Topic 1 hookRevisit Topic 11
Forward Deployed Engineer
AI changes many technical workflows. The valuable person understands the problem, talks to stakeholders, uses agents, verifies outputs, and owns a working system.
Read postTopic 1 Hooks: What AI Means For Students
Use these for a short opening discussion, not a full reading assignment. The goal is to show students that AI literacy is about problem definition, collaboration, verification, and shipping useful work.
Topic 1 hookTopic 8/10 reuse
Problem Definition Wins
A domain expert with a clear problem can outperform faster coders. Students should learn to specify the right thing before asking an agent to build.
Read postTopics 1-3Reflection
AI Fluency Index
Fluent users manage AI rather than delegate blindly: they iterate, ask for missing context, critique polished output, and keep responsibility for the work.
Read postTopics 2-3: Agent Mindset, Limits, and Responsibility
Use these to connect agent setup, human review, and responsible use. Keep provocative claims instructor-framed unless sources are reverified before class.
Topic 2 framing
End of Coding
Useful for macro-action thinking: students direct agents through goals, context, tools, branches, and review rather than only typing code by hand.
Read postTopic 2/8/10
Frustration Shifts Upward
AI reduces syntax friction but makes specification, architecture, diagnosis, and review more important. This supports the whole agent-based course design.
Read postTopic 3 reflection
AI Brain Fry
Student-facing case for cognitive load, tool switching, verification fatigue, and why a small deliberate AI stack can beat more tools.
Read postTopics 8-10: Building With Agents
Topic 8/10
Problem Definition Wins
Return to this as a full app-building case: before students prompt an agent, they must define audience, workflow, constraints, and success criteria.
Read postTopic 10
Frustration Shifts Upward
Use when students hit app complexity. The issue is often not "how do I code it?" but "did I choose the right structure?"
Read postTopic 11 and Topic 14: Career, Learning, and Portfolio
Topic 11
Career-Ops
Agent workflow for evaluating postings, tailoring materials, tracking applications, and preserving human control. Good homework case: do not auto-submit.
Read postTopic 11/14
Agent Portfolio Evidence
Students should show prompts, setup, agent instructions, tests, and verification habits, not just polished screenshots of final projects.
Read postTopic 11 optional
Context-Rich Learning
Optional study-skills case: learning works better when AI connects vocabulary or facts to authentic context rather than isolated memorization.
Read postTopic 12: Prompting and Model Behavior
Topic 12
Shortest Prompt Lines
Use as compact prompt-review moves after students learn structured prompts: ask for missing context, disagreement, uncertainty, and critique.
Read postTopic 12 optional
Bilingual Prompting
Optional extension for bilingual students and model-behavior discussion. Verify numeric research claims before teaching them as facts.
Read postHidden Teacher Training
These belong mainly in the hidden teacher-training chapter. They are about educator workflows, feedback, rubrics, and the line between support and outsourcing.
Teacher training
Teacher Factory
Frames AI as a teacher material-production system: custom activities, rubrics, examples, and differentiated supports.
Read postTeacher training
NotebookLM Grading Paradox
Use for instructor-only discussion of rubric-grounded feedback, privacy, academic integrity, and human responsibility in grading.
Read postRubric input
AI Fluency Index
Use as a teacher-facing rubric input: look for iteration, critical review, uncertainty handling, and collaboration settings.
Read postTopic 13 and Advanced Enrichment
Use teacher-selected excerpts. These are useful for evidence quality and expert verification, but too specialized for required CS-215 reading without simplification.
Topic 13 excerpt
Mythos Verification Crisis
Evidence-quality case: how should students read model-risk claims, vendor claims, restricted-access claims, and independent evaluations?
Read postAdvanced excerpt
Automated Alignment Researchers
Advanced case in evaluation and AI research workflows. Useful for discussing benchmark integrity and scalable oversight.
Read postDiscussion only
Medical AI Scientist
High-stakes caution case. Use for ethics discussion only, not student emulation or workflow imitation.
Read postTopic 5 optional
AI Quant Tools
Optional domain application for finance/time series. Too specialized for required Excel/admin workflow.
Read postTopic 2 optional
Subquadratic Attention
Advanced architecture and skepticism example. Keep out of core flow unless students are ready for attention/computation tradeoffs.
Read postAdvanced coding
TurboQuant
Case for human expertise in low-level AI-assisted coding. Useful later, but not required for CS-215.
Read postOptional enrichment
TensorTonic
For students who want ML fundamentals. Good enrichment, but not central to the agent-app course path.
Read postHow To Use This Bank
For student modules
- Use one hook slide, one discussion question, or one short reflection.
- Promote only Topic 11 career and Topic 13 verification cases into heavier reading.
- Keep technical posts as excerpts unless the class has the prerequisites.
For teacher prep
- Use teacher-facing posts to design rubrics, examples, feedback workflows, and disclosure rules.
- Reverify external sources before making numeric claims in class.
- Favor cases that reinforce human judgment, not tool worship.