¡Obsidian reventó la IA con MEMORIA GRATIS e ILIMITADA!… — Transcript

Discover how Obsidian offers free, unlimited memory with AI-powered organization to act as your second brain for knowledge management.

Key Takeaways

Obsidian provides a powerful, free, and unlimited platform for personal knowledge management.
AI integration enhances Obsidian’s ability to organize and relate complex information automatically.
The tool emphasizes privacy and independence from third-party services.
Users can easily scale their knowledge base with extensions and cross-device synchronization.
Obsidian’s local storage in Markdown ensures data control and longevity.

Summary

Obsidian is a free, unlimited, and private knowledge management tool launched in 2020.
It functions as a second brain by organizing, linking, and visualizing information seamlessly.
The platform supports multiple devices including iPhone, Android, Windows, Mac, and Linux.
Users can create vaults (folders) to store and organize notes without registration.
Obsidian’s privacy is strict, with no third-party dependencies and anonymous company leadership.
The video demonstrates five use cases including automatic organization, concept linking, and visualization with canvases.
Integration with AI agents like Codex allows restructuring and relating stored information for better understanding.
The presenter imports raw YouTube video transcripts into Obsidian to show how AI can convert them into structured knowledge.
Optional paid features include note synchronization and publishing tools but core features remain free.
Obsidian stores data locally in Markdown format, ensuring data ownership and portability.

Full Transcript — Download SRT & Markdown

Speaker A

Storing information isn't the same as having knowledge, but finding it, understanding it, and turning it into action is. Information companies are becoming increasingly capable, offering more features and providing us with the information we need. However, when we move from the generic to a more specific context, they often forget and become overwhelmed. But

Speaker A

Obsidian isn't just a solution that improves this; it can act as a second brain.

Speaker A

To see for yourself, we'll explore what it is, who it's for, and its possibilities with five use cases that will allow you to do things like automatically organize all your information, easily relate concepts to each other, visualize it with canvases,

Speaker A

scale it with its extensions, or even have it count everything, among other uses. The best part is that you can apply everything today because of its simplicity, and it's also completely free, unlimited, and requires no registration. Sounds good? Let's get to know this tool.

Speaker A

Obsidian is a platform that was born in 2020 during the pandemic with a clear manifesto: to be accessible, timeless, private, flexible, and independent of third parties.

Speaker A

The privacy is so strict that the CEO, known as Kepano, wrote that even the users of the platform don't know who they are. To see what it's capable of, let's jump to Obsidian. This is the platform I'll be

Speaker A

linking to below in the description. Once you land on this page, you'll find its focus on sharpening your thinking through a free, flexible, and private application. Keep in mind that this tool, if we click on "More Platforms," is available on iPhone, Android, Windows, Mac, and Linux.

Speaker A

Basically, all devices. Now we'll go through the installation process so you can see how easy it is. But before that, I want to show you how their pricing pages clearly state that this tool is free and unlimited, requiring no registration or

Speaker A

conditions. Keep in mind that they also monetize to be an independent company through some optional extras that are primarily to support the project.

Speaker A

The first one simply synchronizes all your notes across all your devices, although you can also get this for free. The second one is more focused on publishing everything you can do, which we'll see throughout the video, without having to rely on third-party tools. Then there are

Speaker A

some one-time payment plans without a subscription, which, beyond giving you a small community advantage, mainly support the project. With these plans, the platform itself and all its features are 100% unlocked. Let's see what it entails. To do this, going back to the download section, all we have to do is click on download

Speaker A

for our device. Once it's downloaded and installed, you'll see this interface, which doesn't even require registration. From here, we'll have three options. First, we need to create a new vault, which will essentially be the space where we store all the information in an organized way. We could also open a

Speaker A

folder on our computer or sync it with other devices. I'm going to create one from scratch, so I'll click on "create." Now all we have to do is give this vault, which is really just a folder, a name. So here I'm going to put the word "memory."

Speaker A

In the location field, I've selected my downloads folder. And if I click on "create," we'll go directly to the platform where we can start working.

Speaker A

From here, keep in mind that the interface is very simple. We simply have the notes section in the middle where we can put all the information. These can also have different links to connect different areas of knowledge. Notice that if

Speaker A

I go back, here on the right, I'll be linking "Welcome" with "Create a link." And next to this, on the left side, we can also add notes from here.

Speaker A

Add folders and even subfolders here, and the rest is just organization. From here we can expand everything, from here we can compress it, and on the left we only have a series of shortcuts. So, the area where we really

Speaker A

need to dedicate more time is this whole section of organization by folders and notes, and defining all that knowledge base and memory we want to have organized, although with artificial intelligence we can already automate all of this.

Speaker A

Starting with the first use case, we'll be learning about Carpac's method. As the co-founder of OpenAI and having worked as an executive in artificial intelligence at Tesla, he has just optimized, with the help of AI agents, a skill that allows us to organize all the information relationships

Speaker A

we input using Obsidian. To see what this is all about, let's jump to its capability.

Speaker A

This information, which I'll also include in the description below, will show us how this instruction allows any information stored in our folder or vault to be restructured and related, making it much easier for both humans and AI agents to understand. To

Speaker A

do this, all we have to do is copy this entire instruction. I recommend clicking this button here so you can simply copy everything with Ctrl+A and Ctrl+C. Once copied, keep in mind that we can use this with any of its agents, such as Google Antigravity, Cloud Code, Codex, and others.

Speaker A

For this example, I'll be using Codex, as it's one of my favorites and the simplest way to present everything. So, the first thing I'm going to do is grant it access to the project. To do this, we simply need to work on a project. I'll click

Speaker A

on "Add new project." I just opened the folder we created as a vault through Obsidian, so I'm going to open it, and that way the agent will have access. But now we're going to make it smarter so it can structure all the future

Speaker A

information. With the whole promo pasted, I'm going to click send, and then I'll start analyzing that we're looking at a new vault, and I'll be setting up everything to restructure all the future information. While this is happening, I've prepared

Speaker A

the raw transcripts of my latest YouTube channel videos here. Videos that, as you know, are between 20 and 30 minutes long. Here, in fact, if I open one of them, we'll be able to find information like the title, key information, the description I had,

Speaker A

and further down we'll see the complete transcript of all the content of that video. So here I have that with 20 videos, and of course, I have a lot of information here that's really independent of each other. In fact, to see this, let's first look at the

Speaker A

setup I've already had done, creating a directory called "raw" and another called "wiki," which is where people will work. And if I go back to Obsidian, I do indeed find those two folders. On one hand, there's the "raw" folder, which Ritmi tells me I should

Speaker A

put all the raw material I want to convert. And then in the "wiki" section, we can see other options that are currently empty, where people would convert everything into concepts, entities, and other key points. To begin, I'm using a

Speaker A

cleaner approach. I'm going to delete the previous folder, and with that, we're ready to start working. Therefore, I'm going to open the folder we created in Obsidian with all these copied transcripts— that's also crucial to keep in mind. Every time we talk

Speaker A

about a new vault, we're literally talking about a new folder. In other words, regardless of whether we delete the application, this will always remain locally in a Markdown language on any other platform. So, knowing this, I'm going to click on this section and from there

Speaker A

Here I'm going to import all those transcripts. Once I've imported them, everything will be ready, so from Codex I'm going to tell it to transform them. Look, I just imported the transcripts from my last 20 videos on my YouTube channel of

Speaker A

artificial intelligence tutorials. So keep in mind that I talk about tools, use cases, and so on. With all the information you already have in the project, structure it using the workflow you've learned. I click submit, and it's going to start working on it.

Speaker A

Notice that it's going to process all the transcripts we have in the first raw folder into the wiki folder, where it will reorganize and create all the file.

Speaker A

category that might be necessary. So now all we have to do is wait. And so, after about 4 minutes—this will also depend on the model you use and the amount of information, although in all the tests I've done from

Speaker A

Codex, I'm usually around this time—it seems to have processed the 20 transcripts, each averaging about 30 minutes in length. It tells me it has created 20 pages as source code, 11 as concepts, and 30 as AI tools or entities.

Speaker A

It briefly explains how it organized everything, but to see it better, I'm going to go into Obsidian. And look what we have here; without touching anything, the AI agent has already reorganized everything. Before we look at the graphical view,

Speaker A

notice that besides having the raw data, which is what we had before, we now have everything well organized in the wiki with concepts. We also have the entities here with all the tools I've been talking about in those videos I uploaded, along with some

Speaker A

additional information. To see all of this more clearly, I'm going to look at the graphical view, and you'll see how everything is interconnected. Here we can see all these concepts and tools. And if I go to any node, like the Google ecosystem,

Speaker A

we can see how it links to videos such as the seven bombshells dropped by Gemini.

Speaker A

It also links to tools like Google AI Studio, which in turn connects to other things, such as concepts like research and synthesis with artificial intelligence. And each of these interconnected nodes has a corresponding documentation. In fact, if I click on "research and synthesis with AI," we'll go to this document created by

Speaker A

the AI agent, which includes a brief summary, key points to consider, the tools mentioned, and all the videos where I've been discussing this topic. This would also be repeated, for example, if I click on any tool like Gemini, and in

Speaker A

the same way, all that information would be interconnected, such as key points, videos where I mention this, and the interconnections we would have even with the concepts. And look, even though we have a lot of raw data, now, through this methodology and the

Speaker A

visualization using Obsidian, we're going to be able to interrelate all those concepts and really optimize the learning and use of all this data. As we've just seen, we now have several tools and workflows that open up a lot of possibilities. And this, with

Speaker A

a free and limited approach, can also be applied in professional environments. That's why I want to tell you about Chatfuel, a platform that frees you from responding so you can lead.

Speaker A

They've just launched a new feature that allows you to automate your Instagram completely free, with no limits or fine print. Chatfuel combines its own artificial intelligence with being an official Meta partner, allowing for native and stable automation. Doing this would be

Speaker A

as easy as coming to this platform, which I'll also link to below in the description. I'll log in with your account. Once here, you'll enter your phone number and you'll automatically be connected to a conversation with Sofia, an AI agent.

Speaker A

which will configure everything automatically. Here you simply have to enter your Instagram name. In this case, I'm going to use "artificial academy". With this, people will read the profile, discover audience questions, draft the best automatic responses, and prepare

Speaker A

the entire space. Just a few seconds later, we would only have to activate it. And notice how, automatically, it tells me that Sofia has configured the word "info" to write a public comment and also a direct message. And if you write the word "plaza",

Speaker A

the same thing would happen, but with information related to it, since it has deduced that it's an artificial intelligence academy. Furthermore, as it progresses, it will show us how this will look. And here in the first image, we can see how every time someone writes the word "info",

Speaker A

it writes a natural way of sending all the details via direct message, both with "info" and the word "plaza". And along with this, we can also see a preview of what the private message will look like. Once we have this, we can tell it to change something, or conversely,

Speaker A

we can tell it to look good, and we'll have everything unlocked so that when someone types this keyword, they'll always get a personalized response 24/7 through an AI agent. Furthermore, thanks to Chatfield sponsoring this video, beyond providing this functionality for free (albeit limited),

Speaker A

they've given me a special link, which I'll leave in the description below, that will also give you access to the pro features for free for 30 days.

Speaker A

I'll leave the link in the description below, and now that we know this, let's look at the second use case.

Speaker A

In this second use case, we'll have an even broader view, as we'll also cover how we can easily extract information. This is very easy to do thanks to an Obsidian extension called Webclipper. This extension, which is also completely free, will allow us to highlight, capture, and save

Speaker A

any information we find on the internet with a single click. To do this, all we have to do is click on this "Add to Chrome" button, although keep in mind that it's also available for practically all other browsers. Since I have it in Chrome, I've already

Speaker A

installed it. Also, notice the good ratings we have. Now that this is available, I 'm going to go to this Wikipedia page about the Earth's timeline, where we have a lot of information, and we're going to see what this Google

Speaker A

Chrome extension can do. To do this, the first thing I'm going to do is click on this icon, which, by the way, if you haven't pinned it, you'll find here. All you have to do is click on "Pin." And once

Speaker A

it's open, notice how it has automatically extracted all this information in an organized way, including the title, the source URL, the site (Wikipedia), the date it was created, and all the information that's actually on this page.

Speaker A

But before importing all this information and converting and organizing it, notice that we also have some very interesting features we can start working with, such as this option here, which, when clicked, directly switches us to a much easier-to-read, simplified interface without distracting dots.

Speaker A

We also have other features, such as the one I could use to highlight this phrase here. I click on "highlight," and that way, when I import it, it will have higher priority. However, I'm going to remove what I just highlighted, since I'm going to import it as is. But keep in mind

Speaker A

that from here you can modify everything, including increasing the font size, changing the font itself, and even making bulk changes. However, I'm going to work directly with the artificial intelligence agent using the methodology described earlier. To do this, you can click on "add to Obsidian." This

Speaker A

It will take you directly to the last project you have open. So, first of all, I'm going to create a new vault. To do this, keep in mind that in Obsidian, on the vault you're currently in, you can click on the last section in the vault at the bottom left and

Speaker A

select "Manage Vaults." Once you've done that, you'll be back on the home screen, so you can click "Create." I'm going to name it "Land," since that's what the information is about.

Speaker A

I've selected the Downloads folder again so that everything is saved. I click " Create," and now that this new vault is empty, I'm going to import it. But first, in the folder where I want it saved, I'm going to select "Raw," since

Speaker A

that's the methodology we're following. I click "Add to Obsidian" and confirm. And that way, we'll automatically have that information perfectly pasted, including the images, and everything perfectly organized. And now we could repeat the process, even though we already have the

Speaker A

information here, to see how this would work. I've gone back to Codex, but this time I'm going to start a new chat and we're going to change the folder to the new one we created called "Earth." Click "Open

Speaker A

," and all I would have to do is copy Carpac's instructions again. But since we already have all the information, I'm going to give it a little more context so it works perfectly the first time. Look, keep in mind that here you have the entire methodology I want you to generate

Speaker A

now to organize the information with the article I have on Earth's timeline. But keep in mind that even though the instructions and the information you have in the folder are in English, I'll want everything in Spanish. In addition to all the

Speaker A

new documentation you create, even if there's no information in the original file, I want you to expand it in Spanish with relevant information. I click "Send," and with that, it will start processing again. But now, even though we're working on an article

Speaker A

in another language, and the instructions are also in English, specifying that the instructions should be in Spanish and that it should expand with more information according to the different concepts that appear in that article for all the new files it creates

Speaker A

will really enrich everything in a better way. Here, in fact, we can already see that it says it's dealing with more than 250 events with a clear hierarchy of eons, eras, and periods.

Speaker A

So now all we have to do is wait and see how long it takes to do all this organization. About 5 minutes later, on the first try, I can see that along with the raw file, we find the wiki with all these categories, including concepts like eons,

Speaker A

as well as different key events and other categories such as synthesis. And to see all this more clearly, I'm going to click on this section to open the graphical view. And look here at how we already have all these interconnected nodes.

Speaker A

If I click anywhere, like on the emergence of Homo sapiens, we can see how it links to all these different points here. And again, if I click on this document, I'll see the central idea, why it matters, the context, and all the connections we find, such as the geological timescale

Speaker A

. And here again, we can expand the information with these key points, relationships, connections, and other information that allows us not only to save this in an optimized way, but also so that when I return to a new chat with this agent that has configured everything for me,

Speaker A

I can ask specific questions, and the agent will have access to all that structured and related context to answer me in a more optimal and efficient way. For example, if I ask a question like, "Hey, I'd like you to tell me the key points about Homo sapiens."

Speaker A

Click submit and we can see how, to find the correct location, it will search the wiki where all the relevant contextual information is already stored. I'll begin from here.

Speaker A

Just seconds later, you can write all that information very clearly, thanks to the fact that, beyond having a document about Homo sapiens, it also provides a global and strategic view of all the milestones and key points related to it.

Speaker A

With the third use case, we'll be taking it to the next level, since everything you've just seen is only a methodology, and we can actually automate much more with the help of Obsidian and artificial intelligence agents. Going back to the Obsidian website,

Speaker A

notice that we also have a CLI mode, that is, an option that allows this tool to work using commands. This will give you some more information. We'll see that its installation is very simple, and with it, we can do practically everything,

Speaker A

such as working on its fundamentals to open daily notes, as well as other quick access tools, aspects related to development or automation with all the memory or " brain" we're building. To install this, we simply need to activate it from

Speaker A

the general settings. So, going back to Obsidian, I'll click on settings, and from the first option, scrolling all the way down to where it says "Command Line Interface," we need to enable it. After registering, it will be ready to use. But wait,

Speaker A

we won't even have to delve into the technical aspects, as the CEO has created something very interesting. Specifically, it's this repository here, created by the one we already know from Kepano, where we can find a series of skills

Speaker A

that correspond precisely to these five here. These skills basically allow us to create files in the Mmacdown format, and also create and edit different databases, views, filters , and formulas—in other words, all the configuration directly on this platform. Similarly,

Speaker A

we could create canvas files in Jason, use all the Obsidian CLI commands we just installed, and other functionalities. To do this, keep in mind that all we have to do is copy the command. From Codex or any agent like

Speaker A

Cloud Code or Google Antigravity, we would have to open the terminal and paste that command so that when we send it, clicking on the ISTE will trigger the installation. Alternatively, since we're dealing with an agent, you could copy the skill's URL, select the "Skill Installer" option,

Speaker A

paste it, send it, and the installation process would begin. While this is happening, if you'd like to learn everything related to AI agent skills, there's a video here that really elevates the level of the agents. About 30 seconds later, it would tell me that it had installed all

Speaker A

the skills we've seen before, and now, among other things, it will even be able to create canvases, a feature we haven't seen before. But keep in mind that, going back to Obsidian, by clicking this button, you can create a fully customized mind map.

Speaker A

You can even change the color of these nodes and interrelate them with other cards or even existing vaults. I, for example, am going to link it here to one of the nodes, and I could even link it to different images from more

Speaker A

documentation. Along with this, you can add all the information you need, in any format you want. You'll even be able to create different boards and interconnect them, keeping your entire visualization well-organized in a more strategic way, as well as any concept you want to see differently. This feature is great, although

Speaker A

again, with artificial intelligence agents, we can also automate it. To do this, I've gone back to the AI agent and I'm going to ask it. Look, I'd like you to use Jason's Canvas ability to create a canvas for me where everything is presented in a

Speaker A

clearer, more visual, and organized way so that I can see the most relevant points at a glance.

Speaker A

I click send, and it starts processing. Notice that even though John has selected the The skill, it would tell me that it does indeed use Jason Canvas's skill, since I've mentioned it that way too. A couple of minutes later, we'd have it done. It

Speaker A

would tell me that it passed validation, 31 nok done, and 25 connections. I haven't seen it yet, so I'm going to click on the strategic map in this Canva format. And from here I could see the whole Jason,

Speaker A

since that's how that canvas format is actually structured when we export and import it.

Speaker A

But to see it better, I'll just jump to Obsidian. And from here I'm going to look for the file now. To do that, I'm going to select this icon, and notice that here it says wikisíntesis, mapa estrategiadelatierra.Cvas, which is exactly the name it gave me here.

Speaker A

So I'm going to click on it, and look how we already have that mind map here with all the relationships in a clearer way, having on one side the time column, what the drivers of change have been. With all this information, we can also see the

Speaker A

hinge events and the key pages that we should take into account. Since this is all documentation-based, we have all the information linked here, allowing us to build extensive databases and automatically organize not only the knowledge base but also the canvas visualization.

Speaker A

Moving on to the fourth and penultimate use case, we'll discuss plugins. These are extensions that allow us to enhance and personalize the entire Obsidian experience. To do this, I went back to Obsidian, and to enable them, you had to click on settings. Then, where it says "Community Plugins," you had

Speaker A

to click on it and activate the option to use them. Once you've done this, if you click on "Search," you'll see more than 3,600 extensions that you could install.

Speaker A

However, I recommend viewing them directly on the website, as clicking on the "Community" section will give you a clearer overview. For example, you could click on the "Most Popular" category. And from here you can see, through a rating, how satisfied users have been

Speaker A

with this extension. Here, for example, by clicking on the first one, which has a 100% rating, we can see how this tool makes Obsidian more visual, with more colorful labels, images directly in the previews, and even other views like the calendar. If we find an extension suitable,

Speaker A

all we have to do is click "Add to Obsidian," click here again to confirm, and once inside, simply click "Install." Now, click "Activate." And with that, if I close this here, notice how the experience has changed, with different key relationship points. I can also see

Speaker A

how the folders have a series of icons, as do the labels, which in this case don't have any. Additionally, the entire usability flow has also changed, and in this way, we can really customize the experience even further. Going back to the plugins section,

Speaker A

keep in mind that it also has its categories. For example, if I click on integrations, I'll find others, such as the ability to integrate maps. To see this, I'll click on it, and as it shows us the preview, this will also allow us to integrate

Speaker A

fully interactive maps, to which we can also add a knowledge base. Along with this, you also have other categories, such as the artificial intelligence section, visualization sections, and other settings. What I'd like is for you to leave a comment telling me which one you like best. In addition to all the plugins,

Speaker A

we also have a themes section from which we can completely change the entire interface.

Speaker A

Here, as an example, I'm going to select this Cupertino theme, which is an Apple-style theme. So I'm going to click on "Add to Obsidian." I confirm, install, and it automatically... notice that it...

Speaker A

We've changed this entire visualization again, and here we can also see those images with the previous plugin we installed. Ultimately, we'll have an interface here that's more similar to what we'd actually find on an Apple device.

Speaker A

Moving on to the fifth and final use case, I want to mention how you could have this entire platform 100% private. In fact, it is private because everything is stored on our computer, and we're working with an AI agent that operates on local files. But keep in mind that

Speaker A

we typically run these more powerful AI models in the cloud, whether with Cloud Code, Google Antigravity with Gemini, or OpenAI models through Codex, as we've seen throughout this video. By the way, speaking of Codex, if you'd like to see how to get the most out of it by exploring all its features,

Speaker A

I discussed it in detail in this video. But beyond using all these features, keep in mind that we could also be using open-source platforms. These tools allow you to install AI models locally so that the entire AI layer can operate on your own computer, without

Speaker A

needing to connect to the cloud. Everything remains closed and isolated on your device. Installing local AI is very simple these days , and we can use platforms like M Studio, Yang, or even Olama. Olama also gives you unlimited access to premium tools like Cloud Code

Speaker A

thanks to its ability to interconnect open-source models, something we also covered in this video here if you'd like to learn more.

Speaker A

The truth is, with all this, both AI and we can have a clearer view of the map. It's not about remembering more routes, but about understanding the path better. No.

Topics:Obsidianknowledge managementsecond brainAI integrationfree memorynote takingpersonal knowledge baseprivacyMarkdowninformation organization

Frequently Asked Questions

Is Obsidian really free and unlimited?

Yes, Obsidian offers a free and unlimited core platform with no registration required. Optional paid features support the project but are not mandatory.

How does Obsidian ensure user privacy?

Obsidian is designed to be private and independent, storing all data locally without reliance on third parties. Even the company leadership remains anonymous to users.

Can AI be used with Obsidian to organize information?

Yes, AI agents like Codex can be integrated to automatically restructure and relate notes within Obsidian, making information easier to understand and use.

Get More with the Söz AI App

Transcribe recordings, audio files, and YouTube videos — with AI summaries, speaker detection, and unlimited transcriptions.

App Store Google Play

Or transcribe another YouTube video here →