Claude Code + NoteBookLM = Infinite Memory

Imagine combining Claude with the world's number one intelligence platform, Google's Notebook LM. I did

this and it solves Claude's number one problem, giving Claude a long-term memory that will cut your costs and a lot of capabilities that 99% of people don't even know exists yet. And in this

video, I'll show you the exact setup so you can get this full system with near infinite memory, with virtually zero [music] token cost, with a setup that just takes a few minutes. And if you don't know who I am, my name is Jack

Roberts. I built and sold my last tech

Roberts. I built and sold my last tech startup [music] with over 60,000 customers. And now I teach thousand

customers. And now I teach thousand entrepreneurs and run AI businesses of my own. So if you haven't already, grab

my own. So if you haven't already, grab that coffee and let's dive straight in.

When you combine Claude and Notebook LM, you get a second brain. Now, what I'm going to show you in this video not only works for Claude Co-work, it also works for Claude code as well. Now, here's the

thing. Claude has what we call amnesia.

thing. Claude has what we call amnesia.

What do I mean by that? Well, every

session that you have with it, usually you're starting back from zero, minus any long-term memory that you solved in the system and it can forget things really easily. Now, there's one tiny

really easily. Now, there's one tiny exception to this. I'm going to put a video on screen where I covered this yesterday. So, go and check that video

yesterday. So, go and check that video out on co if you haven't seen it. But

the key point still remains. Now, it can solve this to some extent by reading loads of files. But the problem with reading files is not only to increase your context, it also costs money, right? Which is why people run out of

right? Which is why people run out of these tokens too quickly. And Claude

themselves are working on improving it.

They're doing a phenomenal job. They're

shipping like crazy right now. I have so much coffee in Claude HQ. So, they are working on this stuff on memory. But the

thing about notebook, the world's number one researcher intelligence platform, is it always remembers and it does three things that really change this whole game. Number one is it will improve

game. Number one is it will improve memory for conversations you have in Claude code and also in codework. It

will cut your costs so you're not spending as many tokens to get those answers. And on top of it, it also

answers. And on top of it, it also introduces so many new skills that can move your personal life or your business forward in so many different ways. So

the best way to realize just how good this is is to show you across three use cases that you can implement straight away. But if you don't understand the

away. But if you don't understand the actual benefits and what notebook LM can do, you're not going to be able to reach its full capabilities. And so if you take a look at what notebookm can do within Claude, right, it's pretty cool.

So look, you've got persistent project memory, which means that any decisions that you make, any rational context survives across every session. How

freaking cool is that? It can function in many respects like your personal CRM.

Decision journal is really cool.

Personal archive for different ideas, institutional knowledge. So if you're

institutional knowledge. So if you're onboarding new people into your team and also meeting intelligence, any meetings you do, we can all fit into the book of Lamb. And we are under the careful and

Lamb. And we are under the careful and very studious watch of Claude Code himself with a really cool hairline having a good time. Now even from notebook LM we can do many things. What

we call content multiplication. I'm

going to show you this in a second. One

source anything we talk about any project we can create podcast, video creation, social content mining, newsletter fuel, build presentations, audio overviews, you literally name it.

Now the first thing we need to do before we actually go ahead and begin the skills is pretty much get it installed into Cloud. So how do we do that? Well,

into Cloud. So how do we do that? Well,

I'm going to put all the links for you down below in the description. We're

going to move to my classroom and grab this. Now I am basically created for you

this. Now I am basically created for you a skill that will combine all this together. There's two skills in the

together. There's two skills in the video. The first one I'm going to use

video. The first one I'm going to use here is a Claude notebook LM skill. Not

only will this install the skill and connect it to notebook alarm, but it also has many features that's going to help you connect this to Claude Co if you wish. Beautiful. So now I'm in

you wish. Beautiful. So now I'm in Claude. It's a brand new Mac. So we're

Claude. It's a brand new Mac. So we're

going to be installing this together. So

let's go ahead and drop the skill in. So

what we can do is literally come down here and drop the file in. And all

you're going to say is execute this skill. Simple as that. And then hit

skill. Simple as that. And then hit enter. and let it do its wondrous things

enter. and let it do its wondrous things in the background. So, we're going to go ahead and use claude code initially to get this one installed. And what it's going to ask us to do is simply sign into notebook lm and it's going to grab

a token. So, this is an unofficial

a token. So, this is an unofficial connection. So, it's going to go ahead

connection. So, it's going to go ahead which means that somebody's basically built this Python script cuz we're going to download it from a GitHub repo. But

again, it's already going to do all this stuff for us uh automatically because we explain it all in this skill. Then when

we've done that, we'll basically sign in. We'll get the cookie. And then once

in. We'll get the cookie. And then once we've got that token, anything that you can do in Notebook Alan, the creation of projects, the creation of assets, the asking questions, we can all do from

within Claude. Now, how incredible is

within Claude. Now, how incredible is that? Like that's literally insane. And

that? Like that's literally insane. And

you think about how powerful it is. It's

rag system. It's a completely different way of doing memory. A very efficient way of doing memory. So instead of cramming in like several pages or several hundred pages of context about our business or meetings, we just reach

over and grab the one thing that we need. And then what will happen is cloud

need. And then what will happen is cloud code run some commands and it'll pop up with this. So we're going to say yes,

with this. So we're going to say yes, open up the browser. So if you haven't authenticated yet, you're going to click on this one here and it will open a browser and then we're just going to sign in basically to notebook LM.

Wonderful. So this will pop up and you're just going to enter in your information. And then just like that,

information. And then just like that, you'll be signed into Notebook LM, which means it's all authenticated together.

And once that's complete, it will basically say notebookm is fully up and working and it can see your notebook. So

let's just ask it a question. Hey there,

tell me the title of my last three notebooks. And then from my last latest

notebooks. And then from my last latest notebook, go ahead and download for me the video that was last generated. So

we'll just test and validate that actually works fine. And then it's actually found it, which is the mathematics of aentic drift. Did you

know you were super interested in that?

So was it great? Just save it to my downloads folder as if for example, by the way, Notebook LM can now do cinematic videos, which you're going to see in just a second. Awesome, dude.

Open that up for me, please. And it

literally opens up. And this was made by Notebook LM. So check this out. The

Notebook LM. So check this out. The

artificial intelligence industry is currently shifting from the engineering of single model prompts to the design of agent orchestration. Mation moves down

agent orchestration. Mation moves down the chain with each handoff.

You know what? Not perfect, but genuinely impressive. Notebook LM can

genuinely impressive. Notebook LM can build assets like this for anything that you're working on. It is crazy. Now, the

cool thing is not everybody's going to be using cloud code. A lot of people live in Cloud Co-work. So, all you're going to do now to get into co-work and all your co-workers is the following because they're different environments.

I'll say awesome. Now I'd like you to add notebook lm as a co-work skill please. Now the cool thing here is that

please. Now the cool thing here is that claude code is going to generate the document for us that we can upload to co-work as a skill because claude code by itself can't grab a token. So claude

code solves that for us. Beautiful. So

now it saved it to our desktop as notebook lmkll.workmd. And by the way

notebook lmkll.workmd. And by the way all these instructions are in the file that I gave you. So we can just do the same thing. So then we click on co-work.

same thing. So then we click on co-work.

Then within co-work click on this plus.

Come down here to skills. Click on

manage skills. And over here, you're going to click on this button here at the top and click on upload a skill. And

then just upload that same document. So

here, we just click on this. And then

you click on that skill. As you can see, it's got the description, which is awesome. Then click on open like so. And

awesome. Then click on open like so. And

then we're going to say yeah, upload and replace. I did it earlier on my laptop,

replace. I did it earlier on my laptop, so that's why it's showing there. That's

cool. Now it's all there ready to rock and roll. So if I come over to new task,

and roll. So if I come over to new task, I should be able to validate that. Hey

there. What was the last notebook that I created? Cool. I'm just drop that one

created? Cool. I'm just drop that one down there real quick. And just to validate it's actually fully connected.

And it should use the notebook LM skill by itself. There you go. Running skill

by itself. There you go. Running skill

fetching the result and then it can give us some details. And now you can see in co-work it's now integrated fully with notebookm and it's pulled down the latest thing from that folder. Now there

are three critical use cases that you can implement immediately with the last one being something I think you're really really going to want to use. The

first one though is pretty crazy and that's what I call enrichment. It's the

idea of taking something that's already great in any project and it gives you the ability to get insights that are significantly deeper so you can make better decisions. For example, let's say

better decisions. For example, let's say in co I could be in a project which just makes this even more powerful. But let's

say that I'm having a conversation here about, hey, what is best practice in scaling a YouTube channel to a million subscribers? Let's say that we're having

subscribers? Let's say that we're having this conversation with Claude, right?

And we're going back and forth and we've got all our folders, all our contacts.

And if you haven't seen my last video, I linked it earlier on. And there'll be a link um down below at the end of the video as well where you can check out.

But what's really interesting here is that it can leverage all of your projects, all of your context. And what

we could do then is leverage notebook to get even deeper research to help us with anything that we're working on within Claude Code. For example, what I'm going

Claude Code. For example, what I'm going to say to it now is look, awesome. What

I'd now like to do is to create a brand new notebook based on everything that you know about me and this conversation as well as any other additional context within this project. Once you've done

that, come back to me and then I want you to give me a distilled 10point insight with three core actions I need to do immediately based on everything that you know. Very very cool. Now this

doesn't have to be in co projects. This

could work in cloud code. This could

work in a regular chat. But if you do this in projects, you get additional context that we can leverage. And now

this will run in the background. It'll

run the skill. And all we do guys is we grab our coffees, we relax, and then we come back to some beautiful insights.

And there we go guys. brought down 10 distilled insights based on this entire project. It's got over 70 sources

project. It's got over 70 sources loaded, including your full channel strategy, performance data over the last 30 videos, four of my top performing videos, transcripts, my school community

page, and two deep website research passage on YouTube on scaling. And it's

giving me details like, hey, your channel is a conversion machine. Well,

that's very generous of you to say, but we appreciate that. But what I'm going to say to it now is, awesome. Could you

just open up that notebook, LM, so I can actually check out the full research that's gone into this, please? And just

like this guys, it's got over 70 different resources, which is crazy.

Occasionally, if it times out, it might miss this research. And as you can see, even now, it's adding it all in right there, which is amazing. And if it does that, fair enough, but just make sure you give it that little nudge to say, "By the way, always make sure that you import that, too." Because when you come

back and ask them if they imported it, it says, "Great. Yeah, it timed out. So,

they're going to fix it." So, just make sure you always have that follow-up before getting the data. Now, this

shouldn't happen a lot, but if it does, just ask if it's imported to deep research, give it that prompt ahead of time, and it will just import the data for you. Having great enrichment is one

for you. Having great enrichment is one thing, but when we're connected to notebook LM via cloud code and cloud codework, we can enrich our understanding and create assets programmatically, which is something that we couldn't do before. Let's say,

for example, this time we're in Claude Code, right? And we're asking um Claude

Code, right? And we're asking um Claude code about some interesting things like, hey, what is the best strategy for a five-step sequence email campaign? I'm

launching it with my client's business.

I want to make sure he gets maximum conversions. What would be the best way

conversions. What would be the best way of doing that? So, let's say we've got this right and we're presenting to our client on Tuesday. He's like, "Jack, I need a full email breakdown." Well, one of the things that we can do is obviously chat to Cloud Co to build out

a little bit of an understanding. It can

do research and just like in step one, we could even use Notebook LM to get the power of 40, 50, 60 different notebooks and get that deliberate insight. So,

we've come back here and we got a five-step um email sequence framework, which is really cool, decent. But what

if I say then I'm like, cool, well, I'm presenting this back to the client. What

if I want to create an infographic, right? How about I say something like,

right? How about I say something like, "Awesome. Could you do me a favor? Could

"Awesome. Could you do me a favor? Could

you create for me an infographic in Notebook LM using some gorgeous colors, premium fonts, and then just bring that back to me, please?" Now, why is this cool? This is cool because we don't pay

cool? This is cool because we don't pay anything in Notebook LM. Like, it's free generation for infographics, and you can generate pretty much anything you want to with it. And on top of that, I just want to explain exactly what Notebook LM

can do for research in this context.

It's super important. It can do deep research, give us competitive intelligence. So, it isn't just emails.

intelligence. So, it isn't just emails.

We can do market synthesis. So ingest 20 articles and get one coherent summary due diligence lit review trend spotting capabilities that would either take Claude ages to do or it just isn't as equipped as the systems that we built

over in Google. I say we've built I wasn't involved in it. Maybe they should have brought me on that would have been a great thing. Then on top of that we've got these project and business operations. So postmortm analysis we can

operations. So postmortm analysis we can feed in artifacts ask why what went wrong and why. Quality assurance

grounded proposals feed in RFPs and have that come back. client deliverable

enrichments and even SOP and playbook.

So the idea here is that we can leverage notebook to do things and access capabilities without burning any of our tokens included because we make the API call and then notebook does all the magic and then we just get the actual

diamond the actual specific result back.

And what this can also do is solve the issue that some people run into with obsidian where it blazes up your context usage where people feel they get locked out of co-work or their basically their

tokens just go down so quickly. using

this strategy. It's like a jab and then we get the big results back. We send off a query. It does all the processing on

a query. It does all the processing on Google side. And then we just get the

Google side. And then we just get the beautiful insights that we can build over time. Beautiful. And just like

over time. Beautiful. And just like that, it's got a fivestep email sequence blueprint. Awesome, dude. Open this up

blueprint. Awesome, dude. Open this up for me, please. And let's take a good look at what it looks like. This is

really freaking cool, guys. And again,

you paid nothing for this. This is just all run. Here we go. We can open it up

all run. Here we go. We can open it up and take a look. This is the infographic. So, this is an old

infographic. So, this is an old download. Let's go open it up. And there

download. Let's go open it up. And there

you go, guys. Check this out. the hook,

the story, the value drop, the pitch, the closer. And bear in mind, I can

the closer. And bear in mind, I can specify how I want this to look. We can

do many different again, anything you can do in notebook alarm, you can effectively do this also, which is just crazy. And this leads on to one of the

crazy. And this leads on to one of the largest problems that exist with Claude, and that is it's a long-term memory. And

how do you actually solve it without spending an absolute fortune?

Incredible. Well, it turns out there's a very unique solution, and it leverages this notebook LM connection. Now to do this, we're going to get the second skill of the video which is going to be available for you down below completely for free. All you need to do is come

for free. All you need to do is come down and click on this wrap up skill and explain exactly what this is. So if you think about let me sh get download this here right if you check out this website here which is what I use my community

build out the road map. If you come on to YouTube chat you can chat with me and everyone in my videos and ask questions and get resources from the community.

Now what's happening here under the hood when you ask it a question it is searching a vector database and based on that database we'll pull back guides and stuff right so here's a community post

about rag I click on this I can see it it's there how on earth has it done that well it's got some really interesting ways using a retrieval augmented generative system which is really cool

now notebook lm works exactly the same way it's got a rag system so let's come back over now to claude coowork this works in code in exactly the same way.

Now, what we need to do is add a skill.

So, to do that, I want you to come over here to the bottom left, click on the plus. Then click on skills and click on

plus. Then click on skills and click on manage skills. Then over here, you're

manage skills. Then over here, you're going to click on this one here. You're

going to click on upload a skill. And

then right here, and then click on this wrap-up skill, like so. And once that's done, that will finish loading. And now

it has fully uploaded. Now, how does the wrap-up skill work? So, here's the idea behind it, okay? Is that you have a conversation with her and effectively once you've had a large session, you can activate the wrap-up skill. The wrap-up

skill will take everything from the conversation and store it in Notebook LM. Now, think about this. You're going

LM. Now, think about this. You're going

to basically have this wonderful evergrowing system at Notebook LM that can effectively retrieve any information you want to with one simple call.

Meaning that all of your data stays exactly in one place without you having to build up an insane context window in Claude. So, for example, if I come down

Claude. So, for example, if I come down here and I do the wrap-up skill like so, I click on wrap-up. Once I've done in a large conversation and it's contained lots of useful information, I go slrap-up and this is now going to go

through everything, update the to-do list and run you through a particular process. So what this is going to do is

process. So what this is going to do is see if you have a brain notebook setup and if you don't it will prompt you to create one and effectively we're going to have Jack's brain that lives in

notebook lm and essentally every big session I have we're going to go ahead and we're going to store all the information over there. Now here's the cool thing. I have a Jack's brain

cool thing. I have a Jack's brain already. So, all this is going to do now

already. So, all this is going to do now is go through the entire session and just store all that stuff in long-term memory. Meaning that we can recall it

memory. Meaning that we can recall it whenever we're doing any key projects without having to rely on the kind of temporal nature of the memory within a cord session. Now, guys, check this out.

cord session. Now, guys, check this out.

I've now got Jack's AI brain. I click on it. Now, look how cool this is. This

it. Now, look how cool this is. This

report summarizes a strategic planning session. And look, session summary, the

session. And look, session summary, the 29th of March, 2026. I click on it and look, it's got everything on there. Bis

document provides blah blah blah what we did. and you go down and it summarizes

did. and you go down and it summarizes the entirety of Claude. Now, this is how you take it to an even further high level. I come back over to Claude. Now,

level. I come back over to Claude. Now,

anytime I want to think about a decision or anytime I'm using it, I can always say to Claude, "Hey," and I can even add it. Here's where it gets crazy in your

it. Here's where it gets crazy in your instructions on the right hand side. So

for example, if you are in a project like YouTube, for example, within cloud code work, you can add in instructions here and I can say to it, hey, whenever you for example, if I just have this up and I just do something like context,

okay, what that does is that's a feature I'm using within glider. Glider is what I've been using here, which is really cool. You can actually add these

cool. You can actually add these snippets in where you basically write down a word and it throws in there. But

one of the things I can add in addition to that, so I can come down and I can be like, whenever answering questions about strategy, always consult the jackbrain.mmd within notebook LM. Okay,

jackbrain.mmd within notebook LM. Okay,

beautiful. Done. Then I click save. Why

is that so freaking cool? Because now I do one simple API call. I get access to everything and it does a semantic search. Think about having Harry Potter

search. Think about having Harry Potter a million books or a million books.

Harry Potter Lord of the Rings. If I had to process all that information at the same time, guys, I'd be here a long time. Using it this way, it will find

time. Using it this way, it will find and draw back the very specific things we need. Meaning we get way better

we need. Meaning we get way better insights using this notebook LM strategy within Claude. It works in co-work. It

within Claude. It works in co-work. It

also works in code. it is universal and solves the biggest problem that we've got at the moment. And so hopefully you can see this potential. Now, if you want to get your 7-day robot, I've got a free trial in the group running full Sunday.

So, I'll put a link at the top of the description for you. But it does bring us on to a really important question, which is now we know how to leverage this beautiful memory system. How do we actually get the most out of co-working?

How can we use this project's feature to get light years ahead of everybody else?

Well, we're going to learn that by watching this video right