▶️YouTube

Why Is No One Talking About This INSANE NotebookLM Use Case?

Corey McClain·youtube.com·14 min read·Mar 28

Select text to create a clip

Transcript — click timestamp to jump, select text to highlight

163 segments

I'm about to show you the notebook LM use case that no one is talking about, but it absolutely

changes everything. And that is not an overstatement because in this video, I'm going

to be showing you step-by-step how to upload your entire chat history from chat GPT and

cloud AI into notebook LM. We're talking about three years of data, three years of conversations,

three years of memories, three years of frameworks, and where you can pull up memories that you

forgot, search for different items, and even start creating things and other assets from all of those

conversations instead of just letting it sit there dead and dormant. If you didn't know, you can request

a copy of your data from open AI and claw, and they will send it to you. But what a lot of people don't

know is those files are massive. They're too big to upload to notebook LM directly. And so we have to

break them up. Now, the first time I did this, I went a very complicated route and I didn't know

there was a much simpler and easier way, but I went ahead and took the time to work out all the kinks.

And now I have a very simple copy paste method that you're going to be able to use by the end of this

video to just copy and paste this one command and have your files broken up for you. And so the first

thing you want to do is go to my YouTube channel and find probably this latest video right here,

or even this video right here and click on it. Then come down to the description and click on this link

right here. That's going to take you to a Google form. After you fill out the Google form, you're

going to get a link to my Google drive where you're going to find these two documents. This is the

notebook LM compound intelligence prompt pad. And this notebook has everything you need to understand

about this video. It starts off with simple instructions for how to actually get your files

ready for notebook LM, but it also has some very interesting prompts that you confirm once you have

your chat GPT account uploaded to notebook LM. One quick note, don't try to copy the code from here,

because there's some hidden characters and it's not going to run correctly. That's why I've also

provided this second document. And all you have to do is come down to where it says one shot terminal

command highlight, and then stop right here where you see the P Y in the dotted lines. Now we want to

copy that. And just a quick note, I'm on a Mac. I'm not on windows. So if you're running on windows,

then I can't help you with support for that. If you're having problems finding terminal,

I would suggest looking on YouTube for videos for how to open terminal on windows or command prompt,

which I believe it may be named now. I'm not sure. And from that point forward, everything in this

video should work just the same. But if for some reason you paste this command and you run it,

and it doesn't do what it's supposed to do, then all you have to do is copy everything in the terminal,

both the command and the message that you received back paste it in the chat GPT and ask why it didn't

work. It works on Mac, but it's not working on my windows. And chat GPT is more than capable of

solving it for you in one to two turns. But now that we got that out of the way, the next thing you

want to do is head over to open AI. Then you want to click on your name in the bottom left hand corner.

You want to go to settings, then you want to go to data controls, and then you want to click on

export data at the very bottom, and then confirm export. Next, we want to head over to Claude,

click on our name in the bottom left hand corner, click on settings, go to privacy and then export

data. Claude gives you a little more control over what you export, but I recommend just

exporting everything and click export. This is what your email from Claude is going to look like.

All you have to do is click download data. And after 24 hours, this is what you're going to see

an expired link. And the same thing is true for open AI. You have 24 hours to actually download this data

because after that invalid signature or expire URL, I'm going to zoom in right here for you.

So once you get your data back, you're going to see a folder that looks like this, that says data

2026 0102 13. And if you come over here to the right, you can see that the conversations is 54.1

megabytes. That is too big to upload directly to notebook LM, but because of notebook LM file size limit,

this can be tricky. So with notebook LM, the file cannot be larger than 20 megabytes,

but it also can't be larger than 500,000 words, whichever comes first. What we're going to do is

use the terminal command to break up every file that is in this folder into something that is 450,000

words or less. But before you run the terminal command, this is very important. I want you to

right click on this file and then click rename it. And you're going to rename it to chat GPT

underscore data, no spaces, put it in the same caps because this is the way the command is set up

already. And if you just change the file name that you don't have to worry about changing the command,

which would be far more difficult. So just right click, change the file for the chat GPT data and name

it like this here. And don't worry, we're going to come back to this cloud data in just a minute.

I'm going to open up terminal on my Mac and I'm going to paste in the command and then I'm going

to press enter or return, wait for a few moments. And you can see that it says done input. This is

the downloads chat GPT data conversations.json. That's the file that came in output users,

Corey McLean downloads, and it created a folder notebook LM underscore ready. And now when we come

back to finder, here's the notebook LM ready folder. Let's double click and open it. And I want you to

look over here at the timestamps. It says today at 8 0 8 PM, I'm going to go up to the top of my Mac.

So you can see that it is Sunday, January 4th, 8 10 PM. So these were just created. And the second

thing I want you to look at is the file size. So everything is around 3.23 megabytes. And that's

how we know that these files have been chunked the proper way because 450,000 words should just be the

same file size across the board. Right? So all of them are pretty close to that. So we're good to

go there. The next thing I want you to look at is the kind of files they are. They are markdown text

files. That means that they are human readable. And so just to show you, I'm going to open number 24.

And so if I scroll through this, you can see that this is a full conversation taking place right here.

And so this is exactly what we want because not only is this machine readable, but this is also notebook

LM readable because one of the problems with the files you get back from open AI and probably even

claw is that there is just so much metadata, but not only does this Python script chunk everything

down to 450,000 words or less, it also strips out all of the metadata and throws it away so that the

only thing that is left are the labels between you and the chat bot going backwards and forward.

And so now this is the claw data right here. And now we want to batch this. And so all you want to

do is grab this folder, drop it inside of the chat GPT folder. I'm going to take this notebook

LM folder and move it to the trash. Then I'm going to take the chat GPT conversations and I'm going to

move that to the trash as well. Then I'm going to take my claw conversations and I'm going to drag it

out. I'm going back to terminal. I'm at the bottom. I'm going to paste the command again and run it one

more time. And if you look at the bottom users, Corey McLean downloads, chat GPT data,

conversations dot Jason, and then the output, the notebook LM ready folder. The main thing you

need to remember about this, whenever you're using this script to chunk any data is that you want to

move the data into that chat GPT folder and just make sure that the name is conversations dot Jason,

and it'll find that file and then chunk it. It'll parse it. Now we're back in my finder and you can see

the notebook LM ready folder right here. And so this time it only gave me one of them. So let's close this

up. I don't know why, but the one shot command that I had for chat GPT that was working for claw

yesterday did not work today during the recording of this video. And so I had to come to chat GPT

and troubleshoot it. And that's literally all you want to do whenever these things don't work.

And eventually chat GPT rewrote the command. So now I have one that works for both chat GPT

and claw. And what I did was I told chat GPT to give me a downloadable link right here,

the text file that I could upload to my Google drive. So when you download the notebook LM compound

intelligence prompt pack and you get the link to my Google drive, not only are you going to have

this PDF right here, you're also going to see one that says works for chat GPT. And then this one that

says works for claw and chat GPT. So if you use the first one and it doesn't work, then try the second

one, but everything is still the same. You're going to parse or chunk your chat GPT data. And then if you

have claw data, just move the chat GPT data out of the folder and then move the claw data in the

folder, make sure it's named conversations.json, run it again, and it's going to find it and parse it.

And this is the part we've all been waiting for. You want to come over to notebook LM and click on

create a new notebook, click choose file. Then you want to navigate to where your files are. So these

are the clawed files right here. I already have them uploaded as you probably noticed on my screen a

second ago. And now we're inside of my chat GPT notebook. And you can see the label right here

from 2022 to 2025. And if you look on the left-hand side, you can see that there are 82 different files

and I can click on either one of these and notebook LM is going to open it up. Now, once we start

scrolling down, you can see how detailed this file is by how little this margin right here actually moves.

Now that you uploaded your chat GPT and your claw data to notebook LM, we can do some very creative

things that we just haven't been able to do with notebook LM or chat GPT or claw for that matter.

Now I can ask notebook LM to act as a system analyst and go through three years of chat GPT conversations

and find the frameworks that I've used the most often, or the ones that I've had the most success

with, or that I lean on the most. Or if you want to be creative, instead of dropping this in the chat,

just take your prompts and automatically run them over here in the AI studio. So for instance,

you might go to reports, create your own paste your prompt, and then click generate glancing over

this. I can see how it's giving me a very high level view of everything that I've been using chat GPT

for. And I can see at least two frameworks that it's surface. One of them in particular that I didn't

get to finish actually developing that I actually want to come back to, but another one right here.

So what I want to do is copy those frameworks. And now I'm going to create a slide deck that

asked notebook LM to create a slide deck presentation to remind me about those frameworks,

because I don't remember what they were. I remember the five eyes vaguely, but I don't remember

rapid at all generating digital assets inside of notebook LM based on my three years conversation

history with chat GPT is just the beginning of what I'm able to do because I'm also able to come over

here to Gemini click on the plus button to add files. I can choose notebook LM and then I can

choose my chat GPT conversations and I can choose my clawed AI conversations. I can add them both to

the chat and then I can run the same prompt here in the chat. But instead of asking Gemini to create a

slide deck, I might say, well, you know what? I'm going to load nano banana pro create a,

a, I don't know, a punk rock synth wave, a notebook LM image that explains to me what these two frameworks

actually mean in as few words as possible. Paste in the frameworks, press enter. And now let's let

nano banana pro actually create an image that explains both of these frameworks using chat GPT and clawed.

If I consulted with clawed about it, which I probably did it. And just like that, we have a detailed

image. And so the five I system framework is about imagine, instruct, implement, inspect, inaugurate,

sealing the work, rapid system, rally, emotion, angle, controversy, production, originality,

interaction, velocity, delivery algorithm. I believe this is about content designed for rapid audience

growth. Facebook reels addresses. Ah, I remember this. I think it's coming to me now. I'm getting a

little bit more clear on this, but these are both some great frameworks that I never finished fully

developing. So if I wanted to go back to those, I can definitely pull this stuff up with notebook

LM and Gemini. And if we come back over here to notebook, we can see that there is a full report

that's been generated master frameworks and systems index. Suffice it to say that this is a very well

put together document. And if you look in this AI studio for chat GPT, you can see that there are

so many other assets that I've created to help me capitalize on my three years with chat GPT and make

2026 a better year, create a vision board image that captures what you believe my 2026 is going to look

like. Be brutally honest. If you think that based on my conversations, success lies in my future or

failure, just show me what you think is going to look like based on the trajectory of my conversations

and the things that have happened recently that you can ascertain brutally honest trajectory success is

possible. Failure learning retry. This was very modest future trajectory. So it didn't give me a

straightforward answer, which is cool. I just want you to see what's possible. So if you've made it

this far in the video and you want to upload your data from chat GPT or claw to notebook LM so that you

can start inspecting, researching and discovering a lot of new things about yourself, about the frameworks,

about solutions that you've discovered and forgot about, then make sure you go to the description

and click on the Google form link right here. And once you click on that link, you're going to see a

form like this, fill out this form. Just make sure it says notebook LM compound intelligence prompt pack

at the top. And you're good to go. Once you complete that form, you're going to see a thank you message

with the link to this Google drive folder. If you can't click on the link, copy it and paste it in the

browser. And you're going to see these three files. This is the compound intelligence prompt pack. It's going

to have 16 prompts inside that are going to give you some unique ways to start using this data. But again,

there are no right and wrong ways. So you can start creating any type of assets you want. Number two,

there are going to be two different one shot terminals. So if you only have chat GPT data, feel free to just

use the one that works for chat GPT. And if you also have claw data, make sure you use this one for the claw data.

After you move the claw files into the chat GPT folder, just like we discussed at the beginning of the video.

And one more thing, sometimes this stuff just doesn't work. And I can't tell you why. But one

thing I can tell you is if you paste your terminal messages inside of chat GPT, or if you take a

screenshot and drop it inside a chat GPT and just tell it in plain language, what you're trying to do,

just give it as much data as possible, tell it everything, and then ask it to fix the problem.

It's going to fix it. So if you got value out of this video, make sure you hit the like button.

And if you want to help the channel grow, hit the hype button as well and give it some hype points

and think about subscribing to the channel. And as always take care, have a good day. And if you

want to see some more interesting use cases about how I'm using my chat GPT data with Gemini,

then make sure you check out this video right here.

Clips

No clips yet

Select text above to create your first clip

Loading connections...

Reflect

What part of this is most worth remembering a month from now?

Press Cmd+Enter to submit