I’m enjoying my school holidays now, but when the chance arose for me to deepen my AI understanding with regard to Google NotebookLM, I seized it.
I have had some experience using this generative AI tool. I think that the edge it offers over other similar tools is its ability to produce podcasts. However, despite my prompting, it always generates podcasts that are at least 5-6 minutes long. Which mayn’t seem like a big deal, but when you teach teenagers who are drunk on dopamine-immersive experiences, it’s difficult for them to focus on a podcast that long. So, it was with this enquiry in my heart that I arrived at the workshop venue.
The facilitator, Mr Chan Kuang Wen, used Padlet to structure his workshop. You can view it here and follow his thought process. I thought his systematic and controlled way of bringing us from one aspect to another was commendable.
Right off the bat, we were given 15 minutes to explore any topic that we were interested to know about. I didn’t even realise you could use it like that. I had always uploaded my documents onto NotebookLM and gotten it to generate podcasts/study notes/mindmaps. However, I didn’t hesitate for long. A topic close to my heart is inculcating financial literacy in my son. So, that was what I keyed into the system.
Instantly, NotebookLM collated ten sources that seemed legit:
I could then ask it questions that would elicit the key learning points from all these sources. Like what you would use ChatGPT. I did pick up something useful from its suggestions - find a transparent piggy bank to motivate my son to save money since he can witness his savings grow bit by bit. That’s something I have not thought about!
Tried to generate a podcast. Specifically asked for a short one, but it turned out to be six minutes long. A fellow workshop participant was working with 45 sources; his podcast turned out to be 45 minutes long! I guess the duration of the podcast is somewhat directly correlated to the number of sources. But if I decide to narrow the scope of the sources, I will probably get a shorter podcast - but the responses I receive might not be very comprehensive. Thus lies the dilemma when using NotebookLM.
This is how we can use NotebookLM to generate resources in a nutshell:
Subsequently, the workshop took on an unexpected direction. Kuang Wen first taught us to use the sentence starter “Yes, and…” and allow for the divergent generation of ideas.
He posed us the question: How might we use NotebookLM in unique ways that are rendered unavailable by similar LLM tools? We were then given Post-It notes and instructed to pen one idea per Post-It. He encouraged us not to censor ourselves since our first 7 ideas are bound to be crap anyway,
Okay, I’m not one who can just let myself go. I only came up with four ideas that I thought were reasonably good.
After the individual brainstorming process, we were asked to consolidate our ideas. This is what my group came up with:
We then evaluated the feasibility and pedagogically soundness of our ideas. Since 3 out of 5 of us had written something related to study notes, we applied the starter Yes, and… to build up on each other’s ideas:
After that, Kuang Wen introduced this to the SAMR framework that is apparently commonly used in the tech world:
He prodded us to think deeper and at the transformational level. After all, something like study notes is a function that can be adequately fulfilled by other platforms as well. But what is NotebookLM’s USP, really?
Don’t worry, he didn’t keep us hanging for long. He proceeded to give two suggestions, both of which had been penned down on my Post-It notes.
He suggested that NotebookLM could 1) facilitate interdisciplinary learning because we could input articles that dissect a topic from different perspectives, and 2) help us compare Asian vs Western points of views.
Before long, the workshop was coming to an end. I thought Kuang Wen provided me with an additional angle with which to tackle things. Can a tool enable me to try out new solutions? That tweak in my thinking process might provoke a tsunamic shift in my work outcomes. Who knows what the future might bring?
I hope that my reporting was clear and coherent for you!
"Think about one topic that you would like to learn about". However, that is not how one judges the usability of cognitive solutions. Instead, think about the topic you are most confident about your own knowledge on.