Using NotebookLM AI as an editor in physical science class

I do not use AI detection systems in SC 130 Physical Science for laboratory report submissions as I have reasons to doubt their accuracy. AI is evolving too quickly for detection systems to keep up. In addition, false positives are deeply damaging to the instructor-student relationship. False positives are undoubtedly very rare, but anecdotal evidence suggests that false positives disproportionately affect the most academically capable students.  

The course syllabus does not currently include a statement on the use of AI because as an instructor I am still working to understand what a functional AI policy will look like. I can only encourage the students to write reports in their own words. My intent is to help them improve their own writing skills, not the writing skills of an AI. 

Sometimes I can sense that AI has been used - AI tends to use vocabulary and idioms found in formal scientific writing and technical publications. That is what the AI is trained on, but my students were not. The style of writing, word choices, and phrases sometimes makes me think AI wrote a submission. 

As of summer 2026 the course operates on an honor system: I ask the students to write their own laboratory reports and I trust that they will do so. I also know, however, that students will use AI no matter what I say. And their future workplace is likely to use AI as well. 

As a result I told the students that they should also use AI. But not the AI they have been using, and not the way they have been using it. I instructed the students to use NotebookLM set up as I showed them in class - with the textbook links loaded into the notebook.*  I told the student to write their own text first. Then use NotebookLM to check their spelling and grammar. As an example, the following image shows an actual introduction submitted by a student and the prompt I gave in NotebookLM:

The AI came back with the following corrections:

=== NotebookLM response ===

Your draft is a great start! Here is a version with corrected spelling and grammar:

"Today's lab report focuses on the density of soap. In this lab, we are conducting experiments to discover the density of various objects. We are calculating density by exploring the relationship between mass (m) and volume (v)."

Key Corrections: 

Lab vs. Lap: Use "lab" (short for laboratory) instead of "lap."

Soap vs. Soup: I corrected this to "soap" based on the first source in your notebook regarding "Lux Magical Spell soap."

Density: Remember that density is specifically the ratio of mass divided by volume (D=m/v).

While I used my general knowledge to help with these corrections, NotebookLM works best when you select your sources. If you select your "Density" and "Soap" documents in the source panel, I can help you incorporate specific data points and formulas into your report. 

Would you like me to generate a Study Guide on density to help you with the mathematical models for your report?

=== NotebookLM end of response ===


When AI is used this way, by having the student write first and then check the spelling and grammar using NotebookLM, the student is using AI as an editor. Writers have editors, AI is a student's own personal editor. The work is still their own work. The revised text does not use words the student would not use. The work is still the student's own work, but their work as corrected by an editor. 

NotebookLM is particularly well suited to this editorial task because NotebookLM is functionally a private AI model, informed by the materials loaded into the notebook. NotebookLM is a walled garden model, albeit with semi-permeable walls. Under the hood the AI does have a model that was trained on other material. NotebookLM, however, relies primarily on the material loaded into the notebook including, in many instances, active links back to the sources loaded in the notebook.

A note on the physical science NotebookLM setup

*  A video shows how this is done in MS 150 Statistics, a similar link list populates NotebookLM with the physical science textbook links:

https://sites.google.com/comfsm.edu.fm/danaleeling/courses/statistics/stats-text-00-preface

https://sites.google.com/comfsm.edu.fm/danaleeling/courses/statistics/stats-text-01

https://sites.google.com/comfsm.edu.fm/danaleeling/courses/statistics/stats-text-02

https://sites.google.com/comfsm.edu.fm/danaleeling/courses/statistics/stats-text-03

https://sites.google.com/comfsm.edu.fm/danaleeling/courses/statistics/stats-text-035

https://sites.google.com/comfsm.edu.fm/danaleeling/courses/statistics/stats-text-04

https://sites.google.com/comfsm.edu.fm/danaleeling/courses/statistics/stats-text-05

https://sites.google.com/comfsm.edu.fm/danaleeling/courses/statistics/stats-text-06

https://sites.google.com/comfsm.edu.fm/danaleeling/courses/statistics/stats-text-07

https://sites.google.com/comfsm.edu.fm/danaleeling/courses/statistics/stats-text-08

https://sites.google.com/comfsm.edu.fm/danaleeling/courses/statistics/stats-text-09

https://sites.google.com/comfsm.edu.fm/danaleeling/courses/statistics/stats-text-10

https://sites.google.com/comfsm.edu.fm/danaleeling/courses/statistics/stats-text-11

https://sites.google.com/comfsm.edu.fm/danaleeling/courses/statistics/stats-text-12

Comments

Post a Comment

Popular posts from this blog

Setting up a boxplot chart in Google Sheets with multiple boxplots on a single chart

Image
The  Google Statistics  add-on for Google Sheets can display multiple boxplots in a single chart. The key is the layout of the data. One column should be the variable by which the data is to be grouped, the other column should be the data to be box plotted. Set up the Statistics add-on with the data to be plotted as the variable, and the grouping column as the "by" variable. In this image I had deselected all but the boxplot option, the result was the appearance of the Moment, Standard errors, and Confidence intervals options. The default is apparently a 95% confidence interval for the mean. The result is multiple boxplots on a single chart with a common scale. The new tab that is created also quotes 95% confidence intervals for the mean. Note that as of 2018 the Google Statistics add-on cannot be found by search in the add-ons. In addition, as of May 2018 the add-on no longer verifies, possibly due to the add-on not having been updated since August 2017. One may ha...
Read more

Traditional food dishes of Micronesia

Image
Clara, Jennette, and Joanie presented ground hard taro cooked in an umw, effectively a hot rock oven. Pohnpeian: rotama. Pingelapese: sero. Mwoakillese: rodma. Mortlockese: Amahd. Kosrae: rodoma. After cooking in an umw the rotama is pounded with the petiole of a palm frond. Rotama is made from hard or swamp taro, known as mwahng on Pohnpei, mweian on Pingalap, pula in Mortlockese, and pahsruhk on Kosrae. In traditional times on the outer islands the women had primary responsibility for tending the taro while the men handled fishing and climbing tasks. Joesen and Noeleen presented fermented breadfruit from the outer islands of Yap such as Woleai and Eauripik. The raw ingredients were actually brought in from Yap, Noeleen prepared the maare.  The leaf wrap is ti leaf - Cordyline fruticosa. Pauleen, Barnson, Trisha, Verginia, Con-ray, and Maylanda from Kitti, Pohnpei brought in mahi umw, koahpnoair koakihr, and uht sukusuk. Above is breadfruit umw, mahi umw, o...
Read more

Experimenting with the PlantNet and iNaturalist apps

Image
The application of image recognition technologies as the back end to botany apps is providing increasingly relevant search results. Over the recent winter break I began looking at whether the apps now available might be useful to the ethnobotany class here on Pohnpei. Some apps had a more regional focus in the past and were better at north american or european plants. Pohnpei has a both unique indigenous plants and many introduced plants from around the world -  a challenging environment for any image recognition system. A student working with a plant identification app for the first time My criterion was that the app had to be appropriate for use in my classroom. The app could not contain advertising and the app had to be free, my students typically do not have the ability to pay for an app. The app would also have to be fairly small as my students often use phones with limited memory and processing power. I settled on PlantNet and was impressed with the ability of the a...
Read more