WEBVTT

00:00:00.000 --> 00:00:22.235
If you're following me on this channel, you know that a local folder just replaced all my note taking apps. And this channel was all about note taking and personal knowledge management using all the different tools in the past years. But just with their latest releases of AI that allows us to access local files, everything ended up local. To make my own personal knowledge management that we call personal knowledge assistance

00:00:22.235 --> 00:00:51.825
based on this folder structure that self organizes everything. I never need to link things again manually. So the thing is, this is all nice if you're sitting on a MacBook, but what if you're on a go? I could now connect this to my cloud and I can start typing and even talking to it, but that's maybe not something that you want, that you want to use the apps or you might want to have a much more asynchronous input. And that's where people use then third party tools just to capture information.

00:00:51.905 --> 00:01:04.410
But that's something you don't really need. You can have it all for free just using your iPhone if you have one. I'm pretty sure on Android you can do something similar, but on the iPhone, you have an app that's called shortcuts.

00:01:04.410 --> 00:01:35.540
It's for free. It's already on there, but let's open this up. And here, I just deleted all the shortcuts we can start from scratch. On top right, you can create a folder. I call this folder simply my p k a. Give it a icon. Now we have this folder, now I start creating these shortcuts in here. Something I want to do is capture audio on the go. Whenever I have some insights or inspirations, I want to quickly save it. But I don't want to have it in the voice memos that I need them to extract later. I want to have it in the final destination,

00:01:35.700 --> 00:01:38.100
in my local folder, on my MacBook,

00:01:38.100 --> 00:01:47.675
ready for my AI to pick up and to process. And this is how we do it. We just create a new shortcut. Search here for audio. We can use here record audio,

00:01:47.995 --> 00:01:51.355
and here when we use the drop down, we can change the quality.

00:01:51.515 --> 00:02:15.085
On start recording, we have on tap or immediately. I like this to happen immediately when I call the short cut. And now we would already just record the audio, but we want to have it end up in the folder. So I just simply search for save, save file. I use the drop down. You see it picked already up here, recorded audio. So save the recorded audio from the previous step. Ask where to save. No. I want to have it fully automated.

00:02:15.085 --> 00:02:17.885
So therefore, I will say here, two shortcuts.

00:02:17.965 --> 00:02:20.205
Instead, I will just tap on this shortcuts.

00:02:20.205 --> 00:02:46.805
Now I'm in the file system, and I can go into files, and the amazing thing is I can connect any external cloud services too, like Dropbox, g drive, and so on. You see here already Dropbox. I like to synchronize my local folder with Dropbox. I could also use iCloud to synchron ize this. In this case, I'm using Dropbox. Here's already the folder that I have on Dropbox. Here's my p k. That's the folder that I'm using for my knowledge base, and here's a team inbox.

00:02:46.885 --> 00:02:52.405
And I want to have it saved in here. I can create here a new folder called audio

00:02:51.970 --> 00:02:52.770
captures,

00:02:52.850 --> 00:03:22.905
and here we are, and I'm opening this up, and that's it. I don't need to use a sub path here. I just keep it as it is, and it is as simple as that. And we can give it a go by simply tapping now the play button on the bottom right. Hey, this is a test. Out of the testing of the shortcut, if this is working. I tap again and it just recorded it. It asks me now if it allows it, shows me the result. Okay? So now it's ready. Here it is, save file. We can just hold down and rename this. We just say audio capture.

00:03:25.030 --> 00:03:40.645
We can go to details, and in here, we can have it now also on Apple Watch. So I can, on my Apple Watch, trigger this shortcut and talk something to end up in the local folder. I go to edit and I can quickly here choose a different icon.

00:03:40.725 --> 00:03:51.870
Mic. Here we go. And different color. And there we go. All you capture. So I can just tap on it and run it. But obviously, this takes forever. That's why I can go now to my settings,

00:03:52.110 --> 00:04:07.425
and here I can go to the action button if you have one on the side. And here I can actually pick up a shortcut. So here, if I click here, I can go to show my shortcuts. Here's the audio capture, and there we go. And now, I'm going outside.

00:04:07.585 --> 00:04:20.305
Boom. And I just need to press and hold on the side button. Alright. This is just a test if this is working. I'm not just holding down the side button now to see if this is actually going into the folder on my local drive

00:04:20.740 --> 00:04:31.300
on my MacBook. I tap, and that's it. Nothing else. You can also get the control center, press and hold, and here you can add a control, and here you can search for shortcuts,

00:04:31.380 --> 00:04:46.795
and you tap on shortcut. Now you can choose the shortcut. I can go into here. There's the audio capture. Now I have it. Here it comes up. Now I can make this just small. And now I have it in here, and I can do exactly the same thing. Just drop down, click, hi everybody.

00:04:47.035 --> 00:05:25.410
This is a test. I just want to make a journal entry about x and y, and that's it. I tap and this is it. It recorded it and it saved it to Dropbox. And now when I go to my local folder here on the Mac, here we are. This is my folder team inbox and here we see already audio captures, and here are the things that we just recorded. It instantly appeared here. I can go through them, and the beautiful thing is now there's no title, nothing. As I show you, you can open this now simply in Claude Coburg or anything else. I always prefer to run this in a terminal or even better in code, but for the sake of demonstration,

00:05:25.410 --> 00:05:31.730
I show you just right click and then go down to services and new terminal at folder. I launch Claude,

00:05:31.730 --> 00:05:51.130
and what this does, Claude opens up now in this folder for the sake of it. Here's the folder. And all I now say is process the audio captures in the team inbox folder. And here I could have now re routines and I can even trigger this on autopilot to check every now and then there's anything in my team inbox. So this would happen automatically.

00:05:51.450 --> 00:06:09.535
So it reorganizes everything on the fly. I could even give instructions what it should do next. So I can really go wild when it comes to this to automate the things simply simply by sending audio messages this way. And the thing is you see here already, it's taking up something called Whisper CPP.

00:06:09.535 --> 00:06:12.495
It's completely for free. It's open source. FFmpeg

00:06:12.495 --> 00:06:58.025
is also something that is open source. So all these things do. It converts the audio into the right format, so it's perfect to read. Then it uses the whisper to transcribe it. Obviously, in my shortcut, I could have also added a transcription step. I trust Claude to do a better job here together with this whisper to get a right transcription, and this is what they are doing right now. Uh, we provide this as a whole expansion pack to have all out of out of the box. But if you don't know, you just ask, hey, can you transcribe the audio in my inbox? And it says, no, I can't, and gives you the resolutions. Alright? Because that's very common things that you can download, and Claude will do it for you or Gemini or Codex. They will just download these functionalities,

00:06:58.265 --> 00:07:08.905
and they can convert audio into a different version. They can transcribe it, all the things. So now it says here, the four audio captures, it recognized it, so we can go here. Tom inbox,

00:07:09.360 --> 00:07:32.685
audio captures here. All short cast tests with no content. What should I do with them? So is there really no content? Create one MD file that holds all the transcripts of all four files. So obviously, I should have used proper text to test this properly because now it says there's nothing because it's just tests. See, it even found it recognized there are two others that I did from earlier testing.

00:07:32.765 --> 00:07:42.260
So, yes, I want to have them too. So it even picked them up. Now we have this audio capture folder, so you can do this however you like. And then you train your team accordingly,

00:07:42.260 --> 00:07:53.035
so it knows next time how to process this, and that's how you streamline the process. You could have even, for example, trigger words, where you have a specific trigger word. If this is part of the transcript,

00:07:53.195 --> 00:07:58.075
then Larry and his team or your AI team will process this accordingly,

00:07:58.155 --> 00:08:07.590
and that's how you can really think forward. This is where we are so independent in a local folder using these three tools to build really use specific

00:08:07.590 --> 00:09:01.825
personal knowledge assistance system. So done. All six. It shows me now there is a transcript and where it is and where is it. It's in deliverables because our team works with a deliverables inbox that is here where it shares everything that it was working on. And here it is. Hey. This is a test. There we go. Silent. That was the quick things that I tested. Alright. This is just a test. And it's exactly what I was saying to the phone as you you can go back in the video and double check, but this perfectly worked out of the box, guys. And therefore, all I need to do next, I have an idea, hold down this button and start talking, and I can be sure it ends up in this local folder. What about you? Would you pick this up, or do you rather use a third party tool where you go to the tool and and unlock your phone first and all these things? Or are you migrating now to the solution? Let me know in the comments below. And if you haven't already, subscribe to the channel so I can catch you up in the next one.
