Claude just killed ALL Journaling Apps. Here is proof.

Here's the energy level and the mood.

It is a light mode, and I have even have a search field to search for things I can see the insights.

It even created an inside folder where you already can over time track your mood and energy.

And this is out of the box.

There's no application behind this.

It is just extracting what we will fill in from now onwards into this database.

If you're like me and you always struggle to make daily journaling, this video is for you because it will change everything for you into your personal life.

I have no doubt about this I never saw the reason to make real daily journaling, but what I have done is using the application Day One since 2018 because here is where I started the Paperless Movement, which you might know as myICOR.

And this YouTube channel talking about productivity tools, note taking, apps and so on.

is where I. Saved my conclusions.

so I started just writing down highlights and lowlights, but not daily journaling.

So here, for example, I started my first online store and this is in German now.

So, or here I was featured on no shelf, on their website or here.

When I started accelerating, I created my first merch.

Of the Paperless Movement.

That was my first merch that I did this.

See all the rem rememberings are here.

So here my whole journey about going through the Paperless Movement, growing the Paperless Movement, insights from the Paperless Movement, building a business, this is what I was journaling about, but only if there are really relevant insights.

Like here when I launched the very first forum, and in 2019 when I launched a membership and many things but the beautiful thing about this is that there was serendipity.

So if I go here on this day, you see there are five entries And these entries are from different journals that I have on the side.

And here is the new YouTube strategy that I launched, in 2024.

Here is my woodwork that I started and I built my own table.

And, this is eight years ago when I brought my child for the first time to the kindergarten and, and many things like this that I laughed.

To see on a daily basis what happened all the years ago.

And this is where I kept using day one.

And I kept it very clean because I didn't want to get random information about what I ate that day or how I was feeling that day.

So I always only saw milestones and highlights that are really relevant for me to see over time when I was, for example, celebrating the first thousand subscribers on this YouTube channel.

And I see this every year I stay humble when we are now reaching the a hundred thousand subscribers after all this time and work that I put into this channel.

So you see, I used day one for such a long time, let's say since 2018 until now, which includes.

Eight years of information that I shared there that are very personal insights, but they are still hidden in there.

I wasn't growing based on this.

I got this reminders and I saw . The serendipity, , but I wasn't concluding anything out of this journey, and this is what AI changed and this is why along with Day One, I started using Mindsera, which is an AI based journaling application where you can share your thoughts and AI will give you feedback.

And already, back in 2025 with Claude Code, I built my own AI agent I was able to talk to based on stoicism.

It was giving me recommendations and all these things I'm not even sharing here, but it's all possible . And besides Minder, I also used Stoic.

It is another journaling app that is a combination.

it looks more like the day one, but with AI integrations, well, I tested them all.

And the thing is, your knowledge stays stuck in there and you cannot have the context with everything around you going on.

And this is what Claude Code really changed in this game.

now.

In this video, I want to show you how easy it became to do journaling, why it's worth to now daily journaling, and how you can all build this just on a local folder on your computer.

So let's dive right into this, how we built this from scratch.

Now if we go into my own assistant here and I open up my, my personal.

Version that I have here.

I have a journaling app here.

It's not an app, it's just visualizing the database.

And that's what we will build today.

Something similar.

So here I'm able to simply share a screenshot and talk to AI, and it will extract everything, the mood I had when I was doing this, the energy levels that I had.

And then overall, it gives me even then a recommendation based on socialism and so on.

Moving forward, this is an absolute game changer and whatever comes to mind or insights I get, I cannot share here and have this organized and cross connected with all my other thoughts that I had and people that I'm interacting with that I would have never received in any other applications that I paid money for.

So.

Let's dive into how we build this from scratch.

In a previous video I showed you already how we build from scratch this PK demo folder here where we build from the ground up, how to have an owner's inbox that gets processed, how to have a basic team of AI agents working for you.

All the things you might have seen the video already.

If not, it's worth going back there after this video to get more details about how the basics work.

Um, if we open this up, you see it's again just a simple website that opens up and looks and shares.

And this is what we build in this other video.

And there was already a journal in there where we had a test with the first journal entry.

We had already context contact.

We had already prefilled it with some contacts and that we had there.

And here's Paco that we cont added here and so on.

This is what we built in the past video and I could now build on top of this.

However, in this video, I want to start from scratch with focus just on this daily journal, so we keep things clean.

And in the end of the video I show you how you can easily do things like this to have a sandbox mode to play around with things like this and merge it back into your actual team that you already built previously.

So without further you, let's dive into this.

We just create a new folder here on desktop called Daily Journaling.

And as you can see, we open it up.

It's an empty folder right here.

We could now just search for the terminal.

And start Claude.

And I see already people who didn't watch the previous video say, oh my God, he's using Claude code and now I have to learn code and all this things to make this journal work.

Bear with me.

Watch the video until the end and you will see no code involved.

This is the amazing thing about this and what I'm sharing here.

you can also do on the desktop version, go to Claude Cowork or even Better code, and just find this empty folder that we just created here.

If you want to follow along and you don't want to use the terminal, you can give it a go there too.

So you just open this folder daily journaling in here and this is it.

And you can do the same things that I'm doing now using the terminal.

I just prefer this way of working.

I feel it's more reliable and more efficient using the terminal.

And in fact, I showed you another video that I'm not even only using the terminal, but I'm using VS code, which again implies that there is code involved, which I also demystified another video that's not true.

You just open this folder and I will link all the videos to these things in the description below.

But all I need to do here is open up this new folder on daily journaling in via code, and now you see it's just here representing the folders.

you see there is now the empty folder.

We have Claude code here, which is the extension.

We can stay in this version or we launch a terminal.

And that's what I will do.

I close this extension, which has a nice interface.

you know, that's very personal preferences.

I just like to work in the terminal For now, I go on top, I click on new terminal, or you use the shortcut, it opens up the terminal.

I can write down Claude.

And in fact, I will will Claude dangerously skip permissions.

And there's also a new auto mode that is similar to this one.

And here if you are unsure and you are scared, what it will do with your system, don't do this right?

But I'm confident I'm working in this one folder here and I just don't like to accept all the basic.

Access that is asking for all the time, and that's why I want to use this version.

I see we are already running out of of context, so I probably go over for this video, I've done a lot of coding work today and a lot of things that very token intensive, but what we are building here today is nothing complex at all and you're not using up all your tokens at all.

Okay, so I will just move this box on here.

As I showed you in the other video, again, watch the video after this one if you want to get started here.

It's all about building this journaling application.

So now what we see, we have here Claude code open, we have here our text entry, and then here we have our empty folder.

And you just start talking to Claude.

In fact, in Claude code, you can just press and hold the space bar to talk to Claude.

hey, I just wanted to work on something.

Can you help me But I think it's not reliable enough and what I'm using to talk to Claude code is actually Wispr flow.

I made a dedicated video about Wispr flow already, how I talk to ai.

If you want to watch this, it's on the channel two, so I have a shortcut and I can start talking.

And the good thing about whisper flow is that all the things that I say.

Is remains accessible.

Okay.

and you have a dictionary where you can share the words that you're constantly using.

It even updates itself.

And many things that I can do with whisper flow that the normal voice input from Claude Code is not capable of and usually messes things up.

And even if you have things that you correct yourself along the way, or bullet lists, whisper flow formats, and reorganizes the things along the way, that's just as a side note.

So you're not surprised when I keep talking to Claude and suddenly there's text appearing.

That's what I'm using.

Alright.

We are here in an empty folder.

And what I want to have, I want to build a daily journaling database.

It's important that we are not saying something like a daily journaling application because then Claude will start building an actual application with code and all the things.

That's all overkill.

Something we don't need.

All we need is a just say a local database and Claude will , very likely create an SQlite, that's what it's called, database there.

Just describe what you need and what AI should avoid and you get there Anyway, so that's what I will keep doing here.

Let's not overkill the things, the moment you get started with individual features like daily journaling here, because you can always add later on.

But it's really harder to remove the features once you let them out.

All the things that you want, all the base bells and whistles.

It also makes it harder for AI to get the full scope to really understand how everything works together.

So build something very basic, start using it, and then you can give feedback how it should get improved.

And that's what's called , continuous improvement.

And that's what we will do here too.

I want to have something lightweight on a local folder.

No complex application or anything like this.

It should be simply another base with a table where you track my thoughts that I share with you.

And extract my mood and my energy levels that you feel that my shared information contains.

What I also want is that I'm able to share images with you that I get also stored in this database.

Okay.

See, this is, uh, what I have said here.

That's the prompt.

If you want to copy paste it, feel free.

But again, you see, there's no prompt engineering that I'm doing or anything like this.

Claude became really smart enough to really get the point what you're going for, and you can always revert things back and adjust things along the way.

If you see the outcome is not proper, and that's why I love working in VS code because I always have the connection of my input and the output appearing inside this folder.

Again, keep in mind this is just a local folder on our desktop called daily journaling.

So let's see what happens.

So I don't want to directly let them create everything, and that's the big difference in Claude code that you can switch to plan mode that will switch the.

Approach of Claude to planning things really out.

So you can review first what it's about to do, and then you can confirm or readjust along the way.

It also allows Claude to ask you any clarifying questions along the way before it starts implementing, and that's why I always recommend for when you get started.

First start with the plan mode.

How I switched it says this here.

Shift tab to cycle.

That's where you can switch between the different modes and let's see what planet will come up with.

Now you see already, I didn't mention SQ light at all, but it says already a local SQ light database is the perfect fit here.

That's what I told you.

See, I didn't mention SQ light at all in my prompt, but it probably will always get back to SQ light for local database.

So even without understanding what SQI database means, you can be sure it will pick it up and it's the standard for these things.

And here's the plan.

Local daily journaling database context.

What I want the table it will create in this database.

So the database will be called journal db.

the table with the entry.

You see here all the context, the mood, the energy score, the summary text, even the edit.

Uh, now I have another table for the images that I mentioned.

That's the file structure here and then how it works.

Thomas.

So you can really step by step go through this and understand what is going on and what it is about to implement.

Now you could just say, yes, bypass permission, or you go down here and say anything that it should update or that was misunderstood.

And that's really the crucial step here because this is the perfect moment where you can make clear what you actually want from ai.

But I think it made an amazing chop here out of the box.

And we will just start implementing this.

So now you see it will start cleaning the context so it has a full context window available and it will use this plan file to implement this now.

And then we see already here the first folder appeared.

Here we have the journal DB appearing.

So you see here will be our images that are stored and it creates now this database.

And if we go to the desktop.

Here, that's the folder on desktop you see in parallel.

It just creates the same thing.

So if people are always confused when they hear about Claude code or VS code and it looks so technical and for developers only, once you understand that this is just a more raw version of what is fancy in the front end, it makes it so much easier to use and a lot more powerful to control your AI and know where you

are working in and many other things that we talk on this channel, but also in our membership in much more detail And inside our courses and discussions.

So you see it also created already the Code MD file.

If you don't know yet what the MD means, it is just a text file that we can open and read again.

See, it's just the first file that AI will pick up and see the instructions.

So in this folder we are building a daily journaling database.

So if we start talking to AI from within this folder, it will first load in this to understand what the AI should impersonate.

So in this case, it is my personal journaling assistant, if you will.

And then, it is getting aware that there is a database with these tables and all these things, and these are the coming queries that come in and so on.

So this is something I wouldn't even open because.

You don't need to, but if you want to get the transparency what's going on behind the scenes with ai, you can still read this in plain text with natural language for you.

Easy to understand.

And if there is anything that confuses you, you just copy paste this into Claude and ask about it.

I have no clue what this means, or can you give me any additional descriptions inside the Claude MD file as comments to explain more the different sections of these things.

What I just said here, you can just copy in to use as a prompt to ask that.

That's really the beauty about working with ai, that whenever you hit a wall, just ask it how you can work around it.

And if you are worried about what is going on, you can let AI always double check online if things are working properly.

So now the, journaling database is set up.

Here's what's in the folder.

And that's exactly what we have.

And you see how lightweight this is.

So he is now ready whenever you, I want to start journaling.

So let's start, but I want to start on a fresh mind.

Okay.

So we know that AI doesn't know all about this, so I show you that it's actually working.

What we can do is slash clear.

This will just clear the conversation and it starts from complete scratch.

And now we can just start talking.

I could just close the terminal and restart the thing.

It would be the same thing, and you will see what happens now.

So let's start with the first journal entry.

I just had a meeting today with Paco, our co-founder, and it was amazing because we talked about how we can further double down on AI productivity for our members inside myICOR.

And we concluded that the personal area of productivity is really key here, and that's where we will double down more and more.

And that's why I decided today to make a video about daily journaling and the advantages people have using it.

Quite a long thing.

But, uh, we will use this entry very soon.

So I just shared something without even mentioning that this is a journal entry and now what it will do, it will go to Claude at MD before even answering and thinking about this to load in what this is all about.

And now it has the right context for what I share here.

So let's see what happens.

There we go.

See, sounds like an energizing day.

Here's what I extracted.

Boom boom boom.

And then I could keep talking to it.

So now it started tagging even this entry.

And if we go here, there's no image, there's nothing changed.

We have this DB that we cannot even open and look at.

So how do we access now this information, where does it live?

Well, it lives inside this database file and there are, could be tools that you can download to visualize this database file.

But the thing is, I don't want to have any additional tools around this.

So I will just say, how can I visualize these journals?

Thanks for the entry.

Uh, the thing is I cannot see these entries yet, so can you build me a very simple interface that I can open on my Chrome browser?

To actually see these journal entries and the metadata insights that you've created for me, and also later on the images that I will share with you in a visually pleasing way.

and let's think about some colors.

I usually try love dark mode, colors and so on, but I know there are a lot of people as like light mode colors.

So let's test this.

For this video, I'd like to have a light mode interface that is really visually leasing represents a calming way.

Also, when you choose the font styles and so on.

So I really enjoy reading these journals.

See, I'm just sharing feelings here.

And maybe somebody would call me out on that and say, well, you're not doing proper prompt engineering and things like that.

Well, I've built a whole myICOR application, which is our membership with the courses, with community interactions with many things in there, only with Claude code in the way that I'm telling you how I interact with ai and that's why you can call me out, whatever.

I have proof that the way I'm doing it is working and it is much simpler than so many people out there suggest.

So again, I will switch to plan mode for this because we are planning out a new feature and I want to see.

How he will come up with this while he's thinking about this.

Keep in mind in the other video I showed you how to build an AI team.

So if you have an AI team, this becomes even more efficient because in an AI team, you would hire somebody who is just there to do the journaling with you.

So in my personal setup, I have Seneca, which is a stoic mentor that picks up my daily journaling and gives me feedback on this there.

So this is a specific AI team member, but I also have an AI developer who is just there to build out new features and things in my interface.

And as you can see in this version here, it will just create a simple HTML file that we can double click and open our browser to see everything that we have in our database.

Again, I'm just saying don't start with an overkill and create a team and things like that.

Your individual AI assistant here can go a long way.

All you need to understand is that this assistant here is always loading this basic Claude Code MD file, and therefore, if you go now beyond daily journaling and you start task management and many other things that you could do, that is where things become confusing for AI because everything gets stored inside the Code MD file, and that is the moment where you need to think about a way

to build an AI team and things like that based on simple folder structure.

What I'm using here is probably overkill.

The Opus 4.6 with 1 million context window, it's very expensive running through your tokens very fast.

So if you want to start and you have no idea yet where it's going, maybe you switch to sonnet, the model, which you just do by slash model, and then you can switch to models and start working this way.

it will be also a lot faster than here waiting for Opus because there is no 1 million context window needed yet.

But later on, if you have like me backfilling, my journalings from 8 years of journaling into another base and things like this, this is where this becomes really powerful because it gets the full picture of all the things that you.

Created and stored over all those years.

Think about it.

If you are using Evernote or anything like this where you start all this knowledge, this is the moment to extract it from these tools and build your own PKA system, which is Personal Knowledge Assistance we no longer call it PKM, personal Knowledge management because here I hand over the management to my assistant doing the linking, the cross-linking, and all the things that you will see in the end

of the video, how this works in detail.

Okay?

Now he's suggesting that he wants to create a Python, a architecture plus a single HTML file to open up this thing.

It wants really to make an app out of it instead of, uh, just a simple HTML file I would force now for these videos to show you really how simple it is.

So if you think this is overkill and I have no clue what's happening, and I wanted to have it simpler, ask, it let's do this.

I don't like an application.

Please just give me a HTML file that shows me the content of the journal database.

So this is the way where we just want to see now the database content and not having a proper application.

So later on.

We can always come back to this and expand the things when we need any additional fancy features in there.

But right now it's more than enough to just have a fewer HTML.

That helps me to visualize what is inside the database.

see, got it.

No server.

Just plain HTML file.

And that's it.

This being said, I will make more videos because this no server and so on a like code, it scares people away, but it is so much easier than you think and you can build your own applications in no time.

And in fact, it will become the future.

Without you knowing what's happening under the hood, you will be able to create any tools that you want and I will show you a lot more how you can build simple tools and apps using AI for your specific needs.

And that's the key here.

Journal view implementation plan.

See now it refined the plan.

We just have an HTML.

It's easy to understand the plan.

Here's what's going, here's the colors it's using, that's the layout and that's it.

Boom.

Very lightweight.

So whenever you feel it's too complicated and you have no idea, just get back to AI and say, is there no easier way to make this?

And then you will do it.

Uh, you will find an easier way for sure.

Okay, so now it's executing the plan.

So it inserts inserts today's journal entry.

It builds the fewer HTML update Code of md so it understands now that there is a fewer available as well.

So it will just adhere the information that there is a fewer to show the content and verify fewer in Chrome.

That's what we will do after it finished.

And here we go.

It already opened this up.

And look how beautiful this looks.

You see this is running locally.

There's no, that's not like that.

You need to be afraid that this is now online and other people can see it.

No, it is on your local file order.

You see it Users, Thomas Ruddle, my desktop.

And here is the daily journaling folder.

And this is the fewer HTML.

If we open up the folder again, desktop daily journaling, fewer HTML.

So if I double click this, it opens up the exact same thing.

Okay?

That's the amazing thing.

And now look at this.

I have here the first insight that I just said.

You see, it took what I said it under threat my I Corps.

Instead of my eye core, it extracted a summary.

Here are the text.

Here's the energy level and the mood.

It is a light mode, and I have even have a search field to search for things obviously it's nothing to search here because it will always show this, but, uh, I can, you know, check this mood and filter by it.

I can see the insights.

It even created an inside folder where you already can over time track your mood and energy.

And this is out of the box.

It is a simple HTML file.

There's no application behind this.

It is just extracting what we will fill in from now onwards into this database.

So if you see something like this where the My voice translation was wrong, you can now always go back and say, great.

Start.

But make sure to fix any spelling issues.

One example is my I Corp, which should be myICOR.

And the rest I'm saying and sure to make glossary where you start tracking any specific names or words that are uncommon to you.

Add this to the database too as a separate table.

See, this is how I start building additional features on this.

I recognized an issue.

And now I build a solution for it right on the go.

And what it will do, it will create a new table in the journal.

while he's working, I can already talk to it.

I can say Update.

Also the interface, to show also the glossary.

So we have a individual tab where we can see all the glossary entries.

And then later on I could say, I want to use this interface to add any glossary entries and things like this.

And this is where it might become more complex.

If you want to enter things into the database, you might need to go beyond just visualizing it in a simple H but you can really go a long way with the simple one file database view.

So now it created the glossary.

And the nice thing is if I go here, you see the updates happening live.

So yeah, I just needed to reload and here's the glossary and here's the terms that will be retract.

And I guess he will now be adding my I core.

There we go.

See it updated.

Now they're embedding.

This looks very complex, but this is just what's happening behind the scenes inside your journal thing here, in this fewer HTML And there's no magic behind this, even that this is all code.

Again, as I showed you in the other video, if you use Claude desktop code work and so on, that's what you see there too.

They're just hiding it from you so you're not getting confused.

But in the end, behind the scenes, everything works exactly the same.

And to me, this is just much more transparent to see what's going on.

And now if you go to glossary, boom, you see Paco was corrected too, in addition to myICOR.

I didn't even mention it, but it recognized it.

So the next time he knows when I talk Paco, this is it, my ico, and have this glossary.

Now, isn't this amazing?

And here are the entries.

Let's see.

Boom.

It's fixed already.

And this is my first entry here.

now I want to get one step further and then I let you do the future work and you share whatever you've built in the comments below, because it's always amazing what people individually need.

So there is no one template to use for everything.

You are now empowered to build whatever you need for yourself using ai.

See, it tells me then, even if you know I'm getting confused with all the code and so on.

No, it sums it up.

And if it doesn't, you ask, what did you just change?

Explain it to me.

I have no clue what code is.

And this is what's happening here, by the way, as a small trick here, you can also go to config, and in here you can set up how you Code will work, Code code, and there you can actually go to.

Output style and if you hit space, you can now change this from default to explanatory.

So it will explain you a lot more what's doing or even learning.

So it will actually ask you a lot things that you can actually learn what it is doing, like coding and so on.

I keep it at default, but if you are curious what is actually happening, this is the way to go where you set this to explanatory and it will tell you a lot more details about why it's doing what it's doing.

That might be a good way to start the things.

Okay great.

This is working perfectly.

Uh, what I would like to have next is also a simple CRMA table for my contacts, where you keep track of whenever I mention anything about this person to update their table, like whenever I mention a birthday or a phone number or their address or anything like this that, you know, where you should store this.

So we have always cross connected links.

And then I would also like that these things get actually clickable in my interface.

So whenever a person like Pao has mentioned, I could click it and I open his card with all the details about him, but also all the different daily journal entries.

That mentioned him should appear on this card too.

So I can actually go through them and click again on this mention to open up this journal entry to read more details about why I mentioned them him in this details.

also create a summary on this card.

So whenever I mention this person, it'll automatically also update the summary and how I'm connected to this person and so on, on the fly.

So each time I create a journal entry with a person mentioned, it should double check if it's worth updating the summary.

Yeah, I think this is, uh, about it for now.

You see now I again go into plan mode for this and let him plan out this, but now you see what I'm going for.

that's now the difference between PKM and PKA in personal knowledge management.

If you are on Obsidian Notion, Heptabase, you name it, we all learned over the past years how amazing it is to cross connect information.

In fact, that's something we taught over the years too, because you get a lot more insights if you do it, but it's always a lot of friction.

You need to think about it.

You, you don't know how it will be linked in the future because something else comes up, you forgot about it.

That's where a dream was always this auto connecting but here we can now have it perfectly set up.

In a way that we think it makes sense because now you can imagine here I'm just showing a A people database, right?

But you could say next also create A company database and now connect the people with the company whenever I mentioned them.

And then also you have to cross connections this way.

And AI will keep doing the connections just out of whatever you shared with your natural language, even speaking in whatever you had to share to ai.

if this is not amazing, I don't know what is, because I can tell you, since Paco and I both individually switched, creating our own daily journaling setups like this, the insights we retrieved are way beyond whatever we had With our already highly refined productivity systems. We had before that AI is amplifying any system that you already had in place.

And now Paco and I, we feel it.

We had a solid system.

Running the foundation is right and putting AI on top is just accelerating everything, giving you more insights than ever.

And this is the crazy thing.

what people miss out, they hear about ai, they start using the chat bot and so on, but They don't have no system in place.

And that's why AI keeps failing as if you are confused why you need a system and you're not organized in your daily work and things like that.

AI will be too.

you could say, how can I become more productive?

AI will give you a thousand answers, what to do, and you might have a great conversation with AI and share some hacks, but the real beauty is when you have a proper system in place that is already very efficient and you just replace individual parts of the system with AI that processes things for you, that gave you a lot of friction in the past, this is the moment where you feel

that AI is already changing the way we work.

And anybody saying out there, oh, you have time.

It's not there yet.

, we can stay calm.

And it's just hype, it's nonsense.

I can't just say it as it is.

You can believe this, you can, stay in this bubble and say everything will be fine.

And then be surprised that everybody else is ahead of you, or you just start tinkering around with these things.

Start with the basics, play around with it, and you will see with just simple things like this, how you will improve your overall work.

And this is what the future will be in businesses too, because it's not the business owner putting AI on top and forcing everybody to use this ai.

Now it will be the individual team member, the individual employee, and the business owner himself who will have their own AI teams running for them, their daily journalists, connecting information across the company if it is allowed, and all these things that will make the overall company more productive and not.

The company forcing people into ai because as we see, nobody has a clue about ai.

Usually business owner doesn't know why we, they need two ai and it will end up like everybody's using ai.

So we should use ai.

Uh, here we are and you have it, and we have no clue what it is.

And this is not an AI problem.

This is a problem, especially in incorporate in bigger companies where they feel forced to do something like implementing agile thinking and many things like this without understanding why they actually roll out all these things and therefore falling short.

People start using buzzwords, but they have no clue why this is in place, and this gives frustration and confusion, and that's why the future lays in educating your people in personal productivity using AI instead of forcing them to use something that you probably don't understand yourself.

to keep in mind.

Obviously, if we force everything in just one simple HTML file, this will bloat up things very quickly So we can actually open this up here by clicking here.

So you see the content, what's inside the file, and you see there's already a lot of lines of code in there that is just for visualizing the database.

So eventually, if you add more and more complex features like this, you might add up to build a very lightweight, simple application.

Again, you will be surprised that there is no code involved and you can go a long way with it.

Just keep in mind when you do things like this, as long you do these things locally, no issue.

But if you start sharing things and so on, then think twice.

Don't sell these things if you don't know about how to set up security guardrails, things like that, that you are, passwords don't get exposed and things like this.

That's really important, but as long you have it here, store it on Dropbox folder.

Locally and therefore it's synced on on Google Drive or whatever.

Or on iCloud.

You have it backed up.

You have your journal app there just in a folder.

Very lightweight.

Okay, it updated it.

Let's see what it did.

Here we go.

Now we have a new thing, contacts.

Boom.

There is Paco ero I can click on and boom.

Amazing.

I see here.

Paco, our co-founder here, we see there's no information about him.

But he is working a company and is a co-founder.

And here's the entry.

And if I click here we go.

It scrolls to the entry guys.

And we built this in just one short video that I'm sharing with here.

So I hope you can imagine how far we can get.

So you can see now this is clickable pa inside, boom.

I can open this up and in another video I can even show you how you can build a knowledge graph like you have in obsidian, where everything is just visually representing how things are interconnected.

You can do all this yourself.

You don't need obsidian.

That actually changes the files in a way that you cannot use this elsewhere.

Here you're building everything for yourself you see just in a matter of a few prompts without any background knowledge necessary.

And we have all this already working.

Let's do one more journal entry and see how things work.

I clear again the knowledge and now I will start saying, three days ago I talked with Paco about how we can best update the user experience inside our my co application to make engagement even better.

We had great conversations about this and meanwhile we even implemented all this already, but please.

At this daily journal entry, retrospectively and on the same day.

Later on, I went shopping with my wife, and as you don't know her yet, my wife is Carolyn, and that's it.

I think you get the point, and let's see what he will do now with this entry, how it gets cross connected, how he creates now the new CRM entries.

Let's see.

So what he's doing again, he goes now to Claude MD that is now updated.

You see, that's the beauty of VS. Code.

I can go to any file and see its content.

And here we see all the content, image storage, all these things.

Oh yeah.

Image storage.

Something we didn't test yet and I want to show you for sure.

Okay, here it comes back with a question.

My echo.

Okay.

So there was a misinterpretation of what I was saying.

Again, but see it recognized it.

So I can say it is a misspelling.

It was my I core and my personal assistant.

Actually, for my personal journaling, I'm using German language and you can use whatever language you use.

Obviously my English is not perfect and it might miss bread things a lot quicker than others native speakers do, but perfect.

He recognized that he double checks with me if it was correctly understood, and now it makes the correct entries.

Also great for this video is that we already here in using extra usage.

So let's look at this point of the video.

What we are at.

So we see the current session is right now 800%.

It will reset in 17 minutes, and we are currently at 22 Euros, which is already from previous.

Uh.

Sessions where I went over.

So let's just keep this in mind.

22 euros, and then we will check later on when it resets how much we spend while we've been working on this.

Okay, it finished the two entries.

Let's see what happened.

Boom, here we go.

Two new entries late in the day when shopping with my wife Caroline, I can click it.

Created already these things, it made a note.

Thomas's wife and here has the entry and now I can keep expanding on it.

And the same is here Now talked with Paco.

I feel out my, now we see the two entries appearing here and I say, oh, Paco, well here and well, I already.

Thought about PA and in another context and so on.

And this is how you start cross connecting all this.

Isn't this amazing?

So the last thing I want to test here is sharing screenshots and then add this to my daily chart, how this already works.

I, so I have this screenshot and it's just in my, uh, clipboard, right?

And I will just paste this into the chat where we celebrate now that we have 3000 members already inside my Icorp, this is amazing.

Thanks so much guys, for joining our membership.

We are growing rapidly right now, and this is what we paced in.

You see it says image.

That's all I need.

so I just copy pasted in with Command V from my clipboard.

And now I just give more context.

Today we are celebrating 3000 members inside my eye core.

This is amazing.

And we are so happy that what we teach is helping so many members already, and we are obviously hoping that we, we get a lot more members in because engaging with these people inside the membership is just amazing seeing what people are building.

This is amazing.

And being on this journey together towards using AI for productivity.

Is something that PA and I both deeply enjoy.

Okay, there we go.

So now we have the image in front and I have the text and let's see what it will do.

What a milestone.

And it gets the screenshot, it found the count of members.

And this is the amazing thing.

It can read images, it can extract data from these images and makes the thing.

So now it asks me if this looks all right, if I'm getting more and more confident, I might even say, do whatever you want.

Don't get back to me, and so on.

So, I just say, yes in this case, and this is what it'll get created Now.

Now we see images and here's the image.

Boom.

It is locally stored again, you can always right click here and say Reveal and finder, and it opens up the folder.

Again, desktop daily journaling, images.

Here's your image.

It's a p and g file.

It's just a local file, but with the power that this will get now connected inside your database with the context and it will be visualized in a different way.

And this is the powerful thing that even if you're not happy with this interface anymore, you could just build something from scratch.

Nothing gets lost, all your data remains active, and this is makes you independent from any tool switching, losing your information along the way.

So the powerful thing is if I would be on the go and I want to talk with my daily journal, I can always go here remote slash remote and go to remote control and it will create and launch a remote control.

And now you can just use the Claude.

mobile application to access exactly this chat.

So I could now just go here and write something and it will just appear right in here, is if, if I would be sitting on my computer writing with it.

And this is where you can be on the go and simply just keep sharing images talking and it will create these entries on the go.

Again, no other tool needed.

You just use Claude for this.

Boom.

It created the entry.

Let's see what happened on the front end.

Here we go.

And there we see, uh, that it's not showing the image properly.

It extracted everything from the image it mentioned.

Paco.

See it works.

I can click and I see now three entries for Paco.

but for this, again, it's not working.

So what's the thing?

Well, many things that you can do, you can just talk, but what I prefer is that I just make a screenshot from this situation here, how this looks, and share it back with AI like this and say, as you can see, the image is not showing properly on the interface that you've built.

Can you fix this, please?

I have some ideas why this is, but most people don't, and that's the approach.

You would do it.

I could also say, instead of sharing the image, I could say, go to Chrome and check for yourself until it's working.

So it would go to Chrome using the Chrome extension as I showed you in another video on this channel.

And make screenshots for himself to understand if it's working and it keeps improving things until it's working here.

We might have a back and forth where he tries things, but I guess he get it right now.

So if I reload it, ask me to allow, and here we go.

There's the image.

I can open it up.

See full screen image.

Isn't this amazing?

Now we have the daily journal entry with an image available, and this is the moment where I will stop this implementation and we'll now talk about how could we now merge this back into our existing AI team that we already created earlier.

So obviously if you just watched this video and you created this, you could go from here.

Or if you now watched the previous video and you created the AI team, this is how you can bring things together now.

So keep in mind, we are here in the daily journaling folder.

Let's move to another folder.

So I just say a new window and I will open up in this window.

Now, the PKA demo that we created earlier, as you see with these things.

So you see now on the side, just the content again.

So we have both folders open independently, and if I launch now an A clock and here we are now working inside this folder and not in this folder.

Okay, so what I want to do now, as we had here already a journal created.

So just as a reminder, if we open up now this file inside Chrome, this is what we built previously.

And here was already a journal with an entry.

So now we have two different databases, but we like this one much more.

I, I actually do like this much more what we built here.

and now is the moment where you really see how AI can shine with this.

So I'm here and I can now just say, let's go to this folder that we just created daily journaling, right click and I hold on option to get the path name.

I could also just drag and drop this in here and hold shift so you get the link to it.

So there are two ways to do this and I say we've just created this new journaling approach in another folder that I really like, and I would like you to implement this in our.

PKA demo folder.

We already have some journaling basics there, and we already have created an interface, but please look at this new version, that I like much more.

Also, the looks of it.

I want to have this upgraded in the overall interface, going from this brownish color to the light mode version but also make sure that you compare the different databases.

Make sure to migrate the content that we already created in the previous folder, and update our current folder with it and understand how things are interconnected, how things are working in this new version, and carefully update our current version here with this new approach.

So we can end up deleting the other folder and just keep using the journaling approach in this folder.

As you see, there was a lot of rambling that I did here.

You can also write this down to be sure you're doing it right.

Uh, for the purpose of sharing.

I do it this way now and again, I switched to plan mode for this.

He will load in his Claude MD file that we created in a previous video.

And let's see what happens.

Again, in plan mode, nothing will crack break.

This is exactly the approach.

And then you can still decide if you will do changes or if you feel that there's, uh, more things that you need to look into.

But what it will do, it might also use our team.

Okay, so we have team members.

Sable is the one that we hired in the past video, who is our developer.

So he should now hand this off to Sable in order to work on this.

So first he's exploring everything we could have said.

The team should work on this.

That usually triggers that he really is using the team.

We can reinforce and guardrail that he's not allowed to use this, but he needs to hand this off.

We could have said Nolan, the HR person in the team should hire another person who is, uh, focused on.

Databases and understands migrations and things like that.

That's what I would do in this too.

So he actually has a specific AI team member who is focused on just this type of work and you will get even better outcomes.

But I'm pretty sure as long the things are so small as they are here, it will do a pretty good job to make the migration work without any extra, AI member hired.

Just keep this in mind.

If you have more complex things, this would be the way I would do it.

I would hire a specific person who is specialized in the work that needs to get done because this will upfront do a lot more research about how things should get done, and then you have a proper team member that you always can address the work to.

While this is working on it, I will open up again the usage.

To see what's happening.

So we had $22 euros.

If I reload, we see now the updated costs.

Okay?

All the things that we did so far up to this point was just one euro.

So you get just an idea, of the costs, and I think this is okay, especially as this is now resetting in three minutes.

I'm no longer spending any extra usage and I'm back in my normal $200 plan.

And the same applies for the pro plan or the $100 max plan.

Okay.

And that's the beautiful thing.

While he is thinking, it shows you the insights already, it compared already the new daily journaling versus the current implementation of this.

And now he creates the migration plan and that's where if you see any templates or anything that you can download, that might be a good way to do this too.

Let your existing AI agent review what else there is and see how it can get incorporated in your existing setup rather than just switching to something random that another person created.

Now we see, just switched back to the normal plan without anything me doing while it is working, there's no interruption or whatever.

So it just keeps working on the same thing And here we go.

There's the plan that he created.

Again, by the way, you could say now, Hey, can you give me this plan in a PDF format and in a nice formatted way so it can come, you read through it, all the things are possible if this is too confusing to you.

but here, this is what you will do.

Database migration.

It will migrate our other database into this current one.

It will even make a backup first of the one that you have in place.

See, amazing approach.

that's really proactivity that I really love about Claude here.

And it adds to missing column, see, like birthday address summary that we haven't had in our current PKA system yet.

Create new tables, the entry context, the images, and the glossary.

See that's something we haven't had.

So it recognized all the things.

It imports the four journals that we made in the new version, into this current version.

It, shows me exactly what it will work on the different files doing exports, and then it updates the viewer with our new approach of the visualization.

And then it does the data injections of all the things.

So nothing we need to worry about.

I will just let it work.

And you see this is working now, so you see it's delegating to Sable, our team member here where he delegates the update of all the things now.

So Larry will delegate a full implementation to S with this plan.

So you see the thing that we built in the previous video is working, right?

Larry is recognizing that he needs to delegate this over to Sable.

And S is now the guy working on it.

He's saying, plan approved time to get to work.

I'm sending this to Sable.

She built the current PK viewer and knows the architecture inside out.

blah blah blah blah blah.

Okay?

And now you see Larry launches a subagent, and this subagent is called sable.

So it is really recognizing even that we are not using proper agents here.

That's something we just have in the backend.

We have here team roaster and if you watch the previous video, we built this from scratch and it's just all based on normal file.

So if you open up Sable here, we see what she can do, and how she approaches things and where to find additional information.

And then out of this, it's cross connected to this agent that shares a lot more details about what Sewell can do and where to find additional information and so on.

But this is managed by Larry, so I don't have to worry about this.

All I need to worry about is staying here in the chat and keep interacting with Larry.

And if things go wrong, readjust and say update any relevant information and things like this.

I'm doing this since last year, June.

And I'm doing a lot more complex things than what you see me building here.

And it never failed.

It never failed.

The real skill that people need to learn is how to ask the right questions.

This is really the thing that people need to learn in a new generations too, asking right questions and team management.

I think, everything that we teach in myICOR based on the ICOR methodology tool agnostic and gives you a tool agnostic understanding of your productivity system end to end, which is the foundation of any AI tool that we are building here.

But also building businesses and managing teams. It's all based on the Icom methodology.

So if you're not a member yet, this is the moment to join.

Okay, and here we go.

It's finished.

Let's see what it looks now on the front end.

And here we go.

Beautiful.

This is now light mode.

It's no longer dark mode as we had before.

The brown one, if you remember, which was this one.

So this, if I reload this page, it will now switch to this mode too.

I have this journal.

We see here's the entrance that we just created with the image.

It's perfectly working.

I have here the connection to Paco.

It opens up Paco, with all the context available I. I have here the context that go beyond Paco already, but it added Paco and my wife too.

It even shows me here, Paco three journal mentions.

So you see how everything gets perfectly integrated and now I could have here the meeting nodes and it starts connecting the people and all these things and talking about meeting nodes one of the upcoming videos will be exactly about this, how you leverage your meeting.

No transcripts in the best way possible to enrich your PKA system on autopilot.

This is, that's a real game changer.

So if you haven't subscribed to.

Subscribe now or join us in my icorp to stay connected or access any of our courses or even our coaching calls that we have on a weekly basis with PA and me.

If you need any additional assistance with your PK system, I catch up in the next one.