Lesson Specification: Module 4.2

Metadata

Learning Objectives

By the end of this lesson, students will be able to:

1. Differentiate between ethical and unethical uses of Large Language Models (LLMs) in academic writing.
2. Apply LLMs as brainstorming, outlining, and editing partners, rather than as ghostwriters.
3. Evaluate the risks of AI hallucinations and the necessity of human accountability in research output.
4. Draft a transparency statement disclosing the use of AI tools in a manuscript.

Key Concepts

Concept 1: The "Copilot, Not Autopilot" Principle

• LLMs are statistical pattern matchers, not truth engines. They predict the most likely next word.
• Unethical use: Asking an AI to write a literature review or generate data. (This is academic misconduct).
• Ethical use: Asking an AI to check grammar, suggest structural improvements, translate from non-native English, or summarize a dense paragraph that you already wrote.

Concept 2: Hallucinations & Accountability

• AI models hallucinate citations. They will invent papers, authors, and DOIs that look perfectly real but do not exist.
• The Rule of Accountability: You are 100% responsible for every word and citation in your paper. "The AI made a mistake" is not a defense against plagiarism or falsification charges.

Concept 3: Prompting for Structure, Not Content

• Bad prompt: "Write an introduction about CRISPR."
• Good prompt: "Here is my draft introduction about CRISPR. Does it follow the IMRaD funnel structure? Point out any sentences that are longer than 25 words."

Concept 4: Transparency & Disclosure

• Most major publishers (Nature, Science, Elsevier, IEEE) now have explicit AI policies.
• General rule: AI cannot be listed as an author. The use of AI must be explicitly disclosed in the Methods or Acknowledgments section.

Suggested Hooks & Motivation

• Pain Point: You want to use ChatGPT to speed up your writing, but you're terrified of getting caught for plagiarism or submitting a hallucinated citation to your advisor.
• Wow Moment: Seeing the difference between a lazy "write this for me" prompt (which produces generic, detectable fluff) and a targeted "edit this for me" prompt (which produces sharp, professional academic prose).

Cross-References

• Builds on: Module 4.1 (Academic Writing — you need to know the rules before you can prompt an AI to enforce them).

Sources & Further Reading

1. Nature Portfolio (2023). Artificial intelligence (AI) in research and publishing.
2. Hosseini, M., et al. (2023). An ethics checklist for AI in academic writing.