My Journey with AI: Discovering Notebook LM
I was one of the holdouts when it came to AI. I had tried it when AI first came out. All I found were hallucinations, false information and I felt it was more trouble than it was worth. A friend and peer talked me into trying NOtebook LM in early 2026. What I found was so eye-opening.
A New Beginning
The AI models have improved. It could do things beyond what I expected. I started with Notebook LM—a place I could upload images, articles, documents, books, etc., and the answers to the questions I asked were supposed to come from just those sources I uploaded or websites I linked. This felt safe, because the AI answers were "contained" to the sources I gave it.
My First Success: Creating a Timeline
So I played with Notebook LM. I uploaded 50 documents about a person of interest that I needed to sort through, transcribe, and put in date order, to create a timeline of that person's life.
The timeline was generated super fast. I asked for a table with the date, location, and a summary of the document. I asked for a list of the fan club and in seconds the list was there. I could quickly double-check the performance of Notebook LM, and so far it was spot on.
Real-World Application: Census Instructions
I began to think of ways this new tool could help make things faster. Let's take the US Census instructions for example. I used to open all of the PDF documents (17 of them) and use "control F" to find which document(s) had the answer. It was very time consuming.
With this new tool (Notebook LM) I could add all of the census instructions into one notebook, then ask my question of the tool, and have my answer. I liked that Notebook LM also gave me a citation number. When I clicked on that number, it gave me the file and a snippet of the information. I could now go to that source to verify the AI's answer.
Question 1: When Was the Person Giving Information Marked with an X?
In the 1940 census, the person who furnished information about a household's members was identified by writing an X with a circle around it (ⓧ) in column 7 immediately after their name.
But then the answer continued...
If the enumerator had to obtain information from someone who was not a member of the household (such as a neighbor), they were instead instructed to write that person's name in the left-hand margin of the schedule, opposite the entries for that household.
Now we know enumerators didn't always follow instructions, but now I'm on the lookout for pages from the 1940 census with names written in the left-hand margin.
Question 2: Census Year and Age Calculation Dates
Based on the provided census instructions, here are the census years and the specific dates enumerators were instructed to use for calculating the age of the individuals being enumerated:
A list with the specific instructions for each census year followed.
Question 3: Naturalization Information by Census Year
Information regarding naturalization status and the length of time a person had been in the United States was collected in various ways across different census years. The most detailed record for the specific year of naturalization was requested in the 1920 census.
But then it goes on and breaks my question down further:
Naturalization Status and Year of Naturalization
Had a breakdown of the following census years (1890, 1900, 1910, 1930, 1940, 1950) and what kind of naturalization status was given.
Length of Time in the United States
Had a breakdown of the following census years (1820, 1830, 1870, 1890, 1900, 1910, 1930, 1940) and additional information to the "how long" question.
Why Notebook LM?
Yes, I could have probably asked Google these questions, but Google pulls from the internet, and there is a lot of misinformation out there. I like that my answers came "directly" from the source—the Census Enumerator Instructions.
Notebook LM Studio Features
Notebook LM has other features in their "studio" such as:
- Creating an infographic
- Creating a video that tells an engaging story, using the documents
- Creating flashcards, great for studying a topic
- Creating a quiz, to test your knowledge
- Creating a mindmap
- Creating a slide deck
- Creating an audio like an AI podcast.
- Creating a report
- Creating a datatable
After using Notebook LM for the past 6 months, I realize there is still much more to explore in Notebook LM. I have created 29 notebooks to date, and Google will allow me to have 100 notebooks as a free user.
What's Next?
Stay tuned as I review each of the AI platforms (ChatGPT, Perplexity, Gemini, and Claude) in upcoming blogs.
