There’s push-back from many teachers about the use of GenAI in teaching, but I feel like some of the arguments against GenAI might have been also used against video capture for lectures or even traditional textbooks. Help me map out this space.
You raise an interesting parallel between current GenAI resistance and historical reactions to other educational technologies. Here's a mapping of this space:
-
Authenticity concerns
- GenAI: "Students will submit AI-generated work rather than their own"
- Video lectures: "Students won't attend real classes if they can watch recordings"
- Textbooks: "Students will rely on summaries rather than engaging with primary sources"
-
Pedagogical value concerns
- GenAI: "Students won't develop critical thinking if AI does the work"
- Video lectures: "Students lose interactive learning opportunities"
- Textbooks: "Standardized texts flatten nuanced discourse"
-
Teacher displacement anxiety
- GenAI: "AI might replace teacher expertise and guidance"
- Video lectures: "Star professors could replace local instructors"
- Textbooks: "Author authority supersedes teacher knowledge"
-
Scale and accessibility: GenAI can generate unique content instantly at a scale impossible with previous technologies
-
Mimicry of human thought: Unlike passive technologies, GenAI actively simulates human reasoning processes
-
Rapid evolution: The capabilities gap between early and current GenAI is much steeper than the evolution of previous educational technologies
Educational technologies that initially faced resistance but became normalized typically:
- Found appropriate boundaries of use
- Were integrated as supplements rather than replacements
- Developed clear ethical guidelines for implementation
- Demonstrated unique value propositions
The most productive path forward might involve:
- Acknowledging legitimate concerns while avoiding categorical rejection
- Developing GenAI literacy among educators and students
- Creating thoughtful policies that harness benefits while mitigating risks
- Recognizing that GenAI, like previous technologies, will likely find its appropriate place in the educational ecosystem
Does this mapping help with your thinking? I'd be happy to explore any particular aspect in more depth.
Go back to the common patterns of resistance and unfold them.
- Academic integrity fears: Concern that students will submit AI-generated essays, solutions, or creative works as their own
- Detection challenges: Unlike plagiarism, AI-generated work can be original and difficult to detect
- Identity verification: Questions about who actually completed the work when AI tools are accessible
- Skill representation: Worry that credentials won't accurately reflect students' actual abilities
- Physical presence valued: Belief that "real learning" requires being physically present
- Attendance decline: Evidence that recording availability reduces in-person attendance
- Passive consumption: Concern that students would watch without engagement
- Community loss: Fear that campus community and peer learning would diminish
- Secondary knowledge: Criticism that textbooks distance students from original sources and primary texts
- Standardized thinking: Concern that students would memorize textbook content rather than develop independent perspectives
- Intellectual shortcuts: Worry about students relying on summaries rather than engaging with full arguments
- Authority dependence: Fear that students would accept textbook claims without questioning
- Process vs. product: Concern that AI focuses on generating outputs rather than developing thinking processes
- Cognitive offloading: Worry that students will delegate cognitive tasks to AI rather than developing mental capabilities
- Superficial understanding: Fear that AI-assisted work appears competent without depth of understanding
- Feedback loop disruption: Concern that teachers can't properly assess and guide student development
- Lack of adaptability: Pre-recorded content can't adjust to student questions or confusion
- Passive learning model: One-way communication reduces active learning opportunities
- Reduced spontaneity: Loss of teachable moments and organic classroom discussion
- Pace inflexibility: Students can't influence the direction or speed of instruction
- Knowledge canonization: Textbooks present settled knowledge rather than evolving understanding
- Contextual limitations: Standard texts may not connect to students' diverse backgrounds
- Inquiry reduction: Structured presentation may discourage questioning and exploration
- Outdated content: Physical textbooks can't keep pace with rapidly evolving fields
- Expertise devaluation: Concern that AI could replace specialized knowledge teachers provide
- Personalization capabilities: AI systems increasingly able to adapt to individual student needs
- Continuous availability: AI doesn't have office hours or workload limitations
- Scalability concerns: Fear that educational models might shift toward AI guidance with minimal human oversight
- Star system emergence: Concern that celebrity professors would replace local instructors
- Centralized knowledge production: Fear that a few institutions would dominate content creation
- Role reduction: Worry that teachers would become facilitators rather than instructors
- Economic pressures: Institutional incentives to reduce faculty through recorded content
- Curriculum control: Textbook authors and publishers, not teachers, determining content emphasis
- Deskilling concerns: Teachers becoming implementers of textbook-defined approaches
- Authority competition: When teacher knowledge conflicts with textbook content
- Autonomy reduction: Standardized texts limiting teacher creativity and adaptation
Each wave of educational technology has triggered similar patterns of resistance, though the specific manifestations vary with the technology's capabilities and the educational context of the time. The resistance often reflects deeper values about what constitutes authentic learning and the essential role of human relationships in education.
Transpose this to group by common theme in the resistance while leaving the distinguishing between media (if necessary) to the leaf-level details.
Core concern: Technology disrupts the connection between student work and student learning.
-
Misrepresentation of abilities
- GenAI: Students submit AI-generated work that masks their true capabilities
- Video lectures: Students claim engagement with material they merely played in the background
- Textbooks: Students reference concepts they've memorized but not understood
-
Verification challenges
- GenAI: Difficulty distinguishing between student-created and AI-generated content
- Video lectures: Uncertainty about whether students meaningfully engaged with recordings
- Textbooks: Limited ability to assess if students processed information beyond rote learning
-
Shortcuts to credential acquisition
- GenAI: Using AI to complete assignments without developing underlying skills
- Video lectures: Watching at increased speed or only before exams
- Textbooks: Relying on summaries and study guides rather than comprehensive engagement
Core concern: Technology circumvents cognitive development necessary for deep learning.
-
Critical thinking atrophy
- GenAI: Outsourcing analysis and synthesis to AI rather than developing these skills
- Video lectures: Passive consumption replacing active questioning and discussion
- Textbooks: Accepting authoritative presentation rather than evaluating competing ideas
-
Diminished struggle and effort
- GenAI: Immediate solutions without productive struggle
- Video lectures: Illusion of understanding from watching rather than participating
- Textbooks: Following prescribed examples rather than developing problem-solving approaches
-
Reduced intellectual autonomy
- GenAI: Dependency on AI for ideation and evaluation
- Video lectures: Limited opportunity to direct learning through questioning
- Textbooks: Following standardized knowledge paths rather than self-directed inquiry
Core concern: Technology weakens essential human connections in education.
-
Diminished presence
- GenAI: Interaction with AI replacing human-to-human dialogue
- Video lectures: Physical absence from learning community
- Textbooks: Isolated reading replacing collaborative meaning-making
-
Feedback loop disruption
- GenAI: Teachers unable to identify misconceptions in AI-assisted work
- Video lectures: Delayed or absent opportunities for clarification
- Textbooks: Limited ability to address individual comprehension challenges
-
Relationship impoverishment
- GenAI: Reduced necessity for mentor-student connections
- Video lectures: Fewer spontaneous interactions between faculty and students
- Textbooks: Mediated rather than direct knowledge transmission
Core concern: Technology threatens teacher identity, autonomy, and employment.
-
Expertise devaluation
- GenAI: AI providing expert-like responses in teacher domains
- Video lectures: "Superstar" instructors replacing local faculty expertise
- Textbooks: Published authority superseding teacher knowledge
-
Pedagogical autonomy reduction
- GenAI: Pressure to incorporate AI tools regardless of teacher preference
- Video lectures: Standardized recordings limiting instructional flexibility
- Textbooks: Prescribed curriculum constraining teaching approaches
-
Economic and professional threats
- GenAI: Fears of AI replacing certain teaching functions
- Video lectures: Concerns about reduced faculty hiring when content is recorded
- Textbooks: Diminished role when serving primarily as textbook facilitator
Core concern: Technology changes what counts as knowledge and who controls it.
-
Accuracy and reliability questions
- GenAI: AI hallucinations and confident presentation of errors
- Video lectures: Outdated information persisting in recordings
- Textbooks: Errors or biases becoming canonized in standard texts
-
Centralization of knowledge production
- GenAI: Few AI companies determining information presentation
- Video lectures: Elite institutions dominating educational content
- Textbooks: Major publishers controlling curriculum materials
-
Contextual relevance limitations
- GenAI: Generic responses not tailored to specific learning contexts
- Video lectures: Content not adapted to local student populations
- Textbooks: Standardized examples disconnected from student experience
This organization highlights how resistance to educational technologies often stems from deeper concerns about learning, relationships, and professional identity rather than the specific technologies themselves.
So there’s clearly a lot of commonality. Let’s now synthesize the essential differences. What makes these media most different (reasons not to apply critiques from one media to the next)?
- GenAI: Actively creates new content and responds dynamically; possesses apparent agency and decision-making capabilities
- Video Lectures: Entirely passive and fixed; presents the same content regardless of viewer needs
- Textbooks: Static content with predetermined structure; cannot generate novel responses to queries
Key Distinction: Critiques about passive consumption apply strongly to video and textbooks but less to GenAI, which requires prompting and interaction. Conversely, concerns about surrendering agency apply uniquely to GenAI.
- GenAI: Can tailor responses to individual queries and adapt to specific user needs in real-time
- Video Lectures: One-size-fits-all content with limited personalization (beyond user-controlled playback)
- Textbooks: Standardized content for all users with fixed progression
Key Distinction: Critiques about lack of personalization apply to textbooks and videos but not to GenAI. However, GenAI raises unique questions about the quality and appropriateness of its personalized responses.
- GenAI: Creates content through statistical prediction without true understanding; no identifiable author
- Video Lectures: Human expert deliberately crafting and delivering content with full understanding
- Textbooks: Human authors and editors carefully developing and vetting content through established processes
Key Distinction: Concerns about authority and accuracy apply differently—textbooks and lectures represent human expert knowledge, while GenAI produces plausible-sounding content without expertise.
- GenAI: "Black box" operation that obscures reasoning processes and sources
- Video Lectures: Transparent presentation of human reasoning and knowledge construction
- Textbooks: Clear attribution with citations, references, and explicit reasoning
Key Distinction: Critiques about opacity and trustworthiness apply uniquely to GenAI, while textbooks and lectures typically make their sources and reasoning explicit.
- GenAI: Can generate seemingly original content that mimics human creativity
- Video Lectures: Presents original thinking of the lecturer but doesn't generate new ideas
- Textbooks: Synthesizes existing knowledge but doesn't create novel content for specific situations
Key Distinction: Concerns about plagiarism and intellectual property apply differently—GenAI raises unique questions about originality and idea ownership that don't apply to the other media.
- GenAI: Rapidly evolving capabilities with significant changes in months, not years
- Video Lectures: Relatively stable format with gradual improvements in production quality
- Textbooks: Slow evolution with new editions typically released on multi-year cycles
Key Distinction: The rapid advancement of GenAI makes policy and pedagogical responses particularly challenging compared to the more stable technologies.
- GenAI: Can perform cognitive tasks on behalf of the learner, potentially replacing learning
- Video Lectures: Requires learners to process information themselves, even if passively
- Textbooks: Demands active reading and information processing by the learner
Key Distinction: Concerns about bypassing essential cognitive processes apply uniquely to GenAI, which can complete tasks rather than just providing information.
- GenAI: Blurs the line between tool and collaborator/partner in the creative process
- Video Lectures: Clearly a resource rather than a participant in the learning process
- Textbooks: Unambiguously a tool with no collaborative qualities
Key Distinction: Questions about appropriate boundaries in the learning process apply differently to GenAI, which raises unique questions about co-creation that don't apply to passive media.
These fundamental differences suggest that while historical patterns of resistance provide valuable context, GenAI does present novel challenges that require fresh consideration rather than simply applying traditional frameworks.
Excellent.
Now, it turns out that I’m teaching a technical Teaching with Generative AI graduate course. I want my students to have a structured exercise where they work through this same line of thinking I did with you now. I want them to share their own ideas, and I want them to see yours. I want them to work at at somewhat, and I also want them to (as a byproduct of the exercise) come up with something to submit on the Canvas LMS so that I can track their progress. How should I approach this?
I'm looking for something students can do by themselves, at home, as a kind of generalized Reading assignment where they might chat with Claude or whatever.
- Analyze patterns of resistance to educational technologies across different media
- Identify unique characteristics of GenAI that differentiate it from previous technologies
- Develop critical thinking about appropriate applications of GenAI in education
- Practice effective prompting and dialogue with GenAI tools
-
Reflection Prompt: "Before using any AI tools, write 1-2 paragraphs about what you've heard educators say about concerns with using GenAI in teaching and learning."
-
Historical Comparison: "Identify at least one previous educational technology (e.g., calculators, internet, video lectures, textbooks) and list 3-5 concerns that were raised when it was introduced."
-
Initial Pattern Recognition: "What patterns do you notice between concerns about GenAI and your chosen historical technology? Note at least two similarities."
-
First Prompt: "Engage with Claude or another GenAI tool using this initial prompt: 'There's push-back from many teachers about the use of GenAI in teaching, but I feel like some of the arguments against GenAI might have been also used against [your chosen technology] or even traditional textbooks. Help me map out this space.'"
-
Follow-up Prompts: "Based on the AI's response, ask at least two follow-up questions that help you explore:
- Common patterns of resistance across technologies
- Essential differences that make GenAI unique from previous technologies"
-
Documentation: "Save the complete conversation transcript for submission."
-
Create a Framework: "Based on both your own thinking and your AI conversation, create a structured framework that:
- Identifies 3-5 common themes in educational technology resistance
- Highlights 2-3 unique characteristics of GenAI that differentiate it from previous technologies"
-
Application: "Choose one specific teaching scenario and describe:
- How GenAI could be valuably integrated
- What specific concerns would need to be addressed
- How insights from historical technology adoption might inform implementation"
-
Reflection: "Write a brief reflection (250-300 words) on how this exercise changed or reinforced your thinking about GenAI in education."
Students should submit a document to Canvas containing:
- Their initial reflections and historical comparisons
- The complete AI conversation transcript
- Their synthesized framework
- The application scenario
- Their final reflection
- Depth of analysis in identifying patterns across technologies
- Nuance in distinguishing GenAI from previous technologies
- Quality of AI prompting and follow-up questions
- Thoughtfulness of application scenario
- Critical reflection on personal learning
- Emphasize that the value is in their analysis and synthesis, not just in collecting AI outputs
- Clarify that you're evaluating their thinking process and engagement, not whether they agree with particular perspectives
- Consider sharing selected frameworks in a follow-up discussion to highlight diverse approaches
- You might provide my analysis as a "sample response" after submissions are complete, rather than beforehand, to avoid anchoring their thinking
This exercise allows students to engage with the material independently while producing a concrete submission that demonstrates their learning process and outcomes.
I wonder if I could bake this assignment structure into a kind of system prompt that Claude will use to guide the activity and tell students when they are done (and ready to export their conversation for submission to Canvas). Earlier-gen LLMs couldn’t stay on task well enough for this, but I think things might be different now.
I think we need a compact definition of the assignment that I can share on Canvas. These are the instructions for the student for how to use Claude. It also needs the detailed propmt to paste into Claude.
The prompt would need explain to Claude what the assignment is and how to keep both itself and the student on track. For example, it might instruct both the student and Claude to simply restart the assignment if the structure seems to fall apart. It is better to start fresh with the ideas already loaded into the student's mind rather than try to recover from lost LLM context. Claude needs be told not to do the student's work for them, but it can provide appropriate forms of assistance.
This should be a 30-minute exercises in a conversational format where the student sends many short messages (not composing an essay in the message box).
Rather than instructing them to use the Claude.ai website, have them use this link to our campus AI service: https://bayleaf.chat/?model=anthropic_v2.claude-3-7-sonnet-latest&q=I'll+paste+the+instructions+momentarily.
Assignment: Analyzing Patterns of Educational Technology Resistance (30 minutes)
For this assignment, you will engage in a structured conversation with Claude to explore how resistance to GenAI in education compares with historical reactions to previous educational technologies.
Instructions:
-
Go to our campus AI service: https://bayleaf.chat/?model=anthropic_v2.claude-3-7-sonnet-latest&q=I'll+paste+the+instructions+momentarily
-
Copy and paste the entire prompt below into your first message to Claude.
-
Follow Claude's guidance through the 30-minute structured exercise. Respond with short messages rather than lengthy essays.
-
When you complete the exercise, Claude will indicate you're finished. Copy your entire conversation and submit it to Canvas.
Note: This exercise is designed to capture YOUR thinking. Claude will guide but not do the work for you. If the conversation gets off track, simply tell Claude you want to restart the exercise.
I'm completing an assignment for a graduate course on "Teaching with Generative AI." Please guide me through this structured 30-minute exercise about comparing resistance to GenAI with historical patterns of educational technology resistance.
Here are your instructions:
You should guide me through a structured conversation with these components:
- Initial reflection on concerns about GenAI in education (my own thoughts)
- Identifying a previous educational technology and concerns raised about it
- Exploring patterns between GenAI concerns and historical technology concerns
- Synthesizing common themes across educational technologies
- Analyzing what makes GenAI fundamentally different from previous technologies
- Applying insights to a specific teaching scenario
- Brief reflection on what I've learned
Important guidelines for you (Claude):
- Keep me on track through all 7 components
- Ask one clear question at a time and wait for my response
- Do NOT provide your own examples or analysis until AFTER I've shared my thoughts
- After I share my ideas for each section, you may then share 1-2 additional insights
- If our conversation gets off track, suggest we restart the current section
- When we complete all sections, clearly indicate the exercise is finished
- Remind me to copy the entire conversation for submission to Canvas
Remember that the goal is to capture MY thinking and analysis, not to have you generate the content for me. Your role is to guide the conversation, ask good questions, and provide limited supplementary insights after I've shared my own ideas.
Let's begin!