I Gave an AI Agent the Keys to My Life (Here's What Happened) — Radek Sienkiewicz (@velvetshark-com)

SPEAKER_00

Reza Pannczak Reviewer Are you ready to go? All good, yeah. Please welcome, Radek, everybody.

SPEAKER_00

Hey, I'm Radek, I'm one of the open-claw maintainers, and I want to talk about what happens in my life with OpenClaw when I practically gave the keys to my life to OpenClaw. And it's almost literal and what that actually means. So this happens step by step. It wasn't all at the same time, but it can access my emails, it can access my notes, files, calendars, tools, my operating system, automations. And it builds on top of memory of everything that I do at the computer. So it can do anything with it that is possible to do with the computer.

SPEAKER_00

But it didn't all happen in one big leap. So I install OpenClaw and now it controls my life and does everything for me. That would be silly to do or even silly to expect that this could work. So what happened is that I tried installing, just like everybody does, just one channel. I think at the beginning it was just WhatsApp. Then I migrated to Telegram. Now I'm on Discord. But it was just WhatsApp, just one ability to chat. Okay. So we are there. What's next that we can do? Let's do one simple workflow or one very simple task that we can do. Once we are there, let's go to the next step.

SPEAKER_00

So this is how it happens where I am today. I used to think that I have quite a simple setup with my OpenClaw and what it does because I never did any big change. But when I encounter different Twitter threads, YouTube videos, or talking to other people about how they have it set up, I see that my setup has everything that they have, more on top of that. And most also is more sophisticated than what I see out there. Which was really surprising to me because I felt that it's one small step at a time. I have a pretty simple setup. Works for me. But that's what I want to show how that happened and how it looks like today.

SPEAKER_00

So you already had many talks about how the sausage is made, how we are making it better. You will have more talks about the inside of OpenClaw. I want to show how it looks from the other side. From the first, the simple user, then the power user. Now I am also a maintainer. You don't have to go to the maintainer route. But when I was playing with one of the workflows, I just encountered some errors and submitted a first PR, then the second PR, then just looked into Discord, and then you just got involved. Now I am a maintainer there.

SPEAKER_00

So that's the setup. These are the steps that it usually happens. I see the need. I solve it in a very simple way. And then I add more steps to it. And this is why I usually don't have big issues that people have, that okay, now it broke my computer or it completely bricked during the update. Because I have all these small steps that I take. If something breaks, I take one small step back. Fix it. See what doesn't work. Understand why it didn't work. Have a setup that it never happens again. And take one step further again.

SPEAKER_00

So where it started being more and more helpful and running my life is when I gave it my knowledge base. So I had a lot of stuff in my Obsidian, which I built up for years. So right now I have about 3,000 pages or notes, markdown files in my Obsidian. And this is everything. This is work stuff, personal stuff, tasks, projects, research, articles, an inbox of links that I'm putting there. And it then finds the connections and puts it in perspective and in context to other stuff that I have.

SPEAKER_00

So all of that is now accessible through my OpenClaw with a very good search. I have search and memory. I have normal search. I have QMD search for Obsidian. I have different memory for my workspace. And all of that is interlinked. And that's where that magic happens.

SPEAKER_00

And when I saw recently, that's where it hit me that I probably don't have a simple setup. When I recently saw Andrej Karpathis' tweet that went viral where he talks about LLM knowledge bases. I was reading that and it's just like, yeah, that's exactly what I have. What's super revolutionary about it? And then I understood that okay, so I got there step by step. It works for me. So it's probably worth sharing, telling more about it, showing how it works, showing how you can get to that point as well.

SPEAKER_00

And for example, for Obsidian, do we... Yeah, this is the real screenshot of my vault and all the nodes. And these are different clusters. Some are probably project related, the big clusters. Some are one-off. These are probably bookmarks. And one of the tasks that I'm doing is that when I add something to inbox, it takes that link that I add there, looks at what's there. It could be a tweet, it could be a thread, it could be an article, it could be a YouTube video, analyzes it, adds tags to it, adds context to it, looks at what's already there on this topic in my vault, how it could be helpful in other areas, and adds connections to it.

SPEAKER_00

So what previously was just Twitter bookmarks that you bookmark and never go back to, now it adds more context, builds up my knowledge base, and is much more helpful, and even surfaces things for me when I added bookmarks that okay, so you already had this and this and this about this subject, and this is how it connects. Maybe you should look at those notes, and very often it's just like yeah, I completely forgot about that, and that's a good source of knowledge and thinking about it, because there was a reason why I'm adding this bookmark.

SPEAKER_00

So that's where it's starting to be super useful. On top of that, also at 4am, or 4am is just an example that I have, it happens probably between 3 and 6 more or less. Now it just adds more context, builds up my knowledge base, and is much more helpful, and even surfacing the things for me when I added bookmarks that, okay, so you already had this and this and this about this subject, and this is how it connects. Maybe you should look at those notes, and very often it's just like, yeah, I completely forgot about that, and that's a good source of knowledge and of thinking about it, because there was a reason why I'm adding this bookmark.

SPEAKER_00

So that's where it's starting to be super useful. On top of that, also, at 4am, or 4am is just an example that I have, it happens probably between 3 and 6 more or less. So this is what is happening when I'm sleeping. So when I'm sleeping, again, my agent does everything so that it runs well, it indexes everything, it backs everything, so that worst case, if I lose something, I lose maybe a couple hours of work, of content, of anything else.

SPEAKER_00

So it refreshes all the indexes for QMD, for memory, for my Obsidian Vault, and I start fresh in the morning with whatever waits for me, maybe summary of the emails, of the calendar, everything updated, the latest OpenClaw version is waiting for me, which also took step by step. I have some scripts around it, so that it knows what to do and what not to do when updating, what can break, why it breaks, how to verify it before updating or before restarting your gateway, so that it is able to come back online again.

SPEAKER_00

So that all is also automated and as I get up, it's already waiting for me, fresh and ready for me to start the day. And each OpenClaw is, I'm not a big fan of sharing my exact setup, because that exact setup is very specifically for me, for what I need right now, for what I will need in the near future, for the errors that I encountered, for issues that I want to be solved. But to give you some ideas so that we can talk more also about specifics and not just in general, so these are some five areas or five types of jobs that my agent is doing.

SPEAKER_00

So the first one is, I want to talk more about updating. The first one is, I want to talk about the first thing. So this is what I just showed you. So it does all the updating, it does all the plumbing, it does all the stuff that needs to happen, but I don't need and I don't want to think about.

SPEAKER_00

The second is attention filtering. So this is also super useful because it has access to everything, and because it has all the content, context, actually, so it knows that, for example, when an email comes, and it's something important or urgent, and it knows from Obsidian what's the context and the background behind it, because I keep everything in Obsidian what's the project, about everything else. So it then can proactively tell me that, I think I have here, so these are three very specific examples that I had recently, that when the system notices that something is important and urgent, it just lets me know. So Netflix payment failure, for some reason didn't go through, was fixed within five minutes when it happened.

SPEAKER_00

Domain renewal coming up. I would probably miss that email, but it picked it up, gave me a message on my Discord, renewed my domain. Emails that can already be, with enough context given about the project, for example, it can already give, read the email, understand what's happening, understand what's already done within the project, and just draft the reply, and it's already in draft folder for me to accept or delete or make some changes. So these are some examples of attention filtering. Execution supports, yeah, so that's draft.

SPEAKER_00

Synthesize is that the inbox, and these are on the right, these are the channels that I have in my Discord, that more or less relate to these types of jobs. So general is where I have everything. I just start the conversations, see where it goes, and if enough times I have a type, a certain type of conversation, I add a specific channel for it. So these are real screenshots from this morning. The inbox is where I just drop links, and it builds the knowledge base for me. Consulting is for the clients and all the backgrounds. It knows all the projects.

SPEAKER_00

It knows all the quotes, deadlines, tasks, next steps, everything else. Video research is for YouTube, for researching what's out there to help me with the next episode. Briefing is for morning briefings. Instagram for social posting. YouTube is for creating the videos. Openclaw is for maintainer stuff. And there's also one playground channel, which it changes depending on day, month, or the need. It's for testing. I usually test a different model, a different workspace, different way of setting up the other important files, like memory and everything else. So I just play there, see what works. If something worked, I promote it. If it doesn't, I discard it.

SPEAKER_00

And all of that works because it's not just a system that has many moving parts that work well together. So LLMs is for judgment, understanding the email, understanding the context, making the connections. Then there are all the files, the tools, the scripts that I have built. The scripts are just like, if this happens, do this. It's done. You don't even need judgment. So LLM is even skipped.

SPEAKER_00

And important thing is also to optimize your memory file, your sole MD file. I have also critical rules MD, because even if I had something in agents MD, or in sole MD, it's still managed to forget something or not do something with critical rules, having critical rules helps, and having it mentioned quite high in the agents MD file.

SPEAKER_00

So that's also an improvement. I went through a few different setups of memory, where I had one memory file. Now that I have the whole memory folder, now we also have dreaming, where we have promoting the memories. So this is important to work on these files, but it's easy to do in OpenCL, because everything is inspectable. These are marked down files, editable. You can look at it, you can read it, you can understand it, and it works well.

SPEAKER_00

What gets harder? Bad memory compounds. If the memory is not set up correctly, and your vault, your nodes, your memories grow to thousands, you're going to have an issue, so you need to actively work on that. Brittle automations, especially when it's 10-step automations, it can break, and it probably will break at some point. So it's again, either split it up into simpler ones, or have some guardrails that are more effective. Noisy nodes, I'm getting rid of them, cleaning regularly, and weak boundaries. So those are all the Sol and D and everything else that the files that are important to optimize for your needs.

SPEAKER_00

So what I want you to take away from this is that, do what I did, and then at some point you realize, yeah, this stuff is awesome, and this stuff helps my life. Start with one recurring pain, grow trust incrementally, build the knowledge base, move everything, or move as much as you can or as you want to mark down files, and start making those connections. System inspectable is easy for you, done with OpenClaw, and optimize for the future you. And this is what I want to close with. So a couple years ago, I had an article about the past me, the present me, and the future me. And the past me is just this completely stupid guy.

SPEAKER_00

He does nothing, he's lazy, he doesn't want to do anything, so now I, present me, need to do everything for that past me. And the future me? The future me is just some kind of god creature. It can do anything.

SPEAKER_00

Build the knowledge base, move everything or move as much as you can or as you want to mark down files, and start making those connections. System expectable is easy for you, done for with OpenClaw, and optimize for the future you. And this is what I want to close with. So, a couple years ago, I had an article about the past me, the present me, and the future me. And the past me is just this completely stupid guy. He does nothing, he's lazy, he doesn't want to do anything, so now I, present me, need to do everything for that past me. And the future me? The future me is just some kind of god creature. It can do anything. It can do anything for me. That creature is all powerful, and if I don't do something today, it's fine that that other creature will do it for me. So, that was the issue, and the job for me is to become friends with the future me, to treat that as a person that I want to help with. And that's the job of the agent. So, I don't need to do as much as I used to, because the agent just helps the future me as much as possible, so that when I wake up tomorrow, it's like as much as could be done, but then someone else other than me is done. So, that's the whole purpose of this setup. At least for me, I don't know, it could be different for you. So, that's what I want to leave you with. Thank you.

SPEAKER_00

Thank you. Thank you. Thank you. Thank you. If something breaks, I just like take one small step back. Fix it. See what doesn't work. Understand why it didn't work. Have a setup that it never happens again. And just like take one step further again. So where it started being more and more helpful and kind of like running my life is when I gave it my knowledge base. So I had a lot of stuff in my Obsidian, which I built up for years. So right now I have like about 3,000 pages or notes, markdown files in my Obsidian. And this is everything.

SPEAKER_00

This is work stuff, personal stuff, tasks, projects, research, what else? Articles. Kind of like an inbox of links that I'm just putting there. And it then finds the connections and puts it in perspective and in context to other stuff that I have. So all of that is now accessible through my Open Claw with a very good search. I have search and memory. I have like normal search. I have QMD search for Obsidian. I have different memory for my workspace.

SPEAKER_00

And all of that is interlinked. And that's where that magic happens. And when I saw recently, that's where it hit me that I probably don't have a simple setup. When I recently saw Andrej Karpatis' tweet that went viral where he says about LLM knowledge bases. I was reading that and it's just like, yeah, that's exactly what I have. Like, what's like super revolutionary about it? And then I understood that. Okay. So I got there step by step. It works for me. So it's probably worth, I don't know, sharing, sharing, telling more about it, showing how it works, showing how you can get to that point as well.

SPEAKER_00

And, for example, for Obsidian, do we...

SPEAKER_00

Yeah, this is how, this is the real screenshot of my vault and all the nodes. And these are different clusters. Some are probably project related, like the big clusters. Some are one-off. These are probably more kind of like bookmarks. And one of the tasks that I'm doing and that I have is that when I add something to inbox, it then takes that link that I add there, looks what's there. It could be a tweet, it could be a thread, it could be an article, it could be a YouTube video, analyzes it, adds tags to it, adds context to it, looks at what's already there on this topic in my vault, how it could be helpful in other areas, and adds connections to it.

SPEAKER_00

So, what previously was just like Twitter bookmarks that you bookmark and you never go back to that. Now it just adds more context, builds up my knowledge base, and is much more helpful, and even surfacing the things for me when I added bookmarks that, okay, so you already had like this and this and this about this subject, and this is how it connects. Maybe you should look at those notes, and very often it's just like, yeah, I completely forgot about that, and that's a good source of knowledge and of thinking about it, because there was a reason why I'm adding this bookmark.

SPEAKER_00

So that's where it's starting to be super, super useful. On top of that, also, at 4am, or like 4am is just like an example that I have, it happens probably between 3 and 6 more or less. So this is what is happening when I'm sleeping. So, when I'm sleeping, again, my agent does everything so that it runs well, it indexes everything, it backs everything, so that worst case, if I lose something, I lose maybe a couple hours of work, of content, of anything else.

SPEAKER_00

So, it refreshes all the indexes for QMD, for memory, for my Obsidian Vault, and I start fresh in the morning with whatever waits for me, maybe summary of the emails, of the calendar, everything updated, the latest OpenClaw version is waiting for me, which also took like step by step. I have some scripts around it, so that it knows what to do and what not to do when updating, what can break, why it breaks, how to verify it before updating or before restarting your gateway, so that it is able to come back online again.

SPEAKER_00

So, that all is also automated and as I get up, it's already waiting for me, fresh and ready for me to start the day. And each OpenClaw is, like, I'm not a big fan of sharing my exact setup, because that exact setup is very specifically for me, for what I need right now, for what I will need in the near future, for the errors that I encountered, for issues that I want to be solved. But to give you some ideas so that we can talk more also about specifics and not just like in general, so these are some like five areas or five types of jobs that my agent is doing.

SPEAKER_00

So, the first one is, like, I want to talk more about updating. The first one is, like, I want to talk about the first thing. So, this is what I just showed you. So, it does all the updating, it does all the plumbing, it does all the stuff that needs to happen, but I don't need and I don't want to think about. The second is attention filtering. So, this is also super useful that, because it has access to everything, and because it has all the content, context, actually, so it knows that, for example, when an email comes, and it's something important or urgent, and it knows from Obsidian what's the context and the background behind it,

SPEAKER_00

because, yeah, I keep everything in Obsidian what's the project, about everything else. So, it then can proactively tell me that, I think I have here, so, like, these are like three very specific examples that I had recently, that when the system notices that something is important and urgent, it just lets me know. So, like, Netflix payment failure, for some reason didn't go through, was fixed within five minutes when it happened.

SPEAKER_00

Domain renewal coming up. I would probably miss that email, but it picked it up, gave me a message on my Discord, renewed my domain. Emails that can already be, with enough context given about the project, for example, it can already give, like, read the email, understand what's happening, understand what's already done within the project, and just draft the reply, and it's already in draft folder for me to accept or delete or make some changes. So, these are some examples of, like, attention filtering. Execution supports, yeah, so that's draft.

SPEAKER_00

Synthesize is that the inbox, and these are on the right, these are the channels that I have in my Discord, that more or less relate to these types of jobs. So, general is where I have everything. I just start the conversations, see where it goes, and if enough times I have a type, a certain type of conversation, I add a specific channel for it. So, these are, like, real screenshots from today morning. The inbox is where I just, like, drop links, and it builds the knowledge base for me. Consulting is for the clients and all the backgrounds. It knows all the projects.

SPEAKER_00

It knows all the quotes, deadlines, tasks, next steps, everything else. Video research is for YouTube, for researching what's out there to help me with the next episode. Briefing is for morning briefings. Instagram for social posting. YouTube is for creating the videos. Openclaw is for maintainer stuff. And there's also one playground channel, which it changes depending on day, month, or the need. It's for testing. I usually test maybe a different model, maybe a different workspace, different way of setting up the other important files, like memory and everything else. So, I just play there, see what works. If something worked, I promote it. If it doesn't, I discard it.

SPEAKER_00

And all of that works because it's not just a... It's a system that has many moving parts that work well together. So, LLMs is for judgment, like understanding the email, understanding the context, making the connections. Then there are all the files, the tools, the scripts that I have built. The scripts are just like, if this happens, do this. It's done. You don't even need judgment. So, LLM is even skipped. And important thing is also to optimize your memory file, your sole MD file. I have also critical rules MD, because even if I had something in agents MD,

SPEAKER_00

or in sole MD, it's still managed to forget something or not do something with critical rules, having critical rules helps, and having it mentioned quite high in the agents MD file. So, that's also an improvement. I went through a few different setups of memory, where I had one memory file. Now that I have the whole memory folder, now we also have dreaming, where we have like promoting the memories. So, this is important to work on these files, but it's easy to do in OpenCL, because everything is inspectable. These are marked down files, editable. You can look at it, you can read it, you can understand it,

SPEAKER_00

and it works well. What gets harder? Bad memory compounds. If the memory is not set up correctly, and your vault, your nodes, your memories grow to thousands, you're going to have an issue, so you need to actively work on that. Brittle automations, especially when it's like 10-step automations, it can break, and it probably will break at some point. So, it's again, either split it up into simpler ones, or have some guardrails that are more effective. Noisy nodes, I'm getting rid of them, cleaning regularly, and weak boundaries. So, those are all the Sol and D and everything else that the files that are important to optimize for your needs.

SPEAKER_00

So, what I want you to take away from this is that, like, do what I did, and then at some point you realize, yeah, this stuff is awesome, and this stuff helps my life. Start with one recurring pain, grow trust incrementally, build the knowledge base, move everything, or like, move as much as you can or as you want to mark down files, and start making those connections. System expectable is easy for you, done for with OpenClaw, and optimize for the future you. And this is what I want to close with. So, a couple years ago, I had an article about, like, the past me, the present me, and the future me. And the past me is just like this completely stupid guy.

SPEAKER_00

He does nothing, he's lazy, he doesn't want to do anything, so now I, present me, need to do everything for that, like, past me. And the future me? The future me is just like some kind of, like, god creature. It can do anything. It can do anything for me. That creature is, like, all powerful, and just like, if I don't do something today, it's fine that that other creature will do it for me. So, that was the issue, and the job for me is to become friends with the future me, to treat that as a person that I want to help with. And that's the job of the agent. So, I don't need to do as much as I used to,

SPEAKER_00

because the agent just helps the future me as much as possible, so that when I wake up tomorrow, it's like as much as could be done, but then someone else other than me is done. So, that's the whole purpose of this setup. At least for me, I don't know, it could be different for you. So, that's what I want to leave you with. Thank you. Thank you. Thank you. Thank you. Thank you.