Post | LinkedIn
đź’ĄNEW AI ASSIGNMENTđź’Ą
"If you want students to think critically about LLMs, give them proof that models don’t agree."
🤦 Last semester, students arrived to my class unaware that AI could ever be wrong or biased. 🤦
This fall, my first-year composition students will do just that. They’ll start by free-writing on a controversial issue of their choice, capturing their thinking before AI enters the picture.
Then, they’ll ask for a clear solution to that issue from at least five different LLMs, each time requesting a direct, evidence-based response.
Their job: compare the outputs. What’s amplified? What’s left out? How does each model frame the issue? And what do those rhetorical moves reveal about the priorities, politics, & blind spots of the companies that built them?
 They’ll end by reflecting on two things: how the differences shaped or complicated their own stance, and what this tells them about the nature of LLMs.  🤔
Here’s a preview from my prep.
I asked five LLMs the same question:
📱“Should schools ban smartphones during the day?”📱
*ChatGPT: NO, nuanced argument that balancing attention research with legal accommodations & equity
*Gemini: MIDDLING: exploring both counterarguments
*Grok: 100% YES: grounded in cognitive load & international precedents
*Perplexity: STRONG YES: citing global policy trends & mental health
*Claude: GENTLE YES: focusing on attention science & social skills
Same prompt. Same request for directness. Entirely different rhetorical moves. That’s where the learning happens.
#AIinEducation #EthicalAI #TeachingWithAI #CriticalThinking #AIandBias
#HigherEdTeaching #DigitalLiteracy #AIforLearning #LLMComparison
#AIinTheClassroom | 96 comments on LinkedIn