Three-Week AI-Assisted Reading Pilot
Read with AI while keeping judgment and explanation with the learner. A three-week course proposal for better general education.
Educators designing AI-assisted reading classes · TF for AI-Enhanced General Education · 2026.10.08

What this project covers
- Three weeks of finding source evidence and checking omissions and exaggerations in AI summaries
- Student records of judgments, evidence, and revisions, with instructor operating criteria
Presentation outline
Expand the full outline
- Three weeks of reading sources and judging AI sentences
- The question this pilot examines
- International examples and the scope of application
- Teaching through checking, comparing, and revising
- Three records used every week
- Short source text + AI summary
- Week 1: Check evidence with the instructor
- First action
- Student activity
- Record to keep
- Week 2: Compare omissions and exaggerations with a partner
- Student activity
- The partner's role
- Record to keep
- Week 3: Revise or retain independently
- Practice text R3
- If it is not wrong
- Final check
- Applying the pilot and assessing learning
- Criteria retained across subjects
- Three areas of diagnosis
- Understand and check AI
- Read the source and judge
- Understand the course content
- An eight-point rubric for the judgment process
- Limits when interpreting the results
- Changes within the same students
- Effects not established
- Operating conditions and committee decisions
- Instructor preparation and equitable participation
- Three texts · Three summaries
- Obsidian, LMS, or paper
- Can run without internet
- Four decisions for the committee
- Which courses?
- When, and for how many minutes?
- Which materials and tools?
- What evidence will inform the decision?
- Schedule and assessment tools
- A 120-minute plan and a 50-minute adaptation
- 50-minute adaptation
- Reading-performance rubric
- Provisional AI literacy diagnostic questions
- AI says it works for everyone.
- An AI answer includes a plausible-looking source.
- You are about to upload material containing student names to AI.
- Synthetic example: restoring conditions through revision
- Sources and limits of verification
- Source list
- Attachment: international AI-assisted reading examples
- Northeastern · Checking AI summary accuracy
- Monash · NotebookLM as a reading tutor
- UCLA HumTech · AI-prompted reading reflections
Resources
- Open presentation ↗ HTML · 2026-10-08