WEBVTT
00:00:02.319 --> 00:00:05.919
How long has it been since we featured Notebook LM on this show?
00:00:06.160 --> 00:00:10.400
Apparently, so long, well, it isn't even called Notebook LM.
00:00:10.560 --> 00:00:11.519
It has a new name.
00:00:11.759 --> 00:00:15.919
Yes, the tool once known as Notebook LM is no more.
00:00:16.160 --> 00:00:20.800
But don't worry, one of my favorite AI tools ever didn't go to the Google graveyard.
00:00:20.960 --> 00:00:27.839
It actually just graduated and got upgraded into the official Gemini lineup and is now called Gemini Notebooks.
00:00:27.920 --> 00:00:30.399
But the name isn't the only thing that's new.
00:00:30.559 --> 00:00:41.119
That's because over the last month or so, the team at Google has rolled out a bunch of new features that legit supercharge Gemini Notebooks and make it a GENTIC by default.
00:00:41.280 --> 00:00:45.439
Which sounds really cool and all, but well, why haven't we been talking about it?
00:00:45.679 --> 00:00:50.240
Because many of those upgrades were only for users on the expensive Ultra Plan.
00:00:50.799 --> 00:00:52.799
But this past week that changed.
00:00:53.039 --> 00:00:59.359
Now the updates that have completely changed the face of Gemini Notebook are available to all paid users.
00:00:59.439 --> 00:01:06.959
And that's why we're going to be showcasing the seven most important updates today as we put AI to work on Wednesdays.
00:01:07.120 --> 00:01:11.040
All right, so on today's show, here's what you're going to learn.
00:01:11.200 --> 00:01:16.959
You're going to know why Notebook LM is no more and why Gemini Notebook might actually make more sense.
00:01:17.120 --> 00:01:21.920
You're going to know the one important grounded change that you really need to be aware of.
00:01:22.159 --> 00:01:28.079
And you're going to understand the real reason Gemini Notebook going agentic matters more than ever.
00:01:28.319 --> 00:01:29.359
All right, let's get into it.
00:01:29.439 --> 00:01:30.959
Welcome to Everyday AI.
00:01:31.040 --> 00:01:33.359
My name's Jordan Wilson, and this thing's for you.
00:01:33.439 --> 00:01:42.319
It's your daily, unedited, unscripted live stream podcast and free daily newsletter, helping business leaders like you and me keep up with the nonstop avalanche of AI updates.
00:01:42.480 --> 00:01:47.359
I tell you what matters, what doesn't, how to use it, and you take that information to grow your company and career.
00:01:47.519 --> 00:01:58.000
So it starts here, but please make sure if you haven't already, please subscribe to the podcast on Spotify or Apple and make sure you go to our website at your everydayai.com and sign up for the free daily newsletter.
00:01:58.079 --> 00:02:05.280
We're going to be recapping the highlights from today's show in case you miss anything, as well as all of the other AI news and developments that you need to know.
00:02:05.439 --> 00:02:05.840
All right.
00:02:05.920 --> 00:02:08.080
So let's talk at Gemini Notebooks.
00:02:08.240 --> 00:02:10.159
Yeah, Notebook LM is no more.
00:02:10.319 --> 00:02:10.960
So sad.
00:02:11.199 --> 00:02:11.919
Uh right.
00:02:12.000 --> 00:02:13.840
But it's not gone.
00:02:14.000 --> 00:02:20.479
It just got a facelift, uh, both on the exterior and a lot going on under the hood as well.
00:02:20.639 --> 00:02:28.800
And this is one of those things I think over the last couple of years, a lot of people were kind of nervous because Google has done this before, where they've come out with great products, right?
00:02:28.879 --> 00:02:33.840
And this isn't just in the AI uh age, in the pre-AI uh phase, maybe even more.
00:02:34.000 --> 00:02:41.120
They've come out with very popular products that a lot of people love, almost built like a cult like following, and then they end up killing those products.
00:02:41.199 --> 00:02:49.039
So a lot of people were very nervous over the last couple of years as Notebook LM almost seemed like this side quest that was legit amazing, right?
00:02:49.120 --> 00:02:49.759
No joke.
00:02:49.840 --> 00:02:54.800
It's I think we named it like our AI tool of the year in 2024 and 2025.
00:02:54.960 --> 00:02:57.759
Um, so a lot of people were worried, like, hey, don't take this away.
00:02:57.919 --> 00:03:16.080
It's very uh it's very unique because there's no, you know, the other competitors, um, you know, no one from Anthropic, OpenAI, Microsoft, uh, you know, X, Groc, uh, Meta, like no one else has something like Notebook LM that really just grounds its outputs in your data only.
00:03:16.319 --> 00:03:18.639
So luckily, they didn't get rid of it.
00:03:18.719 --> 00:03:24.400
And it seems like, if anything, uh Gemini Notebook is gonna be along for the long run.
00:03:24.879 --> 00:03:27.759
So uh let's quickly go over, not gonna make you wait.
00:03:27.919 --> 00:03:30.080
Here's the seven new updates to Gemini Notebook.
00:03:30.159 --> 00:03:31.919
I'm gonna be going over them pretty quickly now.
00:03:32.080 --> 00:03:35.520
Uh, and we're gonna be doing some live demos so you can see these seven updates in action.
00:03:35.599 --> 00:03:38.879
And then when we're done with the live demos, I'll come back and talk about these things a little bit more.
00:03:39.120 --> 00:03:41.360
So, number one, new update, already talked about it.
00:03:41.520 --> 00:03:43.520
Gemini Notebook is the new name.
00:03:43.680 --> 00:03:44.319
That's number one.
00:03:44.479 --> 00:03:46.639
Number two, there's something called collections.
00:03:46.719 --> 00:03:49.759
That's a way for you to organize notebooks for easier sorting.
00:03:49.919 --> 00:03:52.560
Number three, and this one's actually really cool.
00:03:52.639 --> 00:03:56.159
Uh, it's now there's automatic Google Drive syncing.
00:03:56.319 --> 00:04:01.439
All right, number four, uh, bringing the Gemini and search access.
00:04:01.599 --> 00:04:10.479
So your um personal notebooks can now be accessed inside of Gemini and soon inside of uh Google searches AI mode, uh, which will be really handy.
00:04:10.560 --> 00:04:12.719
Uh number five, new output formats.
00:04:12.879 --> 00:04:17.120
So yeah, no longer uh constrained to only what Notebook LM offered.
00:04:17.199 --> 00:04:18.319
Now Gemini Notebook.
00:04:18.399 --> 00:04:23.439
You can create PDFs, PNGs, documents, spreadsheets, PowerPoints, markdown files, charts, and images.
00:04:23.600 --> 00:04:29.680
Uh number six, technically each notebook has its own secure cloud computer, uh, which is really cool.
00:04:29.920 --> 00:04:32.560
Number seven, just more agentic intelligence.
00:04:32.639 --> 00:04:43.279
And that's probably the biggest um upgrade that we're gonna see in Gemini Notebook now, powered by Gemini 3.5, uh reasoning and Google's anti-gravity agentic search.
00:04:43.439 --> 00:04:45.839
All right, so let's do it.
00:04:45.920 --> 00:04:49.279
Uh, let's take a look live, shall we?
00:04:49.439 --> 00:04:50.480
What could go wrong?
00:04:50.560 --> 00:04:52.560
Uh, I'm gonna share my screen, live stream audience.
00:04:52.639 --> 00:04:54.959
Do me a favor, let me know if you can see.
00:04:55.360 --> 00:05:01.759
All right, so what I did here is I took all of my notes from yesterday's show, uh, had a lot of them.
00:05:01.920 --> 00:05:06.319
So uh yesterday's show was kind of on RSI recursive self-improvement.
00:05:06.399 --> 00:05:08.000
Um, I have all of my notes together.
00:05:08.079 --> 00:05:11.600
And this is something I routinely do inside of Notebook LM.
00:05:11.759 --> 00:05:14.319
I still use Notebook LM almost every single day.
00:05:14.480 --> 00:05:20.079
Sometimes I'll technically spend like hours inside of Notebook LM, just depends on what I'm preparing for.
00:05:20.240 --> 00:05:28.959
Um, but um to make this a little faster, I essentially just created the same version um of this notebook inside of Gemini Notebook.
00:05:29.040 --> 00:05:31.600
I'm gonna probably call it the wrong thing like five times.
00:05:31.680 --> 00:05:34.160
Um, and then I have some prompts ready to go.
00:05:34.240 --> 00:05:41.680
Um, so for our um podcast audience, FYI, our AI at work on Wednesday show, it's always a hands-on demo.
00:05:41.839 --> 00:05:53.279
Sometimes it's a little more visual, sometimes it's not, but you can always watch the video version of this on our website at your everydayai.com and in the podcast show notes, uh, we always leave a link to today's episode.
00:05:53.519 --> 00:05:55.759
All right, so here we go.
00:05:55.920 --> 00:05:58.240
Um, I'm gonna go ahead and enter these prompts.
00:05:58.399 --> 00:06:07.839
We'll probably check on them a little bit as they go, talk a little bit more about these seven new features, um, and then we'll come back at the end and take a look.
00:06:08.000 --> 00:06:10.800
All right, so I'm just gonna get these started so everyone can see.
00:06:10.879 --> 00:06:12.800
Yes, these are going in real time.
00:06:12.959 --> 00:06:16.000
All right, so there's my first one that I sent.
00:06:16.240 --> 00:06:23.199
Uh, there's my second one, there's my third one, and there is my fourth one.
00:06:23.360 --> 00:06:27.360
All right, uh, so a little bit about the content in here.
00:06:27.519 --> 00:06:29.120
It is long form content.
00:06:29.199 --> 00:06:37.600
I think it was like uh probably 60 pages of notes that I had for the show on recursive self-improvement.
00:06:37.680 --> 00:06:42.720
Um, and if you want to go listen to that, like I said, that's yesterday's episode, episode 833.
00:06:42.879 --> 00:06:47.839
All right, where we talked about RSI explained, when AI starts improving itself and what it means.
00:06:48.000 --> 00:06:54.079
So essentially, I have these four different duplicate duplicate notebooks that I just started going, and I'll kind of read the different prompts.
00:06:54.160 --> 00:07:02.800
Um, some of them I'll read in full because it's actually gonna show us um, you know, some of these new features that we're really pushing uh Gemini Notebook on.
00:07:02.959 --> 00:07:04.800
Well, just number one, see if they work.
00:07:04.959 --> 00:07:09.680
Uh, the first one I think will be pretty um relevant to a lot of us.
00:07:09.759 --> 00:07:13.680
And I want you to also think as I'm going through these demos, think of your use cases, right?
00:07:13.759 --> 00:07:18.959
Um, and probably before I do this, I'm gonna give the quick like 60-second overview of Notebook LM.
00:07:19.120 --> 00:07:25.360
Uh, sorry if this is uh, you know, um a rerun for you, uh, right, but I think it's important to just set that.
00:07:25.439 --> 00:07:30.480
So the biggest thing is notebook, or sorry, Gemini, uh, Gemini Notebook just doesn't sound right.
00:07:30.639 --> 00:07:32.399
All right, there's essentially three different panes.
00:07:32.639 --> 00:07:41.360
There's your source pane on the left-hand side, there's your chat pane in the middle, and then on the right hand side is your studio pane where you can create different kinds of um artifacts and outputs.
00:07:41.519 --> 00:07:54.319
So for your sources, um, in my example, I just copied and paste, but there's different ways you can add sources, you can upload files, um, you know, PDFs, images, docs, audios, uh, you can uh put in websites and YouTube videos.
00:07:54.560 --> 00:07:59.839
You can um auto sync now your Google Drive files, which is really helpful.
00:07:59.920 --> 00:08:06.560
Um, and then like I said, copy uh copy and paste text, or there's kind of like a quick uh fast research and deep research at the top.
00:08:06.639 --> 00:08:08.480
So, number one, you're gonna get your sources.
00:08:08.560 --> 00:08:16.720
And this is where the big difference in Notebook LM comes because when you're chatting in the middle, everything that you ask is going to be grounded in that information.
00:08:16.879 --> 00:08:18.240
And that's the biggest difference.
00:08:18.319 --> 00:08:32.080
And I think what makes Gemini Notebook really special and much different because, you know, for all other large language models, right, they're gonna use a combination of number one, the files you upload, just like Gemini Notebook, but number two, training data, right?
00:08:32.240 --> 00:08:34.000
Which could be a good or bad thing.
00:08:34.080 --> 00:08:34.559
Um, right.
00:08:34.639 --> 00:08:37.279
Sometimes training data isn't always the best, sometimes it is.
00:08:37.440 --> 00:08:41.919
And then last but not least, you know, large language models now by default all connect to the internet.
00:08:42.080 --> 00:08:52.799
So the big difference uh with Gemini Notebook is it's gonna ground those chat responses in the middle in just your sources, which cuts down on your hallucinations by an order of magnitude through the roof.
00:08:52.960 --> 00:08:54.480
All right, so uh that's it.
00:08:54.559 --> 00:08:58.240
And then you can chat in the middle uh with your sources, they're gonna output there.
00:08:58.320 --> 00:09:11.440
But then on the right hand side, you can create a bunch of different multimedia um assets in the studio, uh, audio overview, slide decks, uh, video overview, mind maps, reports, flashcards, quizzes, infographics, and data tables.
00:09:11.600 --> 00:09:18.000
And we've actually done dedicated shows on some of these before, uh, on the videos, uh, I think the infographics, right?
00:09:18.080 --> 00:09:23.360
But essentially, you have some of the best of Google's AI products baked into the studio right here.
00:09:23.440 --> 00:09:26.960
Um, you know, as an example, the visuals are created by Nano Banana.
00:09:27.039 --> 00:09:34.559
Um, you know, the audio overview is created by Google's audio uh model, some of the video, you know, uh you can also customize these things.
00:09:34.799 --> 00:09:41.360
So that there's like a cinematic uh video as an example, and that uses elements of Google's VO model, uh video model.
00:09:41.519 --> 00:09:49.600
So essentially, you know, without even really knowing it, you can just click one button and use uh a lot of different um of Google's, you know, different models.
00:09:49.759 --> 00:09:49.919
All right.
00:09:50.080 --> 00:09:59.679
So let me just quickly read some of these prompts here just so we can understand exactly what's going on working with this uh this long chunky data on uh recursive self-improvement.
00:09:59.759 --> 00:10:02.960
And one thing I will let you know, it's already been three minutes, right?
00:10:03.039 --> 00:10:05.679
And my first prompt is still working, right?
00:10:05.759 --> 00:10:07.440
And it's not even nearly done.
00:10:07.600 --> 00:10:16.320
Um, and that's important because you know, one of the big steps here is this is now powered by Gemini 3.5 um and the anti-gravity harness, right?
00:10:16.399 --> 00:10:27.759
So what that means is before um, you know, notebook LM, and I can say that because before it was Notebook LM, um it wasn't truly agentic in nature, at least not in the way that it is now, right?
00:10:27.840 --> 00:10:31.679
Uh, some of those new updates that I talked about was it has a secure cloud computer.
00:10:31.840 --> 00:10:37.759
So what that means is that each notebook can write and execute code for research, calculations, and data analysis.
00:10:37.840 --> 00:10:40.639
And then this first prompt, that's exactly what I'm having it do.
00:10:40.799 --> 00:10:46.720
Aside from having it create custom um outputs, it's also having to under the hood uh do a lot of that.
00:10:46.799 --> 00:10:48.799
So here's the first prompt what I said.
00:10:49.200 --> 00:11:01.840
I said, using this notebook and its sources, create a personalized AI budget decision package for a non-technical CFO and CIO at a 5,000 person enterprise whose agent usage is growing rapidly.
00:11:02.000 --> 00:11:07.840
Do not ask me questions, make reasonable assumptions, state them clearly, and complete the entire task in one run.
00:11:08.000 --> 00:11:10.320
I'm saying first analyze three strategies.
00:11:10.480 --> 00:11:12.720
Number one, use frontier models for every workflow.
00:11:12.799 --> 00:11:15.279
Number two, use the cheapest available model for every workflow.
00:11:15.440 --> 00:11:22.399
Number three, use a tiered architecture in which a frontier model plans, a cheaper model executes, and an independent systems evaluates.
00:11:22.480 --> 00:11:33.759
Uh, then I'm saying compare them using total costs, successful task rate, retries, human retry, uh, human review time, latency, vendor risk, and adaptability to future price reductions.
00:11:34.000 --> 00:11:41.279
Use the Luna and Terra pricing evidence from the notebook, run the necessary calculations and select one strategy.
00:11:41.440 --> 00:11:48.159
Explain why the alternatives lose, identify the strongest counter argument and state what future evidence.
00:11:48.480 --> 00:11:50.559
Oh, just uh put me to the bottom.
00:11:50.639 --> 00:11:54.879
Um, and state what future where did I lose?
00:11:55.039 --> 00:11:56.480
Uh there we go.
00:11:56.559 --> 00:11:58.799
What future evidence would change the recommendation?
00:11:59.039 --> 00:12:02.399
Then generate these five outputs concurrently.
00:12:02.559 --> 00:12:09.759
Number one, a customized audio overview presented as a CFO CIO discussion that reaches a clear decision.
00:12:09.919 --> 00:12:17.200
Number two, a mind map connecting price changes, agent usage, model routing, evaluation, human review, and total costs.
00:12:17.360 --> 00:12:25.120
Number three, an editable Excel workbook with assumptions, formulas, three scenarios, sensitivity analysis, and a recommendation dashboard.
00:12:25.279 --> 00:12:36.960
Number four, a nano banana powered 16 by nine infographic titled Cheaper AI, Bigger AI budget question mark that explains the selected uh strategy visually.
00:12:37.120 --> 00:12:47.279
And number five, a two-page PDF executive decision brief uh containing the recommendation, supporting evidence, rejected alternatives, risks, and next three actions.
00:12:47.440 --> 00:12:47.840
All right.
00:12:47.919 --> 00:12:59.600
So you can see, yeah, kind of a beefy prompt there, but you can probably already start to imagine how something like this new Gemini Notebook would be valuable for your use case, right?
00:12:59.759 --> 00:13:08.559
Um, one of the important things to keep in mind about uh Gemini Notebook is the output studio now is you can just use it with prompts, right?
00:13:08.639 --> 00:13:21.519
So now you can see how a single prompt can do a lot of heavy lifting versus having to go into right um all of these studio assets individually and trying to build them, which is great because what you can actually do, right?
00:13:21.679 --> 00:13:32.320
Starting to uh you know bridge in other terminal uh terminology and um strategies from other uh you know LLMs uh is well, you can create skills, right?
00:13:32.399 --> 00:13:36.000
I actually have a codex skill that does something like this for me, right?
00:13:36.080 --> 00:13:45.279
It will go and, you know, according to my uh, you know, um needs and personalization and customization, it will just literally control my browser.
00:13:45.360 --> 00:13:48.000
It will go in and do all of these kind of things for me.
00:13:48.159 --> 00:13:59.440
And you can see how it's gonna have a much higher success rate and just be way more token um efficient for codecs, you know, using browser use or computer use to go in and do this.
00:13:59.600 --> 00:14:08.000
When I can just throw that big prompt in the chat, knowing now that Gemini Notebook um runs agentically, it has its own cloud computer.
00:14:08.080 --> 00:14:17.200
Whereas before, to get this kind of personalized output, multiple um, you know, artifacts, it would have taken a lot of work, right?
00:14:17.279 --> 00:14:19.279
If I was to just hand this over to an agent.
00:14:19.440 --> 00:14:24.559
So now that you uh really just have a much stronger harness, right?
00:14:24.639 --> 00:14:30.240
It this is Gemini 3.5 using the anti-gravity um engine underneath, right?
00:14:30.320 --> 00:14:33.200
Now the capabilities are just much, much higher.
00:14:33.360 --> 00:14:33.679
All right.
00:14:33.759 --> 00:14:42.240
So uh technically it's still working, um, it's still creating one of the audio overviews, but the actual chat completion now is done.
00:14:42.399 --> 00:14:49.600
So I'm not gonna go through this uh too much in length, but I do want to describe, especially for our podcast audience, kind of what went on here.
00:14:49.759 --> 00:14:54.240
So uh I should have timed this, but it looks like it took about uh five or six minutes.
00:14:54.399 --> 00:15:04.639
And uh, you know, one thing I actually like about Gemini Notebook that's better than other aspects of Gemini, it actually has a decent chain of thought, which is nice, right?
00:15:04.799 --> 00:15:14.240
Because that's one thing that I think uh Google and Gemini is really lacking in uh compared to OpenAI and Anthropic, is being able to see what's actually going on under the hood.
00:15:14.320 --> 00:15:16.000
So little cheat code for you there.
00:15:16.159 --> 00:15:23.440
Uh Gemini Notebook is actually probably better than the default Gemini in terms of knowing and learning what's going on under the hood.
00:15:23.600 --> 00:15:30.559
So, you know, I'm not gonna read all of the chain of thought because it's a lot, but you can go try this yourself, you know, click thoughts and see what's going on there.
00:15:30.639 --> 00:15:41.120
Uh, but it's developing budgeting solutions, it's evaluating the AI strategies, analyzing the model costs, um, you know, it's retrieving some instructions, analyzing strategies, analyzing pricing, right?
00:15:41.200 --> 00:15:43.600
So I can go through and see exactly what it's doing.
00:15:43.679 --> 00:15:45.519
Um, and I'm looking at some of these things.
00:15:45.600 --> 00:15:48.960
I can see that it's creating, uh, you know, creating different files.
00:15:49.039 --> 00:15:50.480
So it's running some code.
00:15:50.559 --> 00:15:54.240
Um, so you know, it's executing code in there in its dedicated sandbox.
00:15:54.399 --> 00:15:57.039
So this was not possible a week ago.
00:15:57.120 --> 00:16:00.799
So the output here, I'm not gonna read it because it's fairly long, but I'll read the beginning.
00:16:00.960 --> 00:16:03.840
So it says executive AI budget decision package.
00:16:04.000 --> 00:16:11.919
This package is designed for the CFO and CIO of our 5,000-person enterprise to address rapidly growing AI agent usage.
00:16:12.080 --> 00:16:20.879
Below, we analyzed three potential model procurement strategies and delivered five unified synchronized outputs to drive your strategic budgeting decision.
00:16:21.120 --> 00:16:22.559
All right, so we'll see here.
00:16:22.720 --> 00:16:25.440
I'm not gonna look at this, but it is exhaustive, right?
00:16:25.519 --> 00:16:28.639
And the good thing is, um, all of this is sourced as well.
00:16:28.799 --> 00:16:36.159
So, you know, it's saying, you know, workflow tokens, you know, for planning, you know, using 20,000 input, two uh 2,000 output.
00:16:36.240 --> 00:16:39.919
I can hover over that and see exactly where it's pulling that from in my notes.
00:16:40.080 --> 00:16:45.840
Uh so you know, when I'm looking over the different strategies here, it did a pretty good job putting together three different strategies.
00:16:46.000 --> 00:16:55.919
One's uh strategy one was using frontier models only, uh, strategy two was using cheap models only, and then strategy three was a tiered architecture using a combination of both.
00:16:56.000 --> 00:17:02.399
And then it went through and, you know, uh went through and did these kind of combined monthly system costs, right?
00:17:02.480 --> 00:17:18.640
And you know, what's kind of cool in this scenario is the tiered architecture, um, right, which is kind of using the stronger model as an orchestrator, at least according to its calculations, um, that was actually gonna be better, right, than just doing the cheap only.
00:17:18.720 --> 00:17:28.480
And the cheap only was actually gonna be more expensive, uh, presumably because it was gonna go through a lot more tokens and more monthly human review rework costs, right?
00:17:28.559 --> 00:17:31.279
So such a cool uh tool to put together.
00:17:31.599 --> 00:17:37.680
All right, but then let's look over at our uh uh right-hand side in our panel and see what was actually created.
00:17:37.839 --> 00:17:42.559
So we did ask specifically for five different uh deliverables.
00:17:42.720 --> 00:17:46.480
We asked for a customized five-minute audio overview, which is almost done.
00:17:46.640 --> 00:17:50.640
Uh, number two, we asked for a mind map connecting pricing changes.
00:17:50.799 --> 00:17:53.359
We asked for an editable Excel workbook.
00:17:53.599 --> 00:17:58.799
We asked for a nano banana powered 16 by nine infographic and a two-page PDF.
00:17:58.880 --> 00:18:03.200
So, so many of these new, so many of these things, again, were not available last week.
00:18:03.359 --> 00:18:13.440
You know, you really had to stay in this kind of confined output of what uh Notebook LM offered, but now with Gemini Notebook, right, it's really flexible.
00:18:13.599 --> 00:18:20.000
And that is actually really helpful, both from what you can do in the middle pane, because you don't have to be um as rigid, right?
00:18:20.160 --> 00:18:35.039
But in my use case, right, if you really want to have agents working for you, I think this makes it so much easier for a you know clawed desktop or a codex desktop, you know, an agent uh living, you know, on your desktop that can control your browser.
00:18:35.200 --> 00:18:38.880
This is going to allow it to iterate and create much more valuable outputs for you, right?
00:18:38.960 --> 00:18:42.880
Um, I always want to encourage listeners of our show, even if you're a beginner, right?
00:18:43.039 --> 00:18:45.279
We have to get out of the prompting phase, right?
00:18:45.440 --> 00:18:53.039
We have to be able to give these desktop agents as much context and as much information about what we need, um, then let them go do the work.
00:18:53.119 --> 00:18:58.799
And I'll usually have another uh, you know, AI agent audit their work and make sure it, you know, cracks any mistakes that it sees.
00:18:58.880 --> 00:19:06.640
But in this scenario, right, if I were to do this manually, even inside Nobook LM, it would have taken me three to five times longer.
00:19:06.720 --> 00:19:08.720
Um, but that's just because of the new harness.
00:19:08.880 --> 00:19:11.920
So let's just quickly take a look at all of our other sources.
00:19:12.079 --> 00:19:14.720
So here we have our AI budget model.
00:19:14.880 --> 00:19:15.519
All right.
00:19:15.920 --> 00:19:18.400
So downloaded uh an Excel file.
00:19:18.480 --> 00:19:20.640
I'm just gonna screenshot this here.
00:19:20.799 --> 00:19:25.440
Um, let's do that versus having to unshare and reshare my screen.
00:19:25.599 --> 00:19:26.880
Uh let's see.
00:19:28.799 --> 00:19:30.720
It's uploading, it's uploading.
00:19:30.880 --> 00:19:32.400
Oh gosh, let's see.
00:19:32.880 --> 00:19:34.240
All right, that one.
00:19:34.319 --> 00:19:36.160
Let me uh let me try that again.
00:19:36.240 --> 00:19:39.839
Uh the actual Excel sheet looks looks great.
00:19:40.240 --> 00:19:42.079
Um, we have our budget model.
00:19:42.240 --> 00:19:44.480
Uh let's let's see if we can get that again here.
00:19:44.640 --> 00:19:46.880
Uh let's copy this, shall we?
00:19:47.119 --> 00:19:49.759
Um, here we go.
00:19:51.039 --> 00:19:51.680
There we go.
00:19:51.759 --> 00:19:52.559
We got it here.
00:19:52.720 --> 00:19:57.200
Uh so we got a working uh financial model, right?
00:19:57.279 --> 00:20:02.079
I just quickly checked the spreadsheet, um, gave me an XLS, opened it in Excel.
00:20:02.240 --> 00:20:03.839
Everything's working, uh, right?
00:20:03.920 --> 00:20:05.519
It gave me exactly what I wanted.
00:20:05.599 --> 00:20:07.920
It gave me my different uh three strategies.
00:20:08.079 --> 00:20:09.759
Um, it mapped out the cost.
00:20:09.920 --> 00:20:11.599
I can go in and change the formulas.
00:20:11.680 --> 00:20:14.880
There's a little graphics, um, a little dashboard.
00:20:14.960 --> 00:20:16.880
It created exactly what I wanted.
00:20:16.960 --> 00:20:17.359
It works.
00:20:17.440 --> 00:20:19.200
Is it the most beautiful spreadsheet ever?
00:20:19.279 --> 00:20:19.519
No.
00:20:19.680 --> 00:20:21.519
Is it something better than I could have done?
00:20:21.680 --> 00:20:22.640
Absolutely, right?
00:20:22.799 --> 00:20:27.359
I love uh working in spreadsheets, but I'm not necessarily the best at creating formulas, right?
00:20:27.440 --> 00:20:30.960
So again, uh just think of all of the unlocks that something like this brings.
00:20:31.119 --> 00:20:33.920
All right, so that was our AI budget model, did a good job.
00:20:34.000 --> 00:20:39.440
Uh, then we asked for a two-page uh PDF that gives an executive decision brief.
00:20:39.519 --> 00:20:41.359
All right, so now uh here we go.
00:20:41.519 --> 00:20:44.799
Live stream audience can see, but we have a two-page executive brief.
00:20:44.960 --> 00:20:52.160
This is the AI portfolio optimization, uh, the implement uh implementing a tiered model architecture for the 5,000-person enterprise company.
00:20:52.319 --> 00:20:52.880
There we go.
00:20:53.039 --> 00:21:04.640
This is, you know, just kind of a PDF version of what we uh created inside the middle pane chat, um, as well as why, you know, what the uh the best strategy was and the counterarguments for those.
00:21:04.720 --> 00:21:05.440
So perfect.
00:21:05.599 --> 00:21:06.559
It created that.
00:21:06.720 --> 00:21:08.720
Uh let's look at the other ones.
00:21:09.039 --> 00:21:10.640
We have our mind map.
00:21:10.799 --> 00:21:13.680
Okay, so here we have our uh RSI here.
00:21:13.759 --> 00:21:21.599
Uh, then we have the economic shifts in pricing, the tiered agentic architect uh architecture, governance and control, and monthly system cost scenarios.
00:21:21.680 --> 00:21:26.559
And then I can click those out um and explore a little bit more uh in each of those.
00:21:26.720 --> 00:21:27.920
There we go.
00:21:28.240 --> 00:21:31.200
All right, so uh mind map, great job.
00:21:31.359 --> 00:21:36.960
All right, and then let's look at our uh this should be our infographic created by Nano Banana.
00:21:37.119 --> 00:21:39.279
All right, so looks good, you know.
00:21:39.359 --> 00:21:43.680
I'm I'm kind of uh, you know, quickly editing it and it looks good.
00:21:43.839 --> 00:21:47.839
The only thing I see wrong is there's a notebook LM watermark.
00:21:47.920 --> 00:21:49.519
All right, Google, we got to update that, right?
00:21:49.599 --> 00:21:54.799
If we're trying to get rid of notebook LM, uh, we gotta update the the watermark to Gemini Notebook.
00:21:54.880 --> 00:21:57.599
Uh, but the it this actually looks pretty good.
00:21:57.759 --> 00:21:59.680
Um, the uh 16 by nine Nano.
00:22:00.160 --> 00:22:04.319
Banana uh graphic for the CFO, CIO decision maker.
00:22:04.559 --> 00:22:05.039
Great.
00:22:05.200 --> 00:22:15.279
Um looking at the you know three different strategies, it breaks down the human cost, the model cost, the the pros and the cons, uh great charts and graphics, did a really good job.
00:22:15.440 --> 00:22:17.519
And then last but not least, it did just finish.
00:22:17.759 --> 00:22:19.119
I'm just gonna listen here.
00:22:21.279 --> 00:22:22.240
All right, perfect.
00:22:22.400 --> 00:22:28.640
It created a 21-minute deep dive on self-improving AI slash token prices.
00:22:28.960 --> 00:22:32.960
All right, so um okay, interesting.
00:22:33.119 --> 00:22:36.160
Um, so it even did put in the custom prompt.
00:22:36.240 --> 00:22:37.200
Uh, so that's cool.
00:22:37.279 --> 00:22:40.400
It went and did everything correctly, right?
00:22:40.480 --> 00:22:42.480
So, hey, the demo actually worked.
00:22:42.640 --> 00:22:46.480
All right, I'm just gonna quickly look at some of the other use cases just so we can see what we did.
00:22:46.720 --> 00:22:54.000
Uh in the other notebook, I said create three quizzes of ascending difficulty focusing on a different topic from my sources.
00:22:54.079 --> 00:22:55.119
Uh, this is a good one.
00:22:55.279 --> 00:23:00.400
This was actually um a Gemini notebook uh example prompt, which I thought was a really good idea.
00:23:00.559 --> 00:23:05.119
Um, if you're trying to learn something, you know, they have this quiz module, which is really good, right?
00:23:05.279 --> 00:23:13.359
But uh, you know, sometimes you might really want to be intentional about how you learn things and break it off and start easier so you can go through the paces and relearn things as you go.
00:23:13.519 --> 00:23:16.160
Uh so this one, it created an easy quiz.
00:23:16.319 --> 00:23:16.799
All right.
00:23:16.880 --> 00:23:19.039
Uh, let's should I do one question?
00:23:19.200 --> 00:23:20.480
Watch, watch me get it wrong.
00:23:20.640 --> 00:23:25.119
All right, what is the core definition of recursive self-improvement in the context of AI?
00:23:25.279 --> 00:23:29.599
A, is it a method where humans manually rewrite every line of code to make AI faster?
00:23:29.839 --> 00:23:44.160
B, a process where an AI uses its own capabilities to design and build a more advanced version of itself, C, an AI system that is only capable of performing one specific task like plain chess, or D, a way for AI to search the internet and summarize existing news articles better.
00:23:44.240 --> 00:23:45.440
So I know it's B.
00:23:45.599 --> 00:23:46.640
All right, so that's right.
00:23:46.799 --> 00:23:51.039
So it has three different quizzes, did a great job doing that of ascending difficulty.
00:23:51.200 --> 00:23:52.960
All right, let's look at our next one.
00:23:53.119 --> 00:23:57.359
Uh, so for this one, I said create a five-level RSI maturity ladder.
00:23:57.440 --> 00:24:00.079
And I wanted an infographic and a PDF.
00:24:00.240 --> 00:24:01.599
So it did answer everything.
00:24:01.759 --> 00:24:02.720
This is really good.
00:24:02.799 --> 00:24:04.559
Uh, it answered everything in the chat.
00:24:04.640 --> 00:24:10.799
Uh, my PDF here, it broke down the five uh kind of ladders of RSI, did a great job.
00:24:10.880 --> 00:24:12.640
And then let's look at the infographic.
00:24:12.720 --> 00:24:13.680
Uh, cool, yeah.
00:24:13.759 --> 00:24:20.400
This infographic actually came out a little cleaner than I thought, considering that uh we didn't uh customize anything.
00:24:20.480 --> 00:24:23.599
So yeah, we have the uh level one, AI assists researchers.
00:24:23.680 --> 00:24:26.319
Level two, AI executes a human design method.
00:24:26.480 --> 00:24:28.880
Level three, AI improves AI infrastructure.
00:24:29.039 --> 00:24:31.920
Level four, AI helps train or improve another AI.
00:24:32.000 --> 00:24:35.039
And then level five, AI controls the complete improvement loop.
00:24:35.279 --> 00:24:36.240
All right, let's go.
00:24:36.319 --> 00:24:37.200
Let's look at our next one.
00:24:37.279 --> 00:24:43.440
I said using the old Luna and Terra pricing in this notebook, create an edible Excel calculator for agentic workflows.
00:24:43.680 --> 00:24:45.440
I have my uh Excel sheet there.
00:24:45.519 --> 00:24:47.440
I'm opening it up in another tab.
00:24:47.839 --> 00:24:50.559
Um yeah, looks good.
00:24:50.640 --> 00:24:51.359
I'm checking.
00:24:51.519 --> 00:24:54.480
There's formulas, everything's working live, perfect.
00:24:54.640 --> 00:25:01.680
All right, and then uh one more thing uh before we wrap up here that I wanted to showcase.
00:25:01.920 --> 00:25:02.559
All right.
00:25:04.559 --> 00:25:07.039
The middle is agentic now.
00:25:07.200 --> 00:25:18.160
All right, so even though it's still going to be grounded by default, uh, I'm gonna go into one of my things here and I'm gonna type who won the 2016 World Series.
00:25:18.240 --> 00:25:19.839
All right, now I'm gonna ask that question.
00:25:20.000 --> 00:25:23.599
So by default, it should come back to me and say, hey, I don't know.
00:25:23.759 --> 00:25:26.880
However, I can go out and find that information.
00:25:27.039 --> 00:25:29.680
All right, so why am I bringing this up?
00:25:30.160 --> 00:25:34.960
Because that is the one important grounded change that you need to be aware of.
00:25:35.200 --> 00:25:42.720
All right, so what's interesting here, um, and I didn't know this, it looks like because it had already started pulling things agentically on the web.
00:25:42.799 --> 00:25:51.279
Um, because in some of my other testing, when I did this for the first time, um, you know, so I just opened up another uh chat to show you this.
00:25:51.440 --> 00:25:53.519
I said, who won the 2016 World Series?
00:25:53.599 --> 00:25:57.279
In this example, it says your sources do not contain information about this.
00:25:57.440 --> 00:25:59.519
Do you want me to research this on the web?
00:25:59.599 --> 00:26:00.400
I said yes.
00:26:00.559 --> 00:26:08.079
So it looks like maybe if web research had already been active, uh had already been activated, that it might just answer that by default.
00:26:08.160 --> 00:26:16.000
So that's an important thing to think about because previously, um, notebook LM in the middle chat was always 100% grounded in your information.
00:26:16.160 --> 00:26:18.319
So that's a big difference and a big change.
00:26:18.480 --> 00:26:22.720
So in some instances, it's a big upside, but you have to be aware of that.
00:26:22.799 --> 00:26:37.279
Um, that you know, you have to be able to go back and look at the chain of thought in a lot of these prompts, it went and it was doing a lot of web research in addition to the information that I gave it, because in the prompts, I was kind of pushing it past the limitation of the sources.
00:26:37.440 --> 00:26:41.920
So this does kind of, in my opinion, change uh Gemini Notebook.
00:26:42.000 --> 00:26:49.519
And maybe that's good that we are getting away from the notebook LM in inserting Gemini Notebook because there's pros and cons to that, right?
00:26:49.680 --> 00:27:02.160
Like you saw in this example here, because it looks like because I had already triggered the anti-gravity harness um in that first prompt and then asked it information that was not in my sources, it just well, it just responded.
00:27:02.319 --> 00:27:02.799
All right.
00:27:02.880 --> 00:27:11.440
Uh, but uh it did say it says I had I haven't imported those sources about the 2016 World Series.
00:27:11.920 --> 00:27:12.720
Oh, wait.
00:27:12.960 --> 00:27:13.759
I looked at it.
00:27:13.920 --> 00:27:15.119
Okay, I didn't read.
00:27:15.279 --> 00:27:15.839
It didn't.
00:27:16.079 --> 00:27:16.480
All right.
00:27:16.559 --> 00:27:19.359
Um, it just said that I hadn't looked at those.
00:27:19.759 --> 00:27:21.359
Uh, do you want me to look them up?
00:27:21.440 --> 00:27:23.039
Uh I I got ahead of myself.
00:27:23.200 --> 00:27:24.640
Uh so yeah, I can say yes.
00:27:24.799 --> 00:27:25.119
All right.
00:27:25.200 --> 00:27:35.839
So, all right, scratch what I just said the last 60 seconds that um I thought because I had already activated the anti-gravity uh harness and it was already doing uh tool calls that it was gonna go out and respond.
00:27:35.920 --> 00:27:37.279
I literally just didn't read it.
00:27:37.359 --> 00:27:39.759
I just saw the response and I'm like, oh my gosh, it responded.
00:27:39.920 --> 00:27:42.000
But no, me uh not sleeping enough.
00:27:42.079 --> 00:27:46.240
It literally said, I haven't imported those sources about the 2016 World Series into your notebook yet.
00:27:46.400 --> 00:27:47.759
Would you like me to import them first?
00:27:47.920 --> 00:27:48.799
And then I said yes.
00:27:48.880 --> 00:27:51.920
And now, right now, I can click and see that it's going out to grab those sources.
00:27:52.000 --> 00:27:53.200
All right, I stand corrected.
00:27:53.519 --> 00:28:02.960
So it's still important to know though that you know, once you do give Gemini Notebook access to, it is going to bring in information that are outside of the sources.
00:28:03.039 --> 00:28:13.440
So it does change, I think, again, for the good and the bad, but at least uh it does caution you uh or make you kind of manually approve that's saying, Hey, you don't have this information in your sources.
00:28:13.599 --> 00:28:15.759
Do you want me to go grab that information for you?
00:28:15.920 --> 00:28:18.559
All right, so that is a wrap.
00:28:18.720 --> 00:28:21.680
But to quickly recap, what do you need to know?
00:28:22.000 --> 00:28:26.400
All right, so uh number one, what's the big difference here?
00:28:26.880 --> 00:28:33.759
Gemini Notebook represents a shift away from just having that, you know, notebook LM, all those sources.
00:28:33.920 --> 00:28:41.839
And, you know, essentially it was a non-reasoning model that could use Google's different multi uh multimedia AI.
00:28:41.920 --> 00:28:43.119
And now it's different, right?
00:28:43.200 --> 00:28:49.839
And I think that's the the big reason for the name change is because now with the Gemini Notebook, it is a gentic by default, right?
00:28:49.920 --> 00:29:00.319
Having that cloud computer, being able to write and execute code, being able to, when you give it the okay, search the web, it really changes what Notebook LM was into what Gemini Notebook will be in the future.
00:29:00.720 --> 00:29:04.400
You know, the important grounded change that I talked about right there, you just have to know.
00:29:04.480 --> 00:29:08.240
And I think ultimately that will be a good thing as long as you understand what's going on.
00:29:08.319 --> 00:29:14.480
Um, and then the real reason I think notebook, uh Gemini Notebook going agentic matters more than ever.
00:29:14.559 --> 00:29:20.079
Uh, well, you saw just through my example, it makes it much easier to work with, right?
00:29:20.240 --> 00:29:24.079
Uh before, um, you know, you were kind of restrained, right?
00:29:24.240 --> 00:29:28.400
You almost had to think in your mind um in terms of outputs, right?
00:29:28.559 --> 00:29:32.079
You had to prompt notebook LM according to the outputs.
00:29:32.240 --> 00:29:34.240
Now those outputs are so flexible, right?
00:29:34.319 --> 00:29:37.680
Before you couldn't create PDFs, before you couldn't create CSVs, right?
00:29:37.759 --> 00:29:39.759
You couldn't create markdown files.
00:29:39.839 --> 00:29:42.799
So you were much more limited in, you know, PowerPoints, right?
00:29:42.880 --> 00:29:43.920
You can do all these things now.
00:29:44.079 --> 00:29:52.480
So uh, you know, Gemini Notebook has become uh much more of an agentic uh, you know, coworker versus what it was before.
00:29:52.640 --> 00:30:00.240
You know, there's essentially a handful of uh artifacts or a handful of formats, and you really had to use Notebook LM according to those outputs.
00:30:00.400 --> 00:30:03.759
Now it is extremely flexible and even more powerful than ever.
00:30:04.000 --> 00:30:10.160
And I guess that's why Notebook LM is no more, but Gemini Notebook at least is here to say, and it is really good.
00:30:10.319 --> 00:30:12.160
All right, that's a wrap for today's show.
00:30:12.240 --> 00:30:15.519
Now you know Gemini Notebook, the seven new updates and what they unlock.
00:30:15.680 --> 00:30:19.759
If this was helpful, do me a favor, tell someone about it, share this on LinkedIn.
00:30:20.000 --> 00:30:23.680
Uh, you know, go and subscribe to the podcast on Apple or Spotify.
00:30:23.759 --> 00:30:27.759
Then make sure to go to your everydayai.com, sign up for the free daily newsletter.
00:30:27.839 --> 00:30:30.559
We're going to be recapping the highlights from today's show as well as a lot more.
00:30:30.720 --> 00:30:31.519
Thanks for tuning in.
00:30:31.680 --> 00:30:34.559
Hope to see you back tomorrow and every day for more everyday AI.
00:30:34.640 --> 00:30:35.599
Thanks, y'all.
00:00:02.319 --> 00:00:05.919
How long has it been since we featured Notebook LM on this show?
00:00:06.160 --> 00:00:10.400
Apparently, so long, well, it isn't even called Notebook LM.
00:00:10.560 --> 00:00:11.519
It has a new name.
00:00:11.759 --> 00:00:15.919
Yes, the tool once known as Notebook LM is no more.
00:00:16.160 --> 00:00:20.800
But don't worry, one of my favorite AI tools ever didn't go to the Google graveyard.
00:00:20.960 --> 00:00:27.839
It actually just graduated and got upgraded into the official Gemini lineup and is now called Gemini Notebooks.
00:00:27.920 --> 00:00:30.399
But the name isn't the only thing that's new.
00:00:30.559 --> 00:00:41.119
That's because over the last month or so, the team at Google has rolled out a bunch of new features that legit supercharge Gemini Notebooks and make it a GENTIC by default.
00:00:41.280 --> 00:00:45.439
Which sounds really cool and all, but well, why haven't we been talking about it?
00:00:45.679 --> 00:00:50.240
Because many of those upgrades were only for users on the expensive Ultra Plan.
00:00:50.799 --> 00:00:52.799
But this past week that changed.
00:00:53.039 --> 00:00:59.359
Now the updates that have completely changed the face of Gemini Notebook are available to all paid users.
00:00:59.439 --> 00:01:06.959
And that's why we're going to be showcasing the seven most important updates today as we put AI to work on Wednesdays.
00:01:07.120 --> 00:01:11.040
All right, so on today's show, here's what you're going to learn.
00:01:11.200 --> 00:01:16.959
You're going to know why Notebook LM is no more and why Gemini Notebook might actually make more sense.
00:01:17.120 --> 00:01:21.920
You're going to know the one important grounded change that you really need to be aware of.
00:01:22.159 --> 00:01:28.079
And you're going to understand the real reason Gemini Notebook going agentic matters more than ever.
00:01:28.319 --> 00:01:29.359
All right, let's get into it.
00:01:29.439 --> 00:01:30.959
Welcome to Everyday AI.
00:01:31.040 --> 00:01:33.359
My name's Jordan Wilson, and this thing's for you.
00:01:33.439 --> 00:01:42.319
It's your daily, unedited, unscripted live stream podcast and free daily newsletter, helping business leaders like you and me keep up with the nonstop avalanche of AI updates.
00:01:42.480 --> 00:01:47.359
I tell you what matters, what doesn't, how to use it, and you take that information to grow your company and career.
00:01:47.519 --> 00:01:58.000
So it starts here, but please make sure if you haven't already, please subscribe to the podcast on Spotify or Apple and make sure you go to our website at your everydayai.com and sign up for the free daily newsletter.
00:01:58.079 --> 00:02:05.280
We're going to be recapping the highlights from today's show in case you miss anything, as well as all of the other AI news and developments that you need to know.
00:02:05.439 --> 00:02:05.840
All right.
00:02:05.920 --> 00:02:08.080
So let's talk at Gemini Notebooks.
00:02:08.240 --> 00:02:10.159
Yeah, Notebook LM is no more.
00:02:10.319 --> 00:02:10.960
So sad.
00:02:11.199 --> 00:02:11.919
Uh right.
00:02:12.000 --> 00:02:13.840
But it's not gone.
00:02:14.000 --> 00:02:20.479
It just got a facelift, uh, both on the exterior and a lot going on under the hood as well.
00:02:20.639 --> 00:02:28.800
And this is one of those things I think over the last couple of years, a lot of people were kind of nervous because Google has done this before, where they've come out with great products, right?
00:02:28.879 --> 00:02:33.840
And this isn't just in the AI uh age, in the pre-AI uh phase, maybe even more.
00:02:34.000 --> 00:02:41.120
They've come out with very popular products that a lot of people love, almost built like a cult like following, and then they end up killing those products.
00:02:41.199 --> 00:02:49.039
So a lot of people were very nervous over the last couple of years as Notebook LM almost seemed like this side quest that was legit amazing, right?
00:02:49.120 --> 00:02:49.759
No joke.
00:02:49.840 --> 00:02:54.800
It's I think we named it like our AI tool of the year in 2024 and 2025.
00:02:54.960 --> 00:02:57.759
Um, so a lot of people were worried, like, hey, don't take this away.
00:02:57.919 --> 00:03:16.080
It's very uh it's very unique because there's no, you know, the other competitors, um, you know, no one from Anthropic, OpenAI, Microsoft, uh, you know, X, Groc, uh, Meta, like no one else has something like Notebook LM that really just grounds its outputs in your data only.
00:03:16.319 --> 00:03:18.639
So luckily, they didn't get rid of it.
00:03:18.719 --> 00:03:24.400
And it seems like, if anything, uh Gemini Notebook is gonna be along for the long run.
00:03:24.879 --> 00:03:27.759
So uh let's quickly go over, not gonna make you wait.
00:03:27.919 --> 00:03:30.080
Here's the seven new updates to Gemini Notebook.
00:03:30.159 --> 00:03:31.919
I'm gonna be going over them pretty quickly now.
00:03:32.080 --> 00:03:35.520
Uh, and we're gonna be doing some live demos so you can see these seven updates in action.
00:03:35.599 --> 00:03:38.879
And then when we're done with the live demos, I'll come back and talk about these things a little bit more.
00:03:39.120 --> 00:03:41.360
So, number one, new update, already talked about it.
00:03:41.520 --> 00:03:43.520
Gemini Notebook is the new name.
00:03:43.680 --> 00:03:44.319
That's number one.
00:03:44.479 --> 00:03:46.639
Number two, there's something called collections.
00:03:46.719 --> 00:03:49.759
That's a way for you to organize notebooks for easier sorting.
00:03:49.919 --> 00:03:52.560
Number three, and this one's actually really cool.
00:03:52.639 --> 00:03:56.159
Uh, it's now there's automatic Google Drive syncing.
00:03:56.319 --> 00:04:01.439
All right, number four, uh, bringing the Gemini and search access.
00:04:01.599 --> 00:04:10.479
So your um personal notebooks can now be accessed inside of Gemini and soon inside of uh Google searches AI mode, uh, which will be really handy.
00:04:10.560 --> 00:04:12.719
Uh number five, new output formats.
00:04:12.879 --> 00:04:17.120
So yeah, no longer uh constrained to only what Notebook LM offered.
00:04:17.199 --> 00:04:18.319
Now Gemini Notebook.
00:04:18.399 --> 00:04:23.439
You can create PDFs, PNGs, documents, spreadsheets, PowerPoints, markdown files, charts, and images.
00:04:23.600 --> 00:04:29.680
Uh number six, technically each notebook has its own secure cloud computer, uh, which is really cool.
00:04:29.920 --> 00:04:32.560
Number seven, just more agentic intelligence.
00:04:32.639 --> 00:04:43.279
And that's probably the biggest um upgrade that we're gonna see in Gemini Notebook now, powered by Gemini 3.5, uh reasoning and Google's anti-gravity agentic search.
00:04:43.439 --> 00:04:45.839
All right, so let's do it.
00:04:45.920 --> 00:04:49.279
Uh, let's take a look live, shall we?
00:04:49.439 --> 00:04:50.480
What could go wrong?
00:04:50.560 --> 00:04:52.560
Uh, I'm gonna share my screen, live stream audience.
00:04:52.639 --> 00:04:54.959
Do me a favor, let me know if you can see.
00:04:55.360 --> 00:05:01.759
All right, so what I did here is I took all of my notes from yesterday's show, uh, had a lot of them.
00:05:01.920 --> 00:05:06.319
So uh yesterday's show was kind of on RSI recursive self-improvement.
00:05:06.399 --> 00:05:08.000
Um, I have all of my notes together.
00:05:08.079 --> 00:05:11.600
And this is something I routinely do inside of Notebook LM.
00:05:11.759 --> 00:05:14.319
I still use Notebook LM almost every single day.
00:05:14.480 --> 00:05:20.079
Sometimes I'll technically spend like hours inside of Notebook LM, just depends on what I'm preparing for.
00:05:20.240 --> 00:05:28.959
Um, but um to make this a little faster, I essentially just created the same version um of this notebook inside of Gemini Notebook.
00:05:29.040 --> 00:05:31.600
I'm gonna probably call it the wrong thing like five times.
00:05:31.680 --> 00:05:34.160
Um, and then I have some prompts ready to go.
00:05:34.240 --> 00:05:41.680
Um, so for our um podcast audience, FYI, our AI at work on Wednesday show, it's always a hands-on demo.
00:05:41.839 --> 00:05:53.279
Sometimes it's a little more visual, sometimes it's not, but you can always watch the video version of this on our website at your everydayai.com and in the podcast show notes, uh, we always leave a link to today's episode.
00:05:53.519 --> 00:05:55.759
All right, so here we go.
00:05:55.920 --> 00:05:58.240
Um, I'm gonna go ahead and enter these prompts.
00:05:58.399 --> 00:06:07.839
We'll probably check on them a little bit as they go, talk a little bit more about these seven new features, um, and then we'll come back at the end and take a look.
00:06:08.000 --> 00:06:10.800
All right, so I'm just gonna get these started so everyone can see.
00:06:10.879 --> 00:06:12.800
Yes, these are going in real time.
00:06:12.959 --> 00:06:16.000
All right, so there's my first one that I sent.
00:06:16.240 --> 00:06:23.199
Uh, there's my second one, there's my third one, and there is my fourth one.
00:06:23.360 --> 00:06:27.360
All right, uh, so a little bit about the content in here.
00:06:27.519 --> 00:06:29.120
It is long form content.
00:06:29.199 --> 00:06:37.600
I think it was like uh probably 60 pages of notes that I had for the show on recursive self-improvement.
00:06:37.680 --> 00:06:42.720
Um, and if you want to go listen to that, like I said, that's yesterday's episode, episode 833.
00:06:42.879 --> 00:06:47.839
All right, where we talked about RSI explained, when AI starts improving itself and what it means.
00:06:48.000 --> 00:06:54.079
So essentially, I have these four different duplicate duplicate notebooks that I just started going, and I'll kind of read the different prompts.
00:06:54.160 --> 00:07:02.800
Um, some of them I'll read in full because it's actually gonna show us um, you know, some of these new features that we're really pushing uh Gemini Notebook on.
00:07:02.959 --> 00:07:04.800
Well, just number one, see if they work.
00:07:04.959 --> 00:07:09.680
Uh, the first one I think will be pretty um relevant to a lot of us.
00:07:09.759 --> 00:07:13.680
And I want you to also think as I'm going through these demos, think of your use cases, right?
00:07:13.759 --> 00:07:18.959
Um, and probably before I do this, I'm gonna give the quick like 60-second overview of Notebook LM.
00:07:19.120 --> 00:07:25.360
Uh, sorry if this is uh, you know, um a rerun for you, uh, right, but I think it's important to just set that.
00:07:25.439 --> 00:07:30.480
So the biggest thing is notebook, or sorry, Gemini, uh, Gemini Notebook just doesn't sound right.
00:07:30.639 --> 00:07:32.399
All right, there's essentially three different panes.
00:07:32.639 --> 00:07:41.360
There's your source pane on the left-hand side, there's your chat pane in the middle, and then on the right hand side is your studio pane where you can create different kinds of um artifacts and outputs.
00:07:41.519 --> 00:07:54.319
So for your sources, um, in my example, I just copied and paste, but there's different ways you can add sources, you can upload files, um, you know, PDFs, images, docs, audios, uh, you can uh put in websites and YouTube videos.
00:07:54.560 --> 00:07:59.839
You can um auto sync now your Google Drive files, which is really helpful.
00:07:59.920 --> 00:08:06.560
Um, and then like I said, copy uh copy and paste text, or there's kind of like a quick uh fast research and deep research at the top.
00:08:06.639 --> 00:08:08.480
So, number one, you're gonna get your sources.
00:08:08.560 --> 00:08:16.720
And this is where the big difference in Notebook LM comes because when you're chatting in the middle, everything that you ask is going to be grounded in that information.
00:08:16.879 --> 00:08:18.240
And that's the biggest difference.
00:08:18.319 --> 00:08:32.080
And I think what makes Gemini Notebook really special and much different because, you know, for all other large language models, right, they're gonna use a combination of number one, the files you upload, just like Gemini Notebook, but number two, training data, right?
00:08:32.240 --> 00:08:34.000
Which could be a good or bad thing.
00:08:34.080 --> 00:08:34.559
Um, right.
00:08:34.639 --> 00:08:37.279
Sometimes training data isn't always the best, sometimes it is.
00:08:37.440 --> 00:08:41.919
And then last but not least, you know, large language models now by default all connect to the internet.
00:08:42.080 --> 00:08:52.799
So the big difference uh with Gemini Notebook is it's gonna ground those chat responses in the middle in just your sources, which cuts down on your hallucinations by an order of magnitude through the roof.
00:08:52.960 --> 00:08:54.480
All right, so uh that's it.
00:08:54.559 --> 00:08:58.240
And then you can chat in the middle uh with your sources, they're gonna output there.
00:08:58.320 --> 00:09:11.440
But then on the right hand side, you can create a bunch of different multimedia um assets in the studio, uh, audio overview, slide decks, uh, video overview, mind maps, reports, flashcards, quizzes, infographics, and data tables.
00:09:11.600 --> 00:09:18.000
And we've actually done dedicated shows on some of these before, uh, on the videos, uh, I think the infographics, right?
00:09:18.080 --> 00:09:23.360
But essentially, you have some of the best of Google's AI products baked into the studio right here.
00:09:23.440 --> 00:09:26.960
Um, you know, as an example, the visuals are created by Nano Banana.
00:09:27.039 --> 00:09:34.559
Um, you know, the audio overview is created by Google's audio uh model, some of the video, you know, uh you can also customize these things.
00:09:34.799 --> 00:09:41.360
So that there's like a cinematic uh video as an example, and that uses elements of Google's VO model, uh video model.
00:09:41.519 --> 00:09:49.600
So essentially, you know, without even really knowing it, you can just click one button and use uh a lot of different um of Google's, you know, different models.
00:09:49.759 --> 00:09:49.919
All right.
00:09:50.080 --> 00:09:59.679
So let me just quickly read some of these prompts here just so we can understand exactly what's going on working with this uh this long chunky data on uh recursive self-improvement.
00:09:59.759 --> 00:10:02.960
And one thing I will let you know, it's already been three minutes, right?
00:10:03.039 --> 00:10:05.679
And my first prompt is still working, right?
00:10:05.759 --> 00:10:07.440
And it's not even nearly done.
00:10:07.600 --> 00:10:16.320
Um, and that's important because you know, one of the big steps here is this is now powered by Gemini 3.5 um and the anti-gravity harness, right?
00:10:16.399 --> 00:10:27.759
So what that means is before um, you know, notebook LM, and I can say that because before it was Notebook LM, um it wasn't truly agentic in nature, at least not in the way that it is now, right?
00:10:27.840 --> 00:10:31.679
Uh, some of those new updates that I talked about was it has a secure cloud computer.
00:10:31.840 --> 00:10:37.759
So what that means is that each notebook can write and execute code for research, calculations, and data analysis.
00:10:37.840 --> 00:10:40.639
And then this first prompt, that's exactly what I'm having it do.
00:10:40.799 --> 00:10:46.720
Aside from having it create custom um outputs, it's also having to under the hood uh do a lot of that.
00:10:46.799 --> 00:10:48.799
So here's the first prompt what I said.
00:10:49.200 --> 00:11:01.840
I said, using this notebook and its sources, create a personalized AI budget decision package for a non-technical CFO and CIO at a 5,000 person enterprise whose agent usage is growing rapidly.
00:11:02.000 --> 00:11:07.840
Do not ask me questions, make reasonable assumptions, state them clearly, and complete the entire task in one run.
00:11:08.000 --> 00:11:10.320
I'm saying first analyze three strategies.
00:11:10.480 --> 00:11:12.720
Number one, use frontier models for every workflow.
00:11:12.799 --> 00:11:15.279
Number two, use the cheapest available model for every workflow.
00:11:15.440 --> 00:11:22.399
Number three, use a tiered architecture in which a frontier model plans, a cheaper model executes, and an independent systems evaluates.
00:11:22.480 --> 00:11:33.759
Uh, then I'm saying compare them using total costs, successful task rate, retries, human retry, uh, human review time, latency, vendor risk, and adaptability to future price reductions.
00:11:34.000 --> 00:11:41.279
Use the Luna and Terra pricing evidence from the notebook, run the necessary calculations and select one strategy.
00:11:41.440 --> 00:11:48.159
Explain why the alternatives lose, identify the strongest counter argument and state what future evidence.
00:11:48.480 --> 00:11:50.559
Oh, just uh put me to the bottom.
00:11:50.639 --> 00:11:54.879
Um, and state what future where did I lose?
00:11:55.039 --> 00:11:56.480
Uh there we go.
00:11:56.559 --> 00:11:58.799
What future evidence would change the recommendation?
00:11:59.039 --> 00:12:02.399
Then generate these five outputs concurrently.
00:12:02.559 --> 00:12:09.759
Number one, a customized audio overview presented as a CFO CIO discussion that reaches a clear decision.
00:12:09.919 --> 00:12:17.200
Number two, a mind map connecting price changes, agent usage, model routing, evaluation, human review, and total costs.
00:12:17.360 --> 00:12:25.120
Number three, an editable Excel workbook with assumptions, formulas, three scenarios, sensitivity analysis, and a recommendation dashboard.
00:12:25.279 --> 00:12:36.960
Number four, a nano banana powered 16 by nine infographic titled Cheaper AI, Bigger AI budget question mark that explains the selected uh strategy visually.
00:12:37.120 --> 00:12:47.279
And number five, a two-page PDF executive decision brief uh containing the recommendation, supporting evidence, rejected alternatives, risks, and next three actions.
00:12:47.440 --> 00:12:47.840
All right.
00:12:47.919 --> 00:12:59.600
So you can see, yeah, kind of a beefy prompt there, but you can probably already start to imagine how something like this new Gemini Notebook would be valuable for your use case, right?
00:12:59.759 --> 00:13:08.559
Um, one of the important things to keep in mind about uh Gemini Notebook is the output studio now is you can just use it with prompts, right?
00:13:08.639 --> 00:13:21.519
So now you can see how a single prompt can do a lot of heavy lifting versus having to go into right um all of these studio assets individually and trying to build them, which is great because what you can actually do, right?
00:13:21.679 --> 00:13:32.320
Starting to uh you know bridge in other terminal uh terminology and um strategies from other uh you know LLMs uh is well, you can create skills, right?
00:13:32.399 --> 00:13:36.000
I actually have a codex skill that does something like this for me, right?
00:13:36.080 --> 00:13:45.279
It will go and, you know, according to my uh, you know, um needs and personalization and customization, it will just literally control my browser.
00:13:45.360 --> 00:13:48.000
It will go in and do all of these kind of things for me.
00:13:48.159 --> 00:13:59.440
And you can see how it's gonna have a much higher success rate and just be way more token um efficient for codecs, you know, using browser use or computer use to go in and do this.
00:13:59.600 --> 00:14:08.000
When I can just throw that big prompt in the chat, knowing now that Gemini Notebook um runs agentically, it has its own cloud computer.
00:14:08.080 --> 00:14:17.200
Whereas before, to get this kind of personalized output, multiple um, you know, artifacts, it would have taken a lot of work, right?
00:14:17.279 --> 00:14:19.279
If I was to just hand this over to an agent.
00:14:19.440 --> 00:14:24.559
So now that you uh really just have a much stronger harness, right?
00:14:24.639 --> 00:14:30.240
It this is Gemini 3.5 using the anti-gravity um engine underneath, right?
00:14:30.320 --> 00:14:33.200
Now the capabilities are just much, much higher.
00:14:33.360 --> 00:14:33.679
All right.
00:14:33.759 --> 00:14:42.240
So uh technically it's still working, um, it's still creating one of the audio overviews, but the actual chat completion now is done.
00:14:42.399 --> 00:14:49.600
So I'm not gonna go through this uh too much in length, but I do want to describe, especially for our podcast audience, kind of what went on here.
00:14:49.759 --> 00:14:54.240
So uh I should have timed this, but it looks like it took about uh five or six minutes.
00:14:54.399 --> 00:15:04.639
And uh, you know, one thing I actually like about Gemini Notebook that's better than other aspects of Gemini, it actually has a decent chain of thought, which is nice, right?
00:15:04.799 --> 00:15:14.240
Because that's one thing that I think uh Google and Gemini is really lacking in uh compared to OpenAI and Anthropic, is being able to see what's actually going on under the hood.
00:15:14.320 --> 00:15:16.000
So little cheat code for you there.
00:15:16.159 --> 00:15:23.440
Uh Gemini Notebook is actually probably better than the default Gemini in terms of knowing and learning what's going on under the hood.
00:15:23.600 --> 00:15:30.559
So, you know, I'm not gonna read all of the chain of thought because it's a lot, but you can go try this yourself, you know, click thoughts and see what's going on there.
00:15:30.639 --> 00:15:41.120
Uh, but it's developing budgeting solutions, it's evaluating the AI strategies, analyzing the model costs, um, you know, it's retrieving some instructions, analyzing strategies, analyzing pricing, right?
00:15:41.200 --> 00:15:43.600
So I can go through and see exactly what it's doing.
00:15:43.679 --> 00:15:45.519
Um, and I'm looking at some of these things.
00:15:45.600 --> 00:15:48.960
I can see that it's creating, uh, you know, creating different files.
00:15:49.039 --> 00:15:50.480
So it's running some code.
00:15:50.559 --> 00:15:54.240
Um, so you know, it's executing code in there in its dedicated sandbox.
00:15:54.399 --> 00:15:57.039
So this was not possible a week ago.
00:15:57.120 --> 00:16:00.799
So the output here, I'm not gonna read it because it's fairly long, but I'll read the beginning.
00:16:00.960 --> 00:16:03.840
So it says executive AI budget decision package.
00:16:04.000 --> 00:16:11.919
This package is designed for the CFO and CIO of our 5,000-person enterprise to address rapidly growing AI agent usage.
00:16:12.080 --> 00:16:20.879
Below, we analyzed three potential model procurement strategies and delivered five unified synchronized outputs to drive your strategic budgeting decision.
00:16:21.120 --> 00:16:22.559
All right, so we'll see here.
00:16:22.720 --> 00:16:25.440
I'm not gonna look at this, but it is exhaustive, right?
00:16:25.519 --> 00:16:28.639
And the good thing is, um, all of this is sourced as well.
00:16:28.799 --> 00:16:36.159
So, you know, it's saying, you know, workflow tokens, you know, for planning, you know, using 20,000 input, two uh 2,000 output.
00:16:36.240 --> 00:16:39.919
I can hover over that and see exactly where it's pulling that from in my notes.
00:16:40.080 --> 00:16:45.840
Uh so you know, when I'm looking over the different strategies here, it did a pretty good job putting together three different strategies.
00:16:46.000 --> 00:16:55.919
One's uh strategy one was using frontier models only, uh, strategy two was using cheap models only, and then strategy three was a tiered architecture using a combination of both.
00:16:56.000 --> 00:17:02.399
And then it went through and, you know, uh went through and did these kind of combined monthly system costs, right?
00:17:02.480 --> 00:17:18.640
And you know, what's kind of cool in this scenario is the tiered architecture, um, right, which is kind of using the stronger model as an orchestrator, at least according to its calculations, um, that was actually gonna be better, right, than just doing the cheap only.
00:17:18.720 --> 00:17:28.480
And the cheap only was actually gonna be more expensive, uh, presumably because it was gonna go through a lot more tokens and more monthly human review rework costs, right?
00:17:28.559 --> 00:17:31.279
So such a cool uh tool to put together.
00:17:31.599 --> 00:17:37.680
All right, but then let's look over at our uh uh right-hand side in our panel and see what was actually created.
00:17:37.839 --> 00:17:42.559
So we did ask specifically for five different uh deliverables.
00:17:42.720 --> 00:17:46.480
We asked for a customized five-minute audio overview, which is almost done.
00:17:46.640 --> 00:17:50.640
Uh, number two, we asked for a mind map connecting pricing changes.
00:17:50.799 --> 00:17:53.359
We asked for an editable Excel workbook.
00:17:53.599 --> 00:17:58.799
We asked for a nano banana powered 16 by nine infographic and a two-page PDF.
00:17:58.880 --> 00:18:03.200
So, so many of these new, so many of these things, again, were not available last week.
00:18:03.359 --> 00:18:13.440
You know, you really had to stay in this kind of confined output of what uh Notebook LM offered, but now with Gemini Notebook, right, it's really flexible.
00:18:13.599 --> 00:18:20.000
And that is actually really helpful, both from what you can do in the middle pane, because you don't have to be um as rigid, right?
00:18:20.160 --> 00:18:35.039
But in my use case, right, if you really want to have agents working for you, I think this makes it so much easier for a you know clawed desktop or a codex desktop, you know, an agent uh living, you know, on your desktop that can control your browser.
00:18:35.200 --> 00:18:38.880
This is going to allow it to iterate and create much more valuable outputs for you, right?
00:18:38.960 --> 00:18:42.880
Um, I always want to encourage listeners of our show, even if you're a beginner, right?
00:18:43.039 --> 00:18:45.279
We have to get out of the prompting phase, right?
00:18:45.440 --> 00:18:53.039
We have to be able to give these desktop agents as much context and as much information about what we need, um, then let them go do the work.
00:18:53.119 --> 00:18:58.799
And I'll usually have another uh, you know, AI agent audit their work and make sure it, you know, cracks any mistakes that it sees.
00:18:58.880 --> 00:19:06.640
But in this scenario, right, if I were to do this manually, even inside Nobook LM, it would have taken me three to five times longer.
00:19:06.720 --> 00:19:08.720
Um, but that's just because of the new harness.
00:19:08.880 --> 00:19:11.920
So let's just quickly take a look at all of our other sources.
00:19:12.079 --> 00:19:14.720
So here we have our AI budget model.
00:19:14.880 --> 00:19:15.519
All right.
00:19:15.920 --> 00:19:18.400
So downloaded uh an Excel file.
00:19:18.480 --> 00:19:20.640
I'm just gonna screenshot this here.
00:19:20.799 --> 00:19:25.440
Um, let's do that versus having to unshare and reshare my screen.
00:19:25.599 --> 00:19:26.880
Uh let's see.
00:19:28.799 --> 00:19:30.720
It's uploading, it's uploading.
00:19:30.880 --> 00:19:32.400
Oh gosh, let's see.
00:19:32.880 --> 00:19:34.240
All right, that one.
00:19:34.319 --> 00:19:36.160
Let me uh let me try that again.
00:19:36.240 --> 00:19:39.839
Uh the actual Excel sheet looks looks great.
00:19:40.240 --> 00:19:42.079
Um, we have our budget model.
00:19:42.240 --> 00:19:44.480
Uh let's let's see if we can get that again here.
00:19:44.640 --> 00:19:46.880
Uh let's copy this, shall we?
00:19:47.119 --> 00:19:49.759
Um, here we go.
00:19:51.039 --> 00:19:51.680
There we go.
00:19:51.759 --> 00:19:52.559
We got it here.
00:19:52.720 --> 00:19:57.200
Uh so we got a working uh financial model, right?
00:19:57.279 --> 00:20:02.079
I just quickly checked the spreadsheet, um, gave me an XLS, opened it in Excel.
00:20:02.240 --> 00:20:03.839
Everything's working, uh, right?
00:20:03.920 --> 00:20:05.519
It gave me exactly what I wanted.
00:20:05.599 --> 00:20:07.920
It gave me my different uh three strategies.
00:20:08.079 --> 00:20:09.759
Um, it mapped out the cost.
00:20:09.920 --> 00:20:11.599
I can go in and change the formulas.
00:20:11.680 --> 00:20:14.880
There's a little graphics, um, a little dashboard.
00:20:14.960 --> 00:20:16.880
It created exactly what I wanted.
00:20:16.960 --> 00:20:17.359
It works.
00:20:17.440 --> 00:20:19.200
Is it the most beautiful spreadsheet ever?
00:20:19.279 --> 00:20:19.519
No.
00:20:19.680 --> 00:20:21.519
Is it something better than I could have done?
00:20:21.680 --> 00:20:22.640
Absolutely, right?
00:20:22.799 --> 00:20:27.359
I love uh working in spreadsheets, but I'm not necessarily the best at creating formulas, right?
00:20:27.440 --> 00:20:30.960
So again, uh just think of all of the unlocks that something like this brings.
00:20:31.119 --> 00:20:33.920
All right, so that was our AI budget model, did a good job.
00:20:34.000 --> 00:20:39.440
Uh, then we asked for a two-page uh PDF that gives an executive decision brief.
00:20:39.519 --> 00:20:41.359
All right, so now uh here we go.
00:20:41.519 --> 00:20:44.799
Live stream audience can see, but we have a two-page executive brief.
00:20:44.960 --> 00:20:52.160
This is the AI portfolio optimization, uh, the implement uh implementing a tiered model architecture for the 5,000-person enterprise company.
00:20:52.319 --> 00:20:52.880
There we go.
00:20:53.039 --> 00:21:04.640
This is, you know, just kind of a PDF version of what we uh created inside the middle pane chat, um, as well as why, you know, what the uh the best strategy was and the counterarguments for those.
00:21:04.720 --> 00:21:05.440
So perfect.
00:21:05.599 --> 00:21:06.559
It created that.
00:21:06.720 --> 00:21:08.720
Uh let's look at the other ones.
00:21:09.039 --> 00:21:10.640
We have our mind map.
00:21:10.799 --> 00:21:13.680
Okay, so here we have our uh RSI here.
00:21:13.759 --> 00:21:21.599
Uh, then we have the economic shifts in pricing, the tiered agentic architect uh architecture, governance and control, and monthly system cost scenarios.
00:21:21.680 --> 00:21:26.559
And then I can click those out um and explore a little bit more uh in each of those.
00:21:26.720 --> 00:21:27.920
There we go.
00:21:28.240 --> 00:21:31.200
All right, so uh mind map, great job.
00:21:31.359 --> 00:21:36.960
All right, and then let's look at our uh this should be our infographic created by Nano Banana.
00:21:37.119 --> 00:21:39.279
All right, so looks good, you know.
00:21:39.359 --> 00:21:43.680
I'm I'm kind of uh, you know, quickly editing it and it looks good.
00:21:43.839 --> 00:21:47.839
The only thing I see wrong is there's a notebook LM watermark.
00:21:47.920 --> 00:21:49.519
All right, Google, we got to update that, right?
00:21:49.599 --> 00:21:54.799
If we're trying to get rid of notebook LM, uh, we gotta update the the watermark to Gemini Notebook.
00:21:54.880 --> 00:21:57.599
Uh, but the it this actually looks pretty good.
00:21:57.759 --> 00:21:59.680
Um, the uh 16 by nine Nano.
00:22:00.160 --> 00:22:04.319
Banana uh graphic for the CFO, CIO decision maker.
00:22:04.559 --> 00:22:05.039
Great.
00:22:05.200 --> 00:22:15.279
Um looking at the you know three different strategies, it breaks down the human cost, the model cost, the the pros and the cons, uh great charts and graphics, did a really good job.
00:22:15.440 --> 00:22:17.519
And then last but not least, it did just finish.
00:22:17.759 --> 00:22:19.119
I'm just gonna listen here.
00:22:21.279 --> 00:22:22.240
All right, perfect.
00:22:22.400 --> 00:22:28.640
It created a 21-minute deep dive on self-improving AI slash token prices.
00:22:28.960 --> 00:22:32.960
All right, so um okay, interesting.
00:22:33.119 --> 00:22:36.160
Um, so it even did put in the custom prompt.
00:22:36.240 --> 00:22:37.200
Uh, so that's cool.
00:22:37.279 --> 00:22:40.400
It went and did everything correctly, right?
00:22:40.480 --> 00:22:42.480
So, hey, the demo actually worked.
00:22:42.640 --> 00:22:46.480
All right, I'm just gonna quickly look at some of the other use cases just so we can see what we did.
00:22:46.720 --> 00:22:54.000
Uh in the other notebook, I said create three quizzes of ascending difficulty focusing on a different topic from my sources.
00:22:54.079 --> 00:22:55.119
Uh, this is a good one.
00:22:55.279 --> 00:23:00.400
This was actually um a Gemini notebook uh example prompt, which I thought was a really good idea.
00:23:00.559 --> 00:23:05.119
Um, if you're trying to learn something, you know, they have this quiz module, which is really good, right?
00:23:05.279 --> 00:23:13.359
But uh, you know, sometimes you might really want to be intentional about how you learn things and break it off and start easier so you can go through the paces and relearn things as you go.
00:23:13.519 --> 00:23:16.160
Uh so this one, it created an easy quiz.
00:23:16.319 --> 00:23:16.799
All right.
00:23:16.880 --> 00:23:19.039
Uh, let's should I do one question?
00:23:19.200 --> 00:23:20.480
Watch, watch me get it wrong.
00:23:20.640 --> 00:23:25.119
All right, what is the core definition of recursive self-improvement in the context of AI?
00:23:25.279 --> 00:23:29.599
A, is it a method where humans manually rewrite every line of code to make AI faster?
00:23:29.839 --> 00:23:44.160
B, a process where an AI uses its own capabilities to design and build a more advanced version of itself, C, an AI system that is only capable of performing one specific task like plain chess, or D, a way for AI to search the internet and summarize existing news articles better.
00:23:44.240 --> 00:23:45.440
So I know it's B.
00:23:45.599 --> 00:23:46.640
All right, so that's right.
00:23:46.799 --> 00:23:51.039
So it has three different quizzes, did a great job doing that of ascending difficulty.
00:23:51.200 --> 00:23:52.960
All right, let's look at our next one.
00:23:53.119 --> 00:23:57.359
Uh, so for this one, I said create a five-level RSI maturity ladder.
00:23:57.440 --> 00:24:00.079
And I wanted an infographic and a PDF.
00:24:00.240 --> 00:24:01.599
So it did answer everything.
00:24:01.759 --> 00:24:02.720
This is really good.
00:24:02.799 --> 00:24:04.559
Uh, it answered everything in the chat.
00:24:04.640 --> 00:24:10.799
Uh, my PDF here, it broke down the five uh kind of ladders of RSI, did a great job.
00:24:10.880 --> 00:24:12.640
And then let's look at the infographic.
00:24:12.720 --> 00:24:13.680
Uh, cool, yeah.
00:24:13.759 --> 00:24:20.400
This infographic actually came out a little cleaner than I thought, considering that uh we didn't uh customize anything.
00:24:20.480 --> 00:24:23.599
So yeah, we have the uh level one, AI assists researchers.
00:24:23.680 --> 00:24:26.319
Level two, AI executes a human design method.
00:24:26.480 --> 00:24:28.880
Level three, AI improves AI infrastructure.
00:24:29.039 --> 00:24:31.920
Level four, AI helps train or improve another AI.
00:24:32.000 --> 00:24:35.039
And then level five, AI controls the complete improvement loop.
00:24:35.279 --> 00:24:36.240
All right, let's go.
00:24:36.319 --> 00:24:37.200
Let's look at our next one.
00:24:37.279 --> 00:24:43.440
I said using the old Luna and Terra pricing in this notebook, create an edible Excel calculator for agentic workflows.
00:24:43.680 --> 00:24:45.440
I have my uh Excel sheet there.
00:24:45.519 --> 00:24:47.440
I'm opening it up in another tab.
00:24:47.839 --> 00:24:50.559
Um yeah, looks good.
00:24:50.640 --> 00:24:51.359
I'm checking.
00:24:51.519 --> 00:24:54.480
There's formulas, everything's working live, perfect.
00:24:54.640 --> 00:25:01.680
All right, and then uh one more thing uh before we wrap up here that I wanted to showcase.
00:25:01.920 --> 00:25:02.559
All right.
00:25:04.559 --> 00:25:07.039
The middle is agentic now.
00:25:07.200 --> 00:25:18.160
All right, so even though it's still going to be grounded by default, uh, I'm gonna go into one of my things here and I'm gonna type who won the 2016 World Series.
00:25:18.240 --> 00:25:19.839
All right, now I'm gonna ask that question.
00:25:20.000 --> 00:25:23.599
So by default, it should come back to me and say, hey, I don't know.
00:25:23.759 --> 00:25:26.880
However, I can go out and find that information.
00:25:27.039 --> 00:25:29.680
All right, so why am I bringing this up?
00:25:30.160 --> 00:25:34.960
Because that is the one important grounded change that you need to be aware of.
00:25:35.200 --> 00:25:42.720
All right, so what's interesting here, um, and I didn't know this, it looks like because it had already started pulling things agentically on the web.
00:25:42.799 --> 00:25:51.279
Um, because in some of my other testing, when I did this for the first time, um, you know, so I just opened up another uh chat to show you this.
00:25:51.440 --> 00:25:53.519
I said, who won the 2016 World Series?
00:25:53.599 --> 00:25:57.279
In this example, it says your sources do not contain information about this.
00:25:57.440 --> 00:25:59.519
Do you want me to research this on the web?
00:25:59.599 --> 00:26:00.400
I said yes.
00:26:00.559 --> 00:26:08.079
So it looks like maybe if web research had already been active, uh had already been activated, that it might just answer that by default.
00:26:08.160 --> 00:26:16.000
So that's an important thing to think about because previously, um, notebook LM in the middle chat was always 100% grounded in your information.
00:26:16.160 --> 00:26:18.319
So that's a big difference and a big change.
00:26:18.480 --> 00:26:22.720
So in some instances, it's a big upside, but you have to be aware of that.
00:26:22.799 --> 00:26:37.279
Um, that you know, you have to be able to go back and look at the chain of thought in a lot of these prompts, it went and it was doing a lot of web research in addition to the information that I gave it, because in the prompts, I was kind of pushing it past the limitation of the sources.
00:26:37.440 --> 00:26:41.920
So this does kind of, in my opinion, change uh Gemini Notebook.
00:26:42.000 --> 00:26:49.519
And maybe that's good that we are getting away from the notebook LM in inserting Gemini Notebook because there's pros and cons to that, right?
00:26:49.680 --> 00:27:02.160
Like you saw in this example here, because it looks like because I had already triggered the anti-gravity harness um in that first prompt and then asked it information that was not in my sources, it just well, it just responded.
00:27:02.319 --> 00:27:02.799
All right.
00:27:02.880 --> 00:27:11.440
Uh, but uh it did say it says I had I haven't imported those sources about the 2016 World Series.
00:27:11.920 --> 00:27:12.720
Oh, wait.
00:27:12.960 --> 00:27:13.759
I looked at it.
00:27:13.920 --> 00:27:15.119
Okay, I didn't read.
00:27:15.279 --> 00:27:15.839
It didn't.
00:27:16.079 --> 00:27:16.480
All right.
00:27:16.559 --> 00:27:19.359
Um, it just said that I hadn't looked at those.
00:27:19.759 --> 00:27:21.359
Uh, do you want me to look them up?
00:27:21.440 --> 00:27:23.039
Uh I I got ahead of myself.
00:27:23.200 --> 00:27:24.640
Uh so yeah, I can say yes.
00:27:24.799 --> 00:27:25.119
All right.
00:27:25.200 --> 00:27:35.839
So, all right, scratch what I just said the last 60 seconds that um I thought because I had already activated the anti-gravity uh harness and it was already doing uh tool calls that it was gonna go out and respond.
00:27:35.920 --> 00:27:37.279
I literally just didn't read it.
00:27:37.359 --> 00:27:39.759
I just saw the response and I'm like, oh my gosh, it responded.
00:27:39.920 --> 00:27:42.000
But no, me uh not sleeping enough.
00:27:42.079 --> 00:27:46.240
It literally said, I haven't imported those sources about the 2016 World Series into your notebook yet.
00:27:46.400 --> 00:27:47.759
Would you like me to import them first?
00:27:47.920 --> 00:27:48.799
And then I said yes.
00:27:48.880 --> 00:27:51.920
And now, right now, I can click and see that it's going out to grab those sources.
00:27:52.000 --> 00:27:53.200
All right, I stand corrected.
00:27:53.519 --> 00:28:02.960
So it's still important to know though that you know, once you do give Gemini Notebook access to, it is going to bring in information that are outside of the sources.
00:28:03.039 --> 00:28:13.440
So it does change, I think, again, for the good and the bad, but at least uh it does caution you uh or make you kind of manually approve that's saying, Hey, you don't have this information in your sources.
00:28:13.599 --> 00:28:15.759
Do you want me to go grab that information for you?
00:28:15.920 --> 00:28:18.559
All right, so that is a wrap.
00:28:18.720 --> 00:28:21.680
But to quickly recap, what do you need to know?
00:28:22.000 --> 00:28:26.400
All right, so uh number one, what's the big difference here?
00:28:26.880 --> 00:28:33.759
Gemini Notebook represents a shift away from just having that, you know, notebook LM, all those sources.
00:28:33.920 --> 00:28:41.839
And, you know, essentially it was a non-reasoning model that could use Google's different multi uh multimedia AI.
00:28:41.920 --> 00:28:43.119
And now it's different, right?
00:28:43.200 --> 00:28:49.839
And I think that's the the big reason for the name change is because now with the Gemini Notebook, it is a gentic by default, right?
00:28:49.920 --> 00:29:00.319
Having that cloud computer, being able to write and execute code, being able to, when you give it the okay, search the web, it really changes what Notebook LM was into what Gemini Notebook will be in the future.
00:29:00.720 --> 00:29:04.400
You know, the important grounded change that I talked about right there, you just have to know.
00:29:04.480 --> 00:29:08.240
And I think ultimately that will be a good thing as long as you understand what's going on.
00:29:08.319 --> 00:29:14.480
Um, and then the real reason I think notebook, uh Gemini Notebook going agentic matters more than ever.
00:29:14.559 --> 00:29:20.079
Uh, well, you saw just through my example, it makes it much easier to work with, right?
00:29:20.240 --> 00:29:24.079
Uh before, um, you know, you were kind of restrained, right?
00:29:24.240 --> 00:29:28.400
You almost had to think in your mind um in terms of outputs, right?
00:29:28.559 --> 00:29:32.079
You had to prompt notebook LM according to the outputs.
00:29:32.240 --> 00:29:34.240
Now those outputs are so flexible, right?
00:29:34.319 --> 00:29:37.680
Before you couldn't create PDFs, before you couldn't create CSVs, right?
00:29:37.759 --> 00:29:39.759
You couldn't create markdown files.
00:29:39.839 --> 00:29:42.799
So you were much more limited in, you know, PowerPoints, right?
00:29:42.880 --> 00:29:43.920
You can do all these things now.
00:29:44.079 --> 00:29:52.480
So uh, you know, Gemini Notebook has become uh much more of an agentic uh, you know, coworker versus what it was before.
00:29:52.640 --> 00:30:00.240
You know, there's essentially a handful of uh artifacts or a handful of formats, and you really had to use Notebook LM according to those outputs.
00:30:00.400 --> 00:30:03.759
Now it is extremely flexible and even more powerful than ever.
00:30:04.000 --> 00:30:10.160
And I guess that's why Notebook LM is no more, but Gemini Notebook at least is here to say, and it is really good.
00:30:10.319 --> 00:30:12.160
All right, that's a wrap for today's show.
00:30:12.240 --> 00:30:15.519
Now you know Gemini Notebook, the seven new updates and what they unlock.
00:30:15.680 --> 00:30:19.759
If this was helpful, do me a favor, tell someone about it, share this on LinkedIn.
00:30:20.000 --> 00:30:23.680
Uh, you know, go and subscribe to the podcast on Apple or Spotify.
00:30:23.759 --> 00:30:27.759
Then make sure to go to your everydayai.com, sign up for the free daily newsletter.
00:30:27.839 --> 00:30:30.559
We're going to be recapping the highlights from today's show as well as a lot more.
00:30:30.720 --> 00:30:31.519
Thanks for tuning in.
00:30:31.680 --> 00:30:34.559
Hope to see you back tomorrow and every day for more everyday AI.
00:30:34.640 --> 00:30:35.599
Thanks, y'all.
