
GenAI Introduction & Prompt Design: A Double-Edged Sword
Time plan: Pre-work 25 min • Mini-lecture 15 min • Hands-on 65 min
GenAI can accelerate good work—or amplify weak work. This workshop gives you simple prompts, safety rules, and practice so you can use GenAI productively without undermining judgment or academic integrity.
About the workshop
You’ll start with prompt design and the ethics/safety basics that matter in student research. In the hands-on segment, you’ll evaluate GenAI outputs for accuracy and bias, and you’ll see how UVic policies shape responsible use.
What you’ll learn (learning objectives)
By the end of this introductory workshop, you will be able to:
- Explain GenAI in plain English and list where it helps in academic work.
- Write better prompts using audience, format, constraints, and evidence requests.
- Run a basic fact-check loop (ask for claims → verify with two sources → correct errors).
- Adjust privacy settings and read T&Cs with a safety mindset.
- Use GenAI for practical tasks (brainstorming, outlining, formatting) without breaching integrity.
- Give pre-submit feedback on your own assignments using a simple rubric.
- Evaluate output quality for accuracy, relevance, bias, and missing context.
- Follow UVic expectations for academic integrity and responsible use.
What you’ll do today (agenda)
- Mini-lecture (15 min): When GenAI helps vs. hurts; prompt hygiene.
- Hands-on (65 min):
- Prompt hygiene (tighten a weak prompt with audience/format/constraints).
- Idea generation (diverge → converge; keep what’s defensible).
- Critical review (ask for numbered claims, verify 2, correct 1).
- Privacy & redaction (replace names/emails/IDs with placeholders).
- Assignment feedback (build a rubric; get structured, auditable comments).
NEXT STEP: Pre-Workshop Activities
How to talk to GenAI (quick guide for beginners)
- Task: “Summarize a 1-page article for a first-year audience.”
- Context: Paste an excerpt or bullet key facts; say what not to change.
- Constraints: Length (120–150 words), tone (neutral), must-include/must-avoid list.
- Output format: “Outline first, then the paragraph.”
- Evidence: “List 3 verifiable claims with citations or URLs.”
- Verify: Check independently; fix any errors before you submit.
Safety & integrity
• Don’t paste confidential or personal data.
• Attribute sources for factual claims or quotations.
• You own the final result—verify before submitting.
Generative AI Workshop Pathway
This workshop is part of our Generative AI Workshop Pathway for students, faculty, and staff. It spans foundational to advanced research and productivity skills: