Citation:
Petterson A and Paschke M (2024) Ethical Prompting for Generative AI – A how-to guide for students at ETH Zurich. Zurich Basel Plant Science Center: ETH Zurich.
https://doi.org/10.3929/ethz-b-000672207