Brent Warner (0:00): How can we use notebook LM in higher ed? We’re discussing some options for you as our school opens up this tech to the campus. This is the HigherEdTech Podcast season seven, episode 10.

Tim Van Norman (0:13): Welcome to today’s HigherEdTech Podcast. I’m Tim Van Norman, the Interim Assistant Director of Technology Services at Irvine Valley College and Adjunct Professor of Business at Cypress College.

Brent Warner (0:36): And I’m Brent Warner, Professor of ESL here at IVC. We both enjoy integrating technology into the classroom, which is what this show is all about.

Tim Van Norman (0:44): Welcome. We’re glad you’re here with us.

Brent Warner (0:47): All right, Tim. So as this episode drops, our semester is now over. But there are some changes coming for everybody listening, which is education access to notebook LM, is that right?

Tim Van Norman (1:05): At IVC, yes. So Google has offered Gemini and notebook LM for a long time, but a little while ago, back at the AI show—and we talked about this a little bit—we found out that they were now FERPA compliant if you’re in the Google workspace for education account. And so we’ve been since then working with our district to try to get something turned on for us. And it actually was just turned on.

Brent Warner (1:41): So if you’re listening to this in mid-December or just as we’re going into winter break here, we’re hoping that maybe some of this information is like, okay, it’s something I can look at, explore, see how it might work for my classes in the spring semester. Not to say that everybody needs to use this—there’s people who are going to be opposed as we know, and people who are going to be into it as we know. But it’s worth knowing what this is, and it’s available. And a lot of students are using notebook LM to help them build themselves study guides and things like that.

Tim Van Norman (2:17): Absolutely. And for people outside of IVC, some of you have had it for six months, nine months, a year, and some of you can look forward to hopefully having it soon. But that’s part of why we wanted to go through this, because I think there’s a lot of people in the same situation where, well, I don’t really know what notebook LM is, so what do I do? It’s yet another AI, right, just like all the rest?

Brent Warner (2:45): Yeah, yes and no. More focused, I think. We can call it like a focused AI might be a way to talk about it, but let’s jump into it, Tim. There’s lots to discuss. So here we go.

Tim Van Norman (2:58): So let’s start off with Gemini notebook LM—what’s the difference? So they’re both part of Google workspace for education. They’re both FERPA compliant. What’s the difference? I’m giving you similarities. But notebook LM is trained using the data that you give it. Gemini is trained with all of the information that Gemini has access to, and then it will use information you give it as well, but it’s got everything else, right? So one of the key things to remember is, if you want something that’s about accuracy, you would use notebook LM. If you want something that’s very creative, use Gemini, or chat GPT, or something like that.

Brent Warner (3:49): You can kind of get a feeling for this once you’re inside of notebook, because it’s a lot more organized. I think there’s a column on the left, and this is where you specifically upload or link your sources. So hey, here’s the PDFs that I’m running for my class, or here are links to YouTube videos that we’re discussing, or here’s the website with the information we’re looking at. So it pulls from the direct sources that you give to it, and it uses those for the information. And it really restricts how creative it’s getting with the answers and conversation. So as you mentioned, Tim, that creativity side is more for when you’re in Gemini or your general chat bots. This one is much more like, hey, the real goal here is to kind of help you study and figure out what’s going on with these ideas. So if you haven’t seen notebook before, before we get too deep into it, Tim, we should say there’s a couple of sections. It’s broken into three columns where there is your traditional AI chat bot in the middle, and then off to the sides—on the left hand side is your sources and main information, where your research comes from, and then on the right hand side is all of the places where you can create interactive opportunities with the—not always interactive, but create new versions of ways to look at this information. So we’ll get into them in a bit. But there’s things like overviews and mind maps and reports and flashcards and quizzes, etc., right? So we’ll talk about those all in a moment. But that’s kind of the basic idea, is you can really focus in with specific content and then use that as a way to get yourself studying and deeper information on what you’re looking at.

Tim Van Norman (5:44): Absolutely. And as we talk about specific content, a single notebook can handle approximately 25 million words compared to—I’ve seen several times when I run out of, you know, you’ve got to pay for lots of tokens to get to anything useful sometimes. So you’re taking one chapter and doing something on a chapter in Gemini or chat GPT or something like that. And here, 25 million words is a lot. It also has a 50 source limit per notebook. So 50 sources is a lot of information. Think about that type of thing. When you’re thinking about where is this useful, it’s probably not going to be useful where you’re having somebody create a chat with somebody and you need to upload all of that information and have a chat with a historical figure or something like that. That’d be great for Gemini. But if you want them to look through your class and see from your transcripts, ask questions, this would be a great place to have that. And much more accurate.

Brent Warner (7:09): So it’s kind of like that tutor bot conversation around things like that, for sure. So yeah, a lot of that could be really, really good for getting started with things. Again, I think Tim, I want to still be wary of the hallucination thing, right? Because it’s still generating text, it’s still generating language. So even if it’s pulling more accurately because you’re saying a limited amount of sources, it still has to by definition, because that’s all AI is—it’s auto-correct on steroids, that’s how they always phrase it, right? So there’s going to be problems with the output at some level. So I would still consider, hey, make this stuff formative if you’re having students do work. Don’t use this as the way to 100% guide and just say you’re going to do all of your work inside of here. But at the same time, if you’re saying, hey, low level practice, like getting your information going, etc., lots to play with here. So Tim, before we get into the actual features, let’s talk about some ideas. We’re kind of doing this backwards, I think, from how we normally do, but like some of the ways that we can maybe use it. So we can say, oh, okay, let me think about that. And then we’ll talk about what the features are, so that you can kind of say, okay, now I can expand on what some of these ideas are. So I think you had your first idea here.

Tim Van Norman (8:31): Well, first of all, think about data. If you want to analyze data, one of the things I’ve had trouble with is, how do I really get it into chat GPT or something like that? Well, notebook LM is designed for handling data. It can handle, by the way, a website—you don’t have to have just documents. You can have a reference point of a website as well. But like I said before, uploading lecture transcripts for a tutor system is a great way to do it. But one of the things that I’m interested in—and I just finished my semester teaching this semester, so I haven’t uploaded this yet—but I want to take the, I have the students do a survey at the end, a reflection on the class, right? And I would really like to upload all of those into the system. And now, instead of it being me, my subjective view of and remembering what students said, I really like to see, to have an analysis that says, oh, this is what people liked, this is what people didn’t like, this is what people suggested to change, and to get some metrics on it. And now that I’ve gone through three semesters, I could actually do that for all three semesters and really get a feel across the semesters. And I think that would be really, really useful. And that’s a perfect use for notebook LM.

Brent Warner (9:54): That’s great. And because the information is private, you don’t have to worry as much. I would still probably pull your—if there’s any names on there, I’d pull them out. But the data and whatever the information, that’s great. So cool. Okay, so one thing that I was thinking about here, Tim, is we know that there are a lot of teachers who are, they’re subject area experts, but they’re not really pedagogists, right? They’re not really into what does it mean to teach? What does it mean to do all these things? And so they might have had kind of a more traditional education themselves, and then they brought that to their own teaching, which is like, I’m going to put the information on a slide, I’m going to review the slides, and then we’re going to be done, right? And it’s like, well, okay. So there are, now these days, I feel like there are a lot of teachers who are recognizing, okay, hold on a second, I can’t just be presenting information. I actually have to teach and do the learning process. And so one of the things that I was thinking about is, as a teacher, you could upload pedagogical journals or approaches to classroom management, or ways to better engage your students, or things like that as resources. And then you can have it in two parts, right? The actual pedagogy ideas as one set of resources, and then your classroom content as another set of resources. And then say, hey, help me build a better, stronger, more engaging, pedagogically driven classroom lesson for today based on these resources. And so it’s like, okay, so then that can help the teacher who maybe, if they’re wanting to improve their classroom teaching interactions directly with the students or online, right? If you’re doing it online or asynchronous, or whatever else it is, they can get a little bit of a better sense of like, well, I’ve heard these ideas, but I don’t know how to apply them to my classroom and to my setting. I read this cool thing, but it seems like that’s for a math class, not for an English class, right? But it’s like, okay, hold on, this can then help you extract those things. And of course, you could do that in other chats, right? You can upload sources and ask it to do all those things, but once you’re kind of setting it up with the resources that you like, I think that can guide you a lot more specifically into what you’re trying to do. So this is just purely back end, and again, we all need to analyze and look and see, like, is that going to work? And what do I think about it? You know, actually test it in your classes, right? As teachers, that’s what we’re always doing, is trying to find better ways to do things. But that’s a fully back end side of something that you can do with the synthesis of large amounts of data.


Tim Van Norman (12:34): And along those lines, remember, when you’re uploading to Gemini or something like that, you’re uploading all of that information each time. And notebook LM, you put the information in, and now you can say, oh, okay, out of this information, I want to create a class on this topic. And now then you can go back a day later, a week later, an hour later, and say, okay, now using that same information, how about on this topic? And you can just keep on asking it questions, because it basically is trained on what you’ve uploaded. So that’s the real useful part of it. It’s not the one time question, it’s that multiple going through it and analyzing or going through it and being able to revisit that same information, that same data. So one I saw that was really interesting is for teaching. Now, there’s two parts to this. I don’t want to use it for infographics for learning, but for teaching. I hate creating infographics. I love them. I think they’re really useful. I hate doing it because I’m not creative that way. I can come up, I can draw, I can draw this, to this, to this, to this, to this, and it looks like a flow chart, right? And it just isn’t appealing. But it’s got the ability, once you’ve loaded your class information, for instance, in there, you could create a flow chart, or you could create an infographic out of that too. And that type of thing, you already know the data, but it’s sometimes putting it together in a way that is more appealing to your students.


Brent Warner (14:22): Also maybe kind of reveals information to you in a way that you wouldn’t have looked at if you’re just reading through things, right? So I do like that a lot. The infographic feature is in beta right now, but we’ve been seeing a lot. I’ve been seeing a lot of these popping up on LinkedIn recently is like, everybody’s showing their infographics on what they’re uploading. It’s very cool and beautiful. And so the whole new Google nano banana thing, like the graphic generator AI that they’ve got, is making some really interesting things. So yes, for sure. If you’re saying, hey, I just want to have a different way to look at this information, visualize it, let’s do that. But Tim, as you were mentioning, and we kind of talked about before, there’s also a value to students creating, creating fully their own infographics and slowly dealing with the data. Not just like, hey, I just want the output. This is always the AI problem, right? Is like, just give me the output. It’s like, no, hold on. The learning is in the process of doing the work, right? And so I wouldn’t necessarily recommend that—like, ask students to upload stuff and have it make an infographic, and that’s their assignment. That’s not going to do anything for them. They don’t even have to think. But that would be more like, go to Canva, build out actual infographics, show all of your work in process, etc.


Tim Van Norman (15:43): Exactly.


Brent Warner (15:45): Cool. So, and then my last little idea here, I’ll go quickly with this. And this is kind of one of the main, main, main features of it too. There are conversations. So you can click a button and they will make a podcast between two very handsome voices, a man and a woman that—they just sound good looking to me, but they’re fake. They’re not real. They’re not real people. But they have these conversations, and they kind of synthesize the information. They give you a summary. They kind of talk about what’s going on. But there’s a cool interactive feature. You can click a button, and then basically what happens is, you can say, hey, I want to jump into the conversation. They’ll be like, oh, hey, we got someone listening in. And then you say, hey, can you talk a little bit more about whatever you were just saying? And then they’re like, oh, yeah, that’s a common misconception. Let’s talk more about it, right? And so they will actually interact with you as you’re listening to a podcast, which is pretty amazing, right? Like, I mean, that feature alone, it’s just like, oh my gosh, that blows my mind, because they do a good job of pausing, listening in, weaving that into the conversation, then weaving their way back to the original one. It’s just unbelievable. And for people who do well with talking through their ideas and trying to say, like, hey, what does this mean? Let me get more clarification. What about this? Would it still work if I did that, right? Those types of things, it’s like, that’s a really great sounding board for a lot of people who maybe don’t have as much time, or whatever else it is to interact, or to meet the teacher during office hours because they’re only awake, they’re working the midnight shift, or whatever it is. So I gotta say, like, I see the problems and I totally understand and agree, and at the same time it’s like, yeah, but there are some real great opportunities for people who need different levels of help in other ways.


Tim Van Norman (17:50): Absolutely. And that’s really what the strength of AI is. It gives you that ability to have things more customized for how that student learns.


Brent Warner (17:59): Yeah, that’s right. So Tim, I think we should kind of zip through now, now that we’ve thrown out a couple of ideas, and there should be more and more, but I think we should just kind of cover these features so that we can really just understand, like, hey, what’s in here? But let’s do it quickly.


Tim Van Norman (18:12): Yep. So first of all, source grounded AI. What that means is, it’s based only on the sources you upload, it’s not based on its own learning. So I think that’s really important to know. And goes back to that part I was talking about for accuracy versus creativity.


Brent Warner (18:31): Yes, and those sources can be things like PDFs, can be Google Docs, websites, slides, text files, direct links to YouTube videos. But Tim, you mentioned this. I said that before, and you said, hey, but be careful, because it’s not really the video. It is the transcript of the video, which, depending on how old that video is, with older technology, that kind of did not transcribe as well as the newer technology does. You just kind of have to be careful about what that means.


Tim Van Norman (18:59): Absolutely, absolutely. Also, when you upload, it automatically summarizes and generates a summary and identifies key topics for every source, automatically. Which is a really cool thing, but it’s also something you just want to look at and make sure that it understood what you were talking about, or, I mean, not that it understands, but make sure that it followed what you were expecting in that.


Brent Warner (19:30): So I mentioned already the audio overview of the content and the interactive chat inside of there. But there is also a feature called video overview, where it kind of makes like a video slideshow where you can get some visuals alongside with the audio. And so it’ll make a short, you know, four or five minute video. It’ll say, hey, here’s all the things. And then it’s not like a bunch of moving images all over the place. It’s very, very simple at this moment where it’s like clean slideshow with pictures that help support your ideas, and it kind of goes through one by one, but it is in there, so it’s pretty nice.


Tim Van Norman (20:09): Nice. So this is going to be something that you want to remember, but it’s also a cool feature. First of all, the feature, you can save your important or useful responses, right? Save them as notes. The thing to remember is you need to save anything you want to save because outputs in the chat interface are not saved automatically, right? We like that, but it’s also something to, hey, wait, take a look. Make absolutely sure that you have saved what you want to save, and it can be used to train itself. So if you like something that was output, you can put it back in to notebook LM, and use that as a source document later. But you need to actually do that work.


Brent Warner (22:27): Yes, yes. So you do the work, you save it, you kind of—so again, the whole thing, I think in the future, it will come where you can actually just go and edit the text directly inside of here. But it’s not there quite yet. That would be really nice for fine tuning. And especially Tim, for these last few things, I’ll just zip through these, but they have an option for building flashcards, for having quizzes, multiple choice quizzes. In beta, you mentioned infographics and also slide deck creation at full creation, which you could then interact with. All of these are great, but they need to be things in the future, in the near future, that you can go and edit and say, okay, hold on a second, this is not accurate. Let’s update it for more accuracy. But it is pretty cool. And then there are buttons inside of here. So the education thing, I think we need to clarify this, Tim, is you can go and share the actual notebook. For our educational institution, you can share it with people who have a matching school address, right? So they have to have an IVC.edu account in order to access my notebook that I want to share out. I can’t do the public doc like I’ve done with sharing a Google Doc, for example.


Tim Van Norman (23:40): Yeah, so it’s within our district. Anybody in our district can have access to it, and that may be—I don’t know for sure—but that may be a setting that we’ve set, versus a setting that is set by Google. Don’t know, especially with the FERPA compliance. I can see why they might go, you know what? The easiest way to handle that is to block it, and it’s all within its tenant. So just understand that that might be something that you run into at other places as well, or it might not. We don’t know at this point.


Brent Warner (24:00): A little bit of a technical thing here, but it is worth pointing out. There’s not really a single email right now that you can just say, oh, my whole class email. So even if your system has ways to blast one email address to my whole class, you can’t use that. You’ll have to have a list of all of your students’ email addresses and cut and paste it into here. But you can with any of those—they call them artifacts—so if you’re doing the infographic or the quiz or the flashcards, for example, once you’ve made that, you could just share out those flashcards with your students, right? So if you made it, you can give them access to one aspect of this notebook.


Tim Van Norman (24:32): There’s different parts to it. You can share a whole notebook with a link. You could share the artifacts with individual people. It kind of goes back and forth, so you’ll want to play with it in your environment. But yes, the ability to share is one of the most wonderful parts of this.


Brent Warner (24:48): Yes, and I think it will grow a little bit more intuitive in the future. It’s not the most right now, but you can get it out there. You can get that going. So all right, so lots of different features inside of here. I wouldn’t say we’re scratching the surface. Those are kind of the main things that are available right now, but we expect more to continue to grow. But Tim, we also have a few things. We’ve mentioned a few of these already, but let’s just make sure that we kind of hold these considerations in mind. I know you’ve got a few, and we’ll jump back and forth on these.


Tim Van Norman (25:06): So we talked about the creativity versus accuracy. This is about accuracy, not creativity. Although it will be creative, you do need to watch it, make sure that the output is accurate. The other part is the massive capacity. 25 million words is a lot. You can include lots of source documents. You can include emails, stuff like that. But it’s got a really large capacity for what it does. So those are two of the very important considerations.


Brent Warner (25:29): Yeah, okay, so the next thing, I think this one might be one of the most important things, Tim, is I think a lot of people are just going to be tempted to say, oh, I’m just going to share all the links that I can find on this subject, and then it’s going to get that information. Well, whatever you’re putting into it is going to be the quality of the work that’s coming out of it. So you would be much better off uploading two, three, four really high quality documents that you really trust, rather than going out there Googling, you know, Introduction to Business, grabbing the first 10 YouTube videos and the first 10 links out there that are just like, you know, Bob’s fun blog on business, right? And then it’s like, well, what is the quality of that information? So you should be curating good quality, trustworthy resources that you think are good. And the fewer that you have, I think, and this is kind of a neat thing to be played with, but the fewer that you have, the more limited the information that’s output is going to be. And therefore you kind of have a little bit more of a handle on what’s coming out of it. And you, as the person creating this, should be absolutely responsible for what you’re putting into it. And so therefore, when you’re saying, hey, what comes out of it, you know, we know that there’s going to be some making stuff up. I know we keep calling it hallucination. We should just call it making stuff up. And the less careful you are with the resources, the less trustworthy the output will be.


Tim Van Norman (26:39): Absolutely, absolutely.


Brent Warner (26:42): I think we kind of talked about everything here then, Tim, for the most part. We will have several—I wouldn’t say a ton, but we’ll have a handful of resources, some videos out there, some tutorials and things like that on how to get started with it, from Google, right? And from a few people who are engaged with this stuff. So at least to get yourself started, if you need some visuals to walk through some of this, I think there’s a lot of information. But what I would suggest is, we’re going into winter break. Give yourself 20, 30 minutes. Find a fun subject that you like, right? Something you know—Tim, you saw mine. I showed you the back end that I’m studying for my little ham radio license, right? And it’s like, okay, for me, I really—I realized that’s not fun for everybody, but whatever you’re interested or wanting to learn about right now would be a thing that would be useful, because once you see it from the perspective of a learner, then you can also consider it from the perspective of a teacher.


Tim Van Norman (27:42): Exactly, absolutely. Thank you for listening today. For more information about this show, please visit our website at TheHigherEdTechPodcast.com.


Brent Warner (27:53): As always, we want your feedback, so please go to TheHigherEdTechPodcast.com and let us know your thoughts.


Tim Van Norman (27:59): For everyone at IVC that’s listening, if you need help with technology questions, please contact IVC technical support. If you have questions about technology in your classroom, please stop by Library 213, and/or contact me, Tim Van Norman at tvannorman@ivc.edu.


Brent Warner (28:14): And if you want to reach out to me about the show, you can find me on LinkedIn at Brent G Warner.


Tim Van Norman (28:22): I’m Tim Van Norman.


Brent Warner (28:24): And I’m Brent Warner, and we hope this episode has helped you on the road from possibility to actuality. Happy New Year, everybody.