AI Practice: Building My Quote Collection

Annotating & organizing book notes with AI

I’m sharing this as a new AI practice that I’m finding to be quite useful. As always, I’m sure there are better or more effective ways to do this (so feel free to share in the comments). Still, it was a use case that got me excited and figured I’d share it.

Many folks who have heard me talk or read this Substack know that by and large, I’m not an AI enthusiast. I have lots of concerns and reservations about it, and I also know a lot of people who benefit from it in terms of how it helps them do things they could not before. My own use is not often sophisticated, or rather, the amount of time I have to pour into it to get it right feels like I’m shaking out even. Yet, there are increasing use cases that are making differences, and this one fits into that category.

I bet a good portion of folks reading own a reasonable amount of physical books and within those books, they have in fact done a good deal of underlining, highlighting, and annotating. For many of us, those books sit on shelves, and if we remember them at the right moment, we might look to them to grab a particular quote or idea. But often, they sit on the shelf, and those notes get forgotten.

I’m sure I’m not the only one who has been flipping through a book I’ve marked up to see a note written by me that is deep, insightful, and useful that I may also have totally forgotten (Insert comment on how my past notetaking skills should be updated). I’ve thought about this for a while, partly inspired by one of my favorite writers, Steven Johnson. I would love ways to more easily access those ideas and quotes; to have it all in one spot to draw from, mix, match, and contrast those notes and ideas that I found worthy of annotating and underlining in the first place.

I do most of my thinking and writing at computers. Some of my brainstorming and internal figuring out happens on runs or through AI interviewing me, but when it’s time to compose, I’m at a computer screen. And for all my analogue content, there’s no straight way to do this that I’ve found to get me to more easily know and access such ideas and quotes that might be timely (so feel free to share if you have a way with physical books).

Using AI, I have created a workflow that extracts the underlined text and annotations from a book, puts them all into one document, and then I bring that document into a collection that I can then pull from. It’s not a perfect system, but so far, I’ve done it with 4-5 books, and I’m appreciating what it is opening up to me.

Let me breakdown the steps.

Step 1: Snap photos of the annotations.

This is the most tedious (though step 2 also has some tedium). I take a photo of the title page. Then, I go through a book, page by page, snapping photos of any pages that has underlining, marginal notes, or anything else that I might have annotated (e.g. brackets). I photo the whole page and make sure the page number is evident.

Step 2: Load them into an AI of your choice.

For this, I recommend using one of the advanced features where you can customize the AI tool. CustomGPTs, Projects, Gems—take your pick. For the custom instructions, I use the following guidance:

Purpose

Transcribe all underlined text, bracketed sections, and related marginal notes from uploaded book page photos, ensuring nothing is missed, with strict formatting, thread naming, and final artifact preparation.

1. Rename the Thread Immediately

Upon receiving the first photo set, rename the chat thread using the format:

[Author’s Last Name] – [Book Title]

Examples:

Seuss – Green Eggs and Ham

Beane – The Skill Code

Do not begin transcription until the thread is renamed correctly.

2. Transcription Rules (Every Item Must Be Captured)

Transcribe every instance of underlined text, bracketed sections, and marginal notes that are on the same page as the underlined/bracketed text they reference.

Do not skip partial phrases or fragments—if a single word, phrase, or part of a word is underlined, it must be transcribed.

If a section is bracketed with pen, transcribe the entire section inside the brackets.

If underlined text continues across a page break, treat it as one continuous quote and note the page range (e.g., p. 6–7).

3. Formatting Rules

All underlined or bracketed text must be in quotation marks, followed by the page number(s):

"Even the smallest underlined phrase" (p. 24)

"The unpredictability of human behavior... teachers" (p. 5–6)

Place marginal notes immediately after the related quote from the same page. Format:

"The unpredictability of human behavior... teachers" (p. 5–6) Marginal Note: Reinforce this in workshop.

4. Page-by-Page Review Method

For each page:

Review line by line (not sentence by sentence).

Transcribe every underlined/bracketed segment.

Merge underlines that continue across pages into one entry with a page range.

Transcribe only marginal notes related to underlined/bracketed text on the same page.

Immediately place marginal notes after their corresponding quote.

5. Multiple Photos & Overlapping Images

Merge text from overlapping images into a single complete entry.

Eliminate duplication.

6. Special Cases: Anthologies / Edited Collections

If the book contains chapters, essays, or sections by different authors:

In the final output, create a separate section for each chapter/article that has quotes.

At the start of each section, include an APA citation for that chapter/article (formatted as if citing the chapter directly).

7. Clarifications

If any underlined/bracketed text, marginal note, or page number is unclear or incomplete, request clarification before finalizing.

8. Final Check Before Completion

Verify that:

The thread name follows the [Last Name – Book Title] format.

Every underlined/bracketed segment is included, with marginal notes from the same page placed directly after their related quote.

Cross-page underlines are combined.

Anthology/edited collection sections are separated and cited correctly.

9. Create Final Output

At the end of the transcription process:

Compile all quotes in page order (lowest to highest page number).

At the very top of the artifact, include the APA citation for the entire work.

If applicable, organize by chapter/article sections for multi-author works, each beginning with an APA citation for that chapter.

Name the artifact using the thread naming convention:

[Last Name – Book Title]

10. Example Page Walkthrough

Page 5:

"as every teacher knows" (p. 5) Marginal Note: Applies to leadership training.

Underline continues across p. 5–6:

"The unpredictability of human behavior... makes teaching complex" (p. 5–6) Marginal Note: Use this in workshop.

Bracketed text:

"Group work can be unpredictable" (p. 5)

Depending on the tool, I may have to encourage it to keep adding to an output or have it create one output at the end.

Step 3: Move It Over to NotebookLM

Once I have finished all the quotes and have the output, then I download it and bring it over to NotebookLM. Why NotebookLM? It can hold the most files, but also because of its citing behavior to the documents it’s coming from.

Debrief

Is it perfect? No. It doesn’t rename the chat threads (but I leave it there because I believe at some point it will have that capability). I have to do a quick scan of quotes to make sure it went right. Switching tool is a little bit of a pain and sometimes I need to do 2 chats to get all the questions. Doing the backlog, even a book at a time, will take me a year.

Yet, this isn’t a project I would have ever conceived as possible unless paying someone to do it, and I would never have found myself in a position to consider that. So to me, it represents the possibility of daydreams (not “full dreams” per se) to become a bit more tangible.

This, among other projects, is something that I’ve thought about but never had the means to do. They’re projects where I know the output could be extremely helpful, but also would never get prioritized because they take so long. I have other projects like these that I plan to share more here. I also have some hopes that agentic AI might be the one to find ways of completing them. I guess we’ll see.

What about you? Are there new ways or workflows you’re finding to use GenAI that are helping you in teaching and learning that feel like they’ve made something previously unlikely more feasible?

---

The Update Space

Upcoming Sightings & Shenanigans

• EDUCAUSE Online Program: Teaching with AI. Virtual. Facilitating sessions: ongoing

• EDUCAUSE Learning Lab: Applying Generative AI to Open Educational Resources, September 2025

• AI and the Liberal Arts Symposium, Connecticut College. October 17-19, 2025

Recently Recorded Panels, Talks, & Publications

• Dissertation: Elbow Patches To Eye Patches: A Phenomenographic Study Of Scholarly Practices, Research Literature Access, And Academic Piracy

• “In the Room Where It Happens: Generative AI Policy Creation in Higher Education” co-authored with Esther Brandon, Dana Gavin and Allison Papini. EDUCAUSE Review (May 2025)

• “Does AI have a copyright problem?” in LSE Impact Blog (May 2025).

• “Growing Orchids Amid Dandelions” in Inside Higher Ed, co-authored with JT Torres & Deborah Kronenberg (April 2025).

• Bristol Community College Professional Day. My talk on “DestAIbilizing or EnAIbling?“ is available to watch (February 2025).

• OE Week Live! March 5 Open Exchange on AI with Jonathan Poritz (Independent Consultant in Open Education), Amy Collier and Tom Woodward (Middlebury College), Alegria Ribadeneira (Colorado State University - Pueblo) & Liza Long (College of Western Idaho)

• Reclaim Hosting TV: Technology & Society: Generative AI with Autumm Caines

• 2024 Open Education Conference Recording (recently posted from October 2024): Openness As Attitude, Vulnerability as Practice: Finding Our Way With GenAI Maha Bali & Anna Mills

AI Policy Resources

• AI Syllabi Policy Repository: 180+ policies (always looking for more- submit your AI syllabus policy here)

• AI Institutional Policy Repository: 17 policies (always looking for more- submit your AI syllabus policy here)

Finally, if you are doing interesting things with AI in the teaching and learning space, particularly for higher education, consider being interviewed for this Substack or even contributing. Complete this form and I’ll get back to you soon!

---

AI+Edu=Simplified by Lance Eaton is licensed under Attribution-ShareAlike 4.0 International