130 | Beyond the Buzz: Real-World Uses of Notebook LM and ChatGPT Canvas with Isar Meitis

0:00Hello, and welcome to Leveraging AI, the podcast that shares practical ethical ways to leverage to improve efficiency, grow your business, and advance your career. This is Isar Maitis, your host. And in today's episode, we're going to dive into two very helpful day-to-day use cases that are using recently released capabilities. One would be Notebook LM, which is the new and amazing summarization and research tool that came out from Google. And the other is ChatGPT 4.0 Canvas. So this is the side-by-side functionality that was just recently added when I say recently a few days ago. And when I say recently, just a few days ago, that was just added to ChatGPT.

0:38And both of them provide incredible day-to-day capabilities. So I'm going to give you one example of each, but we're going to generalize it for other things that you can do with it in your business. In the next few years, AI technology will change our world dramatically. Whether you are a business executive trying to catapult your business forward, or just somebody who refuses to be left behind and want to advance your career, this is the show for you. I'm your host, Isar Maitis, a serial entrepreneur and an AI enthusiast. You'll hear invaluable practical tips from innovative business leaders, AI practitioners, and some of the brightest AI minds in our world today on how you can leverage AI in ethical ways to advance your career and grow your business.

1:33Let's start with Notebook LM. First of all, to get to Notebook LM, you just type notebooklm.google.com and it will take you there and then you can sign in with your regular Google account and then it will allow you to upload information into a new notebook and you can create as many of these notebooks that you want. By the way, this tool is absolutely free, at least for now, which is insane and magical all at the same time. But what this tool does is it allows you to upload information into a notebook. A notebook is basically a repository of information on a specific topic. So the example that I'm going to give you today has to do with a new concept that I literally tested this past week.

2:16So I was speaking at a conference in San Francisco called AI Realize. The conference had some amazing speakers like Ted Shelton, the COO of Inflection AI, Jeremiah Ouyang, who is one of the founders of Blitz Scaling Ventures, which is the VC of Reid Hoffman, Helen Fang, the head of product at Airbnb, Tolga Kurtgulu, who is the CTO of Lenovo, and people at that scale. So it was a fascinating conference. In the audience were leaders of Fortune 2000 companies who came to learn how to implement AI at that scale of businesses and what to expect into 2025. Fascinating speakers, great conversations in the breakout sessions.

3:02So it was really a great and well done conference. So a lot to learn, right? So what do you do? And I was speaking just for one session, but the most of the time I was sitting in the audience and really sucking in all the information that came out of the other speakers, which was a lot of information. So how do you summarize all of that? And the same thing can be not just from a conference, but sitting in several different meetings on a specific topic in your company or looking at the next product design. So you have PDFs that are in a folder, you have a whole Slack channel of communication, you have recordings of several conversations, whether on Zoom or Teams or in person, all of that exists in several different formats in several different places.

3:42So what I took in this particular case, I took the information and the notes that I have taken pictures of all the slides of all the speakers. So I had those and I had the website of the conference that had additional information about the speakers and the conference and the topics and so on. And I uploaded all of them into Notebook LM. So what is the outcome of uploading all that information? I'm sharing the screen if you are watching this on YouTube or on LinkedIn, but if you are not, I'll explain exactly what's everything that's on the screen. So I have this new notebook, which is basically a webpage, and I've uploaded these documents.

4:23And now it offers me to do one of the following, create an FAQ document, create a study guide, create a table of contents, create a timeline and create a briefing doc. Now there's also a button that allows me to create this mini podcast. And the podcast is maybe one of the most incredible aspects. So what I've created so far is I've asked you to create two things. I asked you to create a table of contents and a summary of the thing. So if I click on the table of contents, it will give me a table of contents in several different levels of bullet points. So in here, I've got, this is the table of contents organized by key information, the event materials for the AI Realize 2024.

5:07And then the first thing is AI Realize 2024 Summit Overview, event focus, target audience, key themes. Then the second segment is speaker insights and summaries. Then he talks about the keynote speakers, tells about what each and every one of them talked about in a couple of sentences, and then the panel discussions, which happen in the afternoon. And then it goes into each and every one of these panels, and then it adds more stuff after that. And then strategic planning for 2025 and beyond as a third section, and then overall takeaways. So a great table of contents, but each and every one with one or two sentences that I can learn from.

5:42So if somebody asked me, how was the conference? I can literally send them this, or if I want to post about this, I've got a great summary that already exists. The second document is a more detailed document with a specific topic in mind, and this one has a slightly different structure and a slightly different format, gives me more information about one of the topics. I can also create a study guide by literally a click of a button, and then we'll create a Q &A for me that I can go back and forth and test myself if I know the material that was there. If it's something I want to learn more about, let's say this was a code review or a product development or a board of directors meeting or whatever the case may be, I can go back and send this to people so they can test themselves to see they've actually learned the materials or even just learning for a test.

6:29I can throw all the test information in there, create summaries, and then have it create questions for me to test myself. As I mentioned, this could be used for almost anything you can imagine as far as learning and researching. I want to let you listen to the little mini podcast, and then I will share with you some other use cases and how I started using it daily. So what you're going to listen to right now was created by literally me just uploading the information and clicking a button. I didn't prompt it. There is no way to prompt it. I literally just upload the information and clicked to create this podcast.

7:06And this is what it generates. I'm going to let you listen from the beginning and then we'll jump to a different section and listen to that. Okay, everyone's talking about AI, but we wanted to get past the hype and really dig into how companies are actually using this stuff to get things done. We did a deep dive into all the materials from the AI Realized Summit, speaker notes, those insider sessions, and let me tell you, this wasn't your typical tech conference. It felt different. Imagine a room full of like really switched on business leaders, all wrestling with the same question. How do we ride this AI wave instead of getting crushed by it?

7:45Makes sense. You've got the stock market going bonkers 50%. So I'll pause it here, but you understand this is a very conversational outcome of a very dry input. So again, the input that I gave it was my notes that are by any means are not in proper English, just me taking notes while people are speaking. screenshots or not screenshot, but phone photography of slides on the screen and a website. And what comes out is this amazingly conversational podcast of two people geeking about the topic, which is absolutely nuts. I'm going to jump ahead to minute 13 and a half. The whole podcast in this particular case is 24 minutes.

8:23I've seen it go as short as five to six minutes, and this is probably the longest that I've seen so far, but it will generate it as long in order to cover the content with enough depth. So I'm going to hit play on where we are right now. Honestly, the experts seem to agree that 2025, that's going to be the year we see a lot of these no-brainer AI use cases become really widespread. No-brainer meaning? Low risk, clear ROI, the stuff that just makes sense. And if companies aren't doing that by then, they're going to be behind. It's like an AI arms race. So who's winning so far? That's a great question.

9:00Interestingly, it seems like smaller businesses. Really? Yeah, the ones with those lower cost support centers. Okay. And also private equity owned businesses. They seem to be ahead of the curve. Because they have to be more agile. They're under more pressure to innovate, to optimize. So they're jumping on the AI bandwagon faster. Makes sense. Because if your competitors are using AI to cut costs. And I'm going to pause it because this is my podcast, not their podcast, but it's absolutely amazing. amazing. So if you're somebody like me who A, enjoys consuming content through podcasting, or you just are like me or not exactly like me, but you like to consume this kind of content when you cannot do something else, meaning while you're driving, while you're walking your dog, doing the dishes, working the yard, folding laundry, whatever it is that you're doing, and you can't be in front of the screen, then this is an awesome way to consume content.

9:53So how do I use it now? Literally almost every single day. I review a lot of AI information all the time. I read articles. I check out different stuff from different people, post on LinkedIn, and so on. And not always I have time to read through the whole thing. So now what I do, every time I see an interesting article that I want to learn a little more about, I will literally just grab the link, dump it into Notebook.lm, give it a title that makes sense to me, and click the Create Podcast button. So literally every single day, I've got two or three of these short podcasts waiting in my queue.

10:25And then when I jump in the car to go and get the kids to school or to pick them up from somewhere or to go shopping, I will go back to Notebook LM, find one of the topics that I want to listen to, and we'll click play. And within five minutes, you get a great summary of what was talked about. Now, I will say something about this method because I've been using it a lot in the past few days. in the first time on the first two or three times you do this it's incredible just the concept of this highly conversational really engaging way to consume content is fascinating once you do this 15 times you understand there's a little too much fluff in the way they deliver the content and i wish there was a way to control it i really hope that's what's going to happen in the next step, but so I don't have too much fluff in my podcast.

11:13I'm going to move ahead from this topic. Quick summary on Notebook LM. You go to notebooklm.google.com. You start a new notebook. You upload whatever information you want. And this could be a mix from multiple sources in your company. It could be a Slack channel, files, link to a website, link to a YouTube channel, all of the above combined together. And then you can create quick, very accurate, detailed summaries, FAQs, questions, briefing documents, and a podcast with a click of a button. But there's also a way to also just ask questions and query. It's just like a regular chat bot would on the information that you uploaded.

11:54So there's a regular chat interface at the bottom of it where you can go ahead and type whatever you want to type and get specific answers about the topic. It's literally changing completely the way I learn new content and the way I summarize information. And I highly recommend you check it out. We have been talking a lot on this podcast on the importance of AI education and literacy for people in businesses. It is literally the number one factor of success versus failure when implementing AI in the business. It's actually not the tech. It's the ability to train people and get them to the level of knowledge they need in order to use AI in specific use cases successfully, hence generating positive ROI.

12:40The biggest question is how do you train yourself if you're the business person or people in your team, in your company, in the most effective way? I have two pieces of very exciting news for you. Number one is that I have been teaching the AI Business Transformation course since April of last year. I have been teaching it two times a month, every month since the beginning of the year, and once a month all of last year. Hundreds of business people and businesses are transforming their way they're doing business because based on the information they've learned in this course. I mostly teach this course privately, meaning organizations and companies hire me to teach just their people.

13:24And about once a quarter, we do a publicly available course. This once a quarter is happening again. So on October 28th of this month, we are opening another course to the public where anyone can join the courses, four sessions online, two hours each. So four weeks, two hours every single week with me live as an instructor with one hour a week in addition for you to come and ask questions in between based on the homework or things you learned or things you didn't understand. It's a very detailed, comprehensive course that will take you from wherever you are in your journey right now to a level where you understand what this technology can do for your business across multiple aspects and departments, including a detailed blueprint of how to move forward and implement this from a company-wide perspective.

14:16So if you are looking to dramatically impact the way you are using AI or your company or your department is using AI, this is an amazing opportunity for you to accelerate your knowledge and start implementing AI in everything you're doing in your business. You can find the link in the show notes. So if you just open your phone right now, find the link to the course, click on it, and you can sign up right now. Now, the other piece of news is that many companies are already planning for 2025, and we are doing a special webinar on October 17th at noon Eastern. So that's a Thursday, October 17th at noon Eastern.

14:58We're doing a 2025 AI planning session webinar when we are going to cover everything you need to take into consideration when you're planning HR, budgets, technology, anything you need as far as AI implementation planning in 2025. We're going to cover all the things you can do right now. So still in Q4 of 2024 in preparation to starting 2025 with the right foot forward, but also the things you need to prepare for in 2025. If that's something that's interesting to you, find another link in the show notes that's going to take you to registration for the webinar. The webinar is absolutely free. So you're all welcome to join us.

15:39And now back to the episode. But as I mentioned, there is something else that happened this week. And that something else is that OpenAI introduced a new functionality that they call Canvas. So what is ChatGPT Canvas? ChatGPT Canvas is very similar to Claude Artifacts. Now, those of you who don't know what Claude artifacts is what I just said is not going to help you whatsoever. So let me explain. On both these scenarios, you are writing the chat, the regular chat, like we always used to on the left, but you're seeing the outcome on the right. So if you ask it to write a document for you, instead of the chat keeping scrolling and you miss your prompts and so on, you're prompting on the left, you're getting very short responses.

16:27Oh, I tried this, I changed that. But then the actual content shows up on the right. The other use case that is extremely helpful is when you create code. When you're writing code, then your communication with the chat and what the chat is telling you that it's going to do shows up on the left and the code that it's generating shows up on the right. So the same thing is happening both in Cloud as well as in ChatUPT Canvas now. So Cloud has been out for a few months. ChatUPT Canvas just came out. You can choose it from the drop-down menu of which large language model version you want to use. So now you can pick ChatGPT 4.0 with Canvas, and that gives you that Canvas functionality.

17:08So that's how these two are the same, but there are other means in which they're very different. So let's talk a little bit about the biggest differences between the ChatGPT capabilities and CloudArtifacts. The biggest benefit of CloudArtifact is the fact that you can see the output or the outcome of the code if you're creating code with it. Now, I don't know how to write code. I don't understand how to read code. And yet I'm generating more and more code with these tools. So you can ask it to create code that does something for you, create a dashboard, create a user interface for something, like create an active way to use like a web page based on data that you have or on documents that you have and just make them interactive and so on.

17:54And it works extremely well. If you do this within Cloud Artifacts, it will actually show you the output within Cloud Artifacts. So on the right side of your screen, you'll be able to toggle between the code and the outcome of the code when it actually runs the code. This is extremely useful, but it's also very limiting because if you want other people to use that code, they need access to your Claude account. So the way around this, you can obviously copy the code and drop it somewhere else and then run it that way. But the benefit is you can have it right there and then. The ChatGPT interface does not provide that.

18:32Meaning ChatGPT Canvas, you cannot see the output of the code. You can just see the code itself. However, they do have some amazing benefits. So if I click on the bottom right corner, I will have several different options. Option number one is to add comments, and it will add comments to the entire code or to just a section that I highlighted. Option number two is to add logs. Option number three is to fix bugs. Option number four, which is completely magical, is called port to a language. And if I click on that, I can translate basically my code. Right now, the code that I'm looking at is Python, but I can translate it to C++ or PHP or JavaScript or TypeScript or Java and so on.

19:19So you can literally translate your existing code that whether you created or it created for you to different languages related to the use case that you need. So this is absolutely magical. The other thing that you can do is you can highlight a section and then a little tooltip pops up that's called Ask ChatGPT. And then you can ask ChatGPT about that section and you can have it change or massage or explain to you what's going on just in that section, literally by highlighting it and clicking a button. What are the use cases of that? They're endless. If you're creating a document, then it becomes a lot easier to write the document.

19:59And over there, you will have different functionality in the pop-up menu, which we're going to discuss in a minute. But right now, I want to dive in with you on a specific use case that I just tried for the purpose of this podcast. So what I was trying to create, I'm looking at different companies and their success tied to AI. In this particular case, I was looking at Microsoft and I wanted to be able to look at Microsoft financial data in the last few years and see how the introduction of AI has impacted their fiscal success and their stock market success as well. So what I did is I wanted to create a dashboard that will allow me to see and analyze their financial reports as they reported them to the SEC.

20:40And so what I did is I've downloaded the reports quarter by quarter in an Excel file, and then I've uploaded them to ChatGPT together with the type of charts that I want to see. So examples from other websites, but I wanted to make it interactive because just creating the chart is not interesting. So then I wrote the following. You are an expert financial data analyst and data presenter. You are also an expert code writer. I'm about to provide you an income statement quarterly as an Excel file. I would like you to create an interactive and easy to follow dashboard that would allow to get insights easily from the file.

21:17The dashboard should generate a graph like the attached one. So I give it an example of what kind of information I'm looking for. I could have attached more than one and it would have generated more than one. The page will allow the user to upload a file or several files in the same format as the attached Excel file that is labeled quarterly data. The dashboard should allow the user to select specific years or specific quarters as a multi-select. I would also like to be able to select which parameters would show up, such as revenue, operating income before taxes, net income, etc. Pay attention to the formatting of the quarter names.

21:59Please create the format as QQ-YY. Example, Q3-19 would mean third quarter of 2019. Please take your time and create the code for the dashboard. And it did. And it's creating an amazing code for me. The problem is I don't have a clue how to read code. And to make it worse, I don't have a clue how to run the code. So it created Python code that presumably does this. So then what I did is I literally just asked it. I said, okay, how do I run this code? And it gave me the step-by-step process on how to install Python on my computer, how to install the right libraries for it to run. And the first time it didn't, and then I told it what doesn't work.

22:42And then it told me what else I need to install and how to install it. And now I'm pretty much an expert at writing, at running Python code on my computer just because I've been playing with this for 45 minutes. But in the end, after some bug fixing, and when I say bug fixing, I obviously didn't fix anything. So what is the outcome of all of this is that I have a web page that I can go to whenever I want that allows me to upload files. So on the top of the page, it says drag and drop selected files here, and I can drop files as long as they're in the same format, meaning now I can take every formal Excel file that comes out of Microsoft SEC filing and drop it in there.

23:27And what it generates a dashboard in which I can select what information I want to have in the dashboard. So in this example that I'm looking at right now, it's showing revenue and gross margin and net income. But if I go to the drop down menu that it has created, I've got product, cost of revenue, research and development, sales and marketing, general administration, operating income, et cetera, et cetera. Basically all the parameters that exist in the income statement of every company. And so I can choose which one I want to show in the graph. I can select specific years and I can select specific quarters.

24:06And if I add them, it's going to add them to the graph and show them side by side next to the others. and it adds the parameters of each and every one of the tabs and so on and so forth. So I can now compare and mix and match different data from filings of companies to the SEC because they're all going to be in the same format. And I have a fully customizable dashboard that shows it to me. And it took me minutes to create, and I don't know how to create code. Now, the reason I really like this solution more than I like the Claude solution that we talked about before is that now it's a webpage that I can share with anybody who wants to use this thing.

24:50And all they will have to do is drag files in the same format in there and they'll be able to use this dashboard. So this functionality gives people like me who don't know how to write code to create different kinds of software to do different things. Again, this dashboard is just one example. I could write a code that will do basically anything that I want. One of the things that I use this for is to connect different automations that I create to the APIs of large language models. How did I do this? I uploaded the API documentation to ChatGPT and I asked it to create the code for it. It gave me the code.

25:28I ran the code and now I can connect it to whatever platform that I want. And so what these tools give us is the capability to now have a much easier and user-friendly user interface than we had before. And we can actually see the outcome. And we can have, as I mentioned, a tooltip and multiple options in the ChatUPT option to come and make changes either to text. If we're creating a post, a blog post, a summary of something, and so on, or a proposal, any kind of text, you'll be able to edit it and add things to it straight within the ChatGPT user interface. Or if you're writing code, and again, those of you who never written code before, I never wrote code before ChatGPT.

26:16I don't know how to read code still this time. So every time it creates a bug or something doesn't work, I literally just copy the error that I'm getting, or I'm explaining what's not working, and it's going to fix it for me. And then I'm going to post a new code. I will refresh the website and I will see whether it's working or not. That's it for today. I really hope you found these two use cases helpful. This functionality in ChatGPT is now available to all paying users. And they said that they're going to make it available to the free users as well, but it's not available to those people yet.

26:53By the way, if you're not paying for ChatGPT, start paying for ChatGPT. It's worth every penny for the value that you're getting, even if it's just for the functionality I described today. That's it for today. If you found this podcast helpful, please open your phone right now and rate this podcast on your favorite podcasting platform, whether it's Spotify or Apple Podcasts or whatever it is you're listening, if it has the ability to do that. And at the same time, click the share button and share it with a few people who can benefit from it as well. I will really appreciate it if you do that. And don't forget, we are holding a webinar on the 17th at noon Eastern that will talk about how you can plan for successful AI implementation in 2025.

27:43If that's something that's not on your radar or on your company's radar yet, You need to start working on this and we're going to give you all the tools and all the information and all the processes and everything you need to know in order to plan properly so you can start benefiting from AI implementation as soon as possible.