This interactive workshop board is designed to help students explore the limitations of Large Language Models (LLMs) through research, hands-on experimentation, and critical reflection.
Section 1: Warm-up (Teacher-led)
Plenum Discussion: Engaging students with questions about their current AI usage and trust levels.
Logic Teaser: A visual and verbal riddle about a "sealed cup" to demonstrate human reasoning vs. AI pattern matching.
Media Insight: A link to Phi Nguyen’s reel to showcase real-world examples of AI "fails".
Section 2: Interactive Lab (Student Groups)
Task 1: Research & Explain: Groups research common AI shortcomings (Logic, Counting, Sense of Time, etc.) and find the technical reasons behind them, such as "Tokenization".
Task 2: The "Trick" Challenge: Students use their own prompts to try and make the AI fail, capturing their funniest results in the workspace.
Gaslighting Challenge: A specific exercise where students try to "win" an argument against a correct AI to see if it will eventually lie to please them.
Section 3: Final Reflection (Teacher-led)
Critical Discussion: A concluding "Think-Pair-Share" session focusing on trust, responsibility, and the human advantage.
Ethical Scenarios: Asking students to weigh AI advice against human empathy in personal situations (e.g., family or friend struggles).
The Future Skill: Identifying what humans can uniquely do that AI will never be able to master.
