You Need An AI Operating System (AIOS). Here's What That Means.

So at this point in my trajectory here with this AI content and you know coming from like the web side of things, I am I am at this point with my business and my clients businesses and the people that I

talk to and the content that I consume on you know YouTube and the colleagues that I talk to. I am pretty deep in this stuff at this point and there are still some problems and I think there's always

going to be problems but like there are specific problems that are still kind of plaguing me and I feel like I am you I always feel like I'm behind but I feel like I'm still objectively very far ahead of like the average person when it

comes to understanding how all this stuff works, the terminology, the the number of reps that I put in, right? And

uh it's it's just kind of it's kind of interesting. So, in this video, okay,

interesting. So, in this video, okay, and I'll be completely transparent and honest with you. I've been trying to kind of I wanted to make this video. I

want I told myself and I told my community and everything like Modern Builder Society. I told everyone that I

Builder Society. I told everyone that I was going to make a AIOS this this week.

And I got that that term initially. I

heard it from Liam Muley, large YouTuber, uh, very successful in the AI space. I got that term from him and I

space. I got that term from him and I think it was because he kind of went against the open claw stuff and then I realized it after I was putting all this together. I'm going to show you some

together. I'm going to show you some stuff on the screen here shortly. Um,

and I'll have some things for you to kind of like take on take on your way and and uh you know the links will be in the description to kind of like do some of this auditing for yourself, for your own business, for your clients, do

whatever you want. But the problem is actually really simple and they constantly crop up in conversations that I'm talk that I talk about to people all

the time. The the problems are context

the time. The the problems are context like not even context bloat or context rot. It's the fact that your context

rot. It's the fact that your context lives in so many different places. If

you're trying to use chatbt or you're trying to use claude or you're trying to use claude code or what have you, maybe lay people don't have or non-technical people like they don't have as big of a problem because they're just in their

chatbt and they just remember stuff about them, but then they don't have the ability to create as many things. They

can create it in the app, but they don't have like the the the barrier there is I was having a conversation this morning and it's like the barrier there was like okay well you have all these other apps, right? and they don't have AI built into

right? and they don't have AI built into them or maybe the AI built into them, you know, the models that they have or whatever are not as good or they're they're just, you know, it's it's it's just nerfed in some sort of way, but you

have all this inference in Chad GBT.

Like you could connect it, right, using the little app thing or in Claude, you can connect it using the connectors.

That's a part that's missing and it's a part that OpenClaw and like Hermes agent now and all these things that you can't keep up with try to solve. But really

for me, I'm more interested and you can see a video on my channel like a few months ago, this idea of like a context system.

And that is the thing that I'm trying to navigate right now and just try to put together for myself literally, you know, in real time. And that's the part that slides into this whole system that really makes or breaks it. So, what I

want to do is I want to show you just kind of a I want to show you two diagrams because it evolved since I wanted to make this video like two days ago and then now. and I'll just like walk you through this. Now, what I would

recommend is this isn't a tutorial. This

is you trying this should be you trying to like understand and wrap your mind around these concepts because these are going to be very fundamental moving forward. I would just asterisk it with

forward. I would just asterisk it with we're very early. Like there's probably going to be products that end up making this easier. There already are. And

this easier. There already are. And

that's kind of what I've started to find and I'm gonna start trying to use them.

But I I still think that even though we're utilizing like AI agent products at some point, we should still understand how to set them up to be as

optimized as possible. And I think this video right here is an example of how we can think about it rather than just jumping

in and like, you know, like setting it up, so to speak. Uh because there's a lot of tutorials like that. I they're

great. Cool. But I also like I just like this. So hopefully you do too and uh

this. So hopefully you do too and uh we'll find out together if this is uh if this is helpful here. So I just want to run through this as quickly as possible.

So the concept of like an AIOS, so an artificial intelligence AI operating system for your business, your life, multiple businesses, whatever. For me, I

think my life and my businesses are so intertwined that I would want basically I would want it segmented, but I'd also want it to all be kind of in one thing.

And this is I'm going to say this later, but like think about your life and the information that you have to process every day as like a big ball of context

that needs to be at all times accessible and connected to all different things and organized to a decent degree. I

don't think a lot of us have that. It's

all just scattered in a bunch of different places. And there's different

different places. And there's different ways to deal with that, but we'll get into it. So I want to show you two

into it. So I want to show you two different diagrams and I'm going to split it here after we get to kind of like this uh thing because I'm going to show you at a high level first. But the

thesis is like one house many doors. You

own the house. The model is who works like inside. And then this is the key

like inside. And then this is the key thing here is I do believe this and there's a and there's some some more to talk about here but ultimately inference is a commodity. So inference is the

actual when people say the word inference that's the LLM doing the work, doing the thinking, right? So Claude

doing the thinking or Chad GBT or you know GPT or whatever doing the thinking or Gemini doing the thinking that's like the inference part that's the part that is really becoming obviously everybody's like oh this is the best no this is the best but it's really becoming

commoditized like that's the part that's going to get changed and slotted out especially if prices go up and then you have to get like local L local AI and a local LM like doing your stuff that is

the part that you want to be you definitely want to be modular because everything else is like I'm not going to call it proprietary but it's more specific to you and you'll understand what I mean in a second. So just I want to show you this as a high level and

then I'm going to switch to the other diagram because it made I made some changes but the architecture. So you

know this is the this is like if you zoom out what do we really need here? We

need a way to interface with this technology whether that's phone voice web UI or like a coding environment. We

need the inference right again any model it's swappable right the maybe you've heard of some of these names before.

Then we need like in this case the folder or like this is like it's kind of bigger like this is like the context right of everything we're doing here. So

there's context skills execution. I'll

talk more about this stuff. There's some

sources down here. Uh, and then there's like different modes for things that you kind of already use. So, I want to talk about all this, but the reason I'm I'm flying through it real quick is because up here, I was getting ready for this and I was like, wait a second. I've had

this I had this hourlong back and forth with Claude to kind of like massage this idea and think about it and have it do some research for me and all that. And

then I was like, this is where I got I was like, okay, cool. And then I came back to this and I was thinking, you know, I have very limited experience with Open Claw because I just, you know, didn't I was doing other things when it was hyped. And then I know there's

was hyped. And then I know there's Hermes agent now because I'm subscribed to everyone. I just don't have time to

to everyone. I just don't have time to check out every single thing. But like

I'm very good at like understanding what this stuff is supposed to do and like what I want it to do. I just haven't like gone and played with those and maybe you haven't either. So this is if I can understand this mental model, I

can either build it for myself or make sure that I know the tool that I choose can actually do this, right? And then I thought, well,

this, right? And then I thought, well, wait a second, let me just go back into that claw conversation that I was having and then let me see like, hey, go I was like, literally go look at openclaw, go

look at Hermes agent and see how close they are. And it basically said like,

they are. And it basically said like, yeah, it's pretty close, bro. Um, but it is it is more of like a it's more of a of a layer to this. So then just the quick the quick change here, right?

Pretty much the same here, but look how different this is. So we have the interface still like we still need to interface somehow because that's kind of part of it, but like there's always going to be like I've had so many issues now. Like all this stuff just comes from

now. Like all this stuff just comes from personal experience, right? It's like

I'm out at uh you know at an event and I only have my phone on me and it's like I want to I want to do something.

Obviously if you've seen OpenCloud and [ __ ] everybody's using Telegram, everybody's using all this. Obviously it

exists, but you still need to understand your use case. Okay, I want to be able to go from my phone, but I also like there's there's there's problems there, right? It's like do you

right? It's like do you like does every is everything still synced? Do you still have it's it's kind

synced? Do you still have it's it's kind of like this like you have chat GP on your computer and you have chat GP GPT on your phone, but do you do you still have all of the uh you know, do you

still have all the apps connected? Do

you still have all the access to all the content that you need? In that case, maybe a little bit of a bad example, but this this example actually happened to me today. I was doing something on cloud

me today. I was doing something on cloud code and VS Code right here. I wanted to do remote control. So, I remote controlled and then I went upstairs and I was like on my phone and I was like, there is no way that I'm going to be able to see the image that I wanted to

create with this thumbnail generator that I have on my local machine down here because I don't have access to all the files in the directory that is on my local machine. So, you know, I had to

local machine. So, you know, I had to think of a creative way for it to like save it somewhere else so I could see it on my phone, right? as an example. If

you don't know what I'm talking about, it's fine. The point is that you need to

it's fine. The point is that you need to be able to interface with these things in different ways and still have a similar experience where you have the whole the whole gamut of everything. I

would also argue there's different modes like kind of that you would want to do, right? So like voice, you don't want to

right? So like voice, you don't want to be able to you don't want to have to just type. You want to be able to do

just type. You want to be able to do that. That's, you know, that's kind of

that. That's, you know, that's kind of like a little bit of an offshoot because you could always like voice into whatever, you know, interface you were dealing with. But that's one thing. A

dealing with. But that's one thing. A

big thing here is I don't like manually moving [ __ ] around or managing tasks anymore in like a traditional SAS sense.

I would much rather just orchestrate.

However, I do want to see it though cuz if I don't see it, I'm basically blind.

And I don't like being blind. I just

don't want to do all the manual clicking and work. So that's an important thing

and work. So that's an important thing and that's a big distinction. So like a mission controller, like a web UI for these types of things. and then

potentially a coding environment because here listen like you can orchestrate code all day from a [ __ ] Telegram chat thing but but ultimately like I still want to be able to see what's

happening because that's that that part of it. I mean I don't purely vibe is

of it. I mean I don't purely vibe is kind of my point there. If you do that great, great for you. But like I still want to be like I don't need to I'm not going to read all the syntax but I need to understand where things are, how things kind of interact. And it's just a

better experience to me like when you're actually doing that because you still need to be able to test things and actually see if stuff works, right? So

there's a little bit there, right? Um

let's move into the harness. So this is the biggest thing that's changed.

This whole thing was added because what Open Claw kind of pioneered in a sense, I mean Claude Code did it and it it still works, but what OpenClaw started to pioneer and now Hermes has kind of

done and these are all kind of like in here. This word harness is the harness

here. This word harness is the harness isn't the model. It's not the inference.

It's like the thing that orchestrates the model a little bit and adds maybe like tool calls and all this other [ __ ] that like kind of gets wrapped together, right? And cloud code obviously does

right? And cloud code obviously does that and cursor does that or whatever.

And open call and Hermes are like kind of in the sameish realm with that. And

you know there's some just some interesting stuff down here just some other constraints like harnesses everything wrapping the model orchestration tools memory state all that sort of stuff. Um, but here's the thing is like on certain things like

this is this is why it's sketchy to go outside of like claude code for instance if you want to use like opus 47 for instance, right?

Because it's been the big one for a while now. Chat GBT 5.5 is pretty good.

while now. Chat GBT 5.5 is pretty good.

GBT 5.5 is pretty good. But the thing is if you want this part I got to skip to here for a second to kind of talk about these together. If you want everything

these together. If you want everything to be super swappable, I mean, you could use the APIs for everything and it'll be super swappable, but it'll be more expensive, at least for now, because the the subscriptions are highly subsidized.

So, it's like if you're subbed to these things and you don't want to use the APIs, then you have to kind of play by the rules. And right now, Anthropic

the rules. And right now, Anthropic doesn't really let you do that [ __ ] Like there's been so much BS that I've watched and heard about and and uh and you know listen to on you know these uh

YouTube channels and the news and everything that if you take your subscription and you go to OpenClaw like you could have like your account like you know uh canceled basically um and banned. So those are things to worry

banned. So those are things to worry about and be like slightly concerned about because they obviously don't want you to be using a different harness right other than other than cloud code.

Now I don't know if that's going to change in the future or what the case is but it's something to consider. So

that's why I was like kind of steering away from some of those things because I I have had a great experience in cloud code and you could build everything in there. However,

there. However, they've they've added Anthropics added a bunch of cool stuff to make that easier like remote control and there was like channels and like dispatch for co-work and all this other [ __ ] but it's still

it's still a coding environment and that's not exactly what I want. It's it's very you know you can customize it and everything like that. But now I'm like thinking and as I

that. But now I'm like thinking and as I as I was preparing for this video, I was like maybe I should try I should probably try Hermes definitely and Open Claw maybe a little bit more and stuff.

But those even these these three layers up here like they're important for sure.

But honestly, and this is what I can't believe like I don't hear more about this. I'm sure it's out there, but like

this. I'm sure it's out there, but like I I had this idea. I I'm not saying I came up with it, but I had this idea like months ago and the video didn't do very well. probably wasn't packaged very

very well. probably wasn't packaged very well, but it's called like a personal context system. And this is the [ __ ]

context system. And this is the [ __ ] part that's so important. And if you're trying to explain this to like a random business owner, like this is

the most important part. Like without a doubt. Obviously, all the other stuff

doubt. Obviously, all the other stuff does the work, but this is the this is why this is important is because you can have the best interface, the best harness, and the best inter in um

inference, and your output is still going to be [ __ ] [ __ ] Because if you don't like if you don't tell a business owner, imagine you're trying to set this

up for a business owner, right? And you

you give them all the tools, all the inference, and all the stuff. If those

tools and that whole system above this doesn't have the context for that business like deep understanding of that context, the that's where you get AI slop because it has nothing to go off

of. It has no data about that business,

of. It has no data about that business, the processes, the you know the decisions of everything like the current state of the business, the you know the methods, the playbooks, how to do things like all that sort of stuff. If it

doesn't have any of that, then it's not going to it's not going to do any of the any of the work that you need it to do.

So, this is the part where like I'm actually struggle to like educate on this because I think it's so so simple but yet so difficult to actually get through to people. It's like listen, you're running a business. It's

relatively successful. Okay, great. It's

a local service business. Whatever it

is, do you have any of your [ __ ] documented? No. Okay. Well, that's

documented? No. Okay. Well, that's

literally like basically our entire job right now. Like all the other stuff is

right now. Like all the other stuff is being solved in so many ways and we could just plug and play things if we wanted to. the like the role of people

wanted to. the like the role of people potentially in this space right now is literally to be content context gatherers in a sense because it it's

context gatherers like wranglers and organizers.

Now you don't have to do all the work but you have to know how to get that context and extract that out of a business because this layer is the most important layer to defend against AI

slop in the thing. And if you if you output AI slop for yourself or for another business, that's that is literally categorically not a good result. So that's the whole thing. And

result. So that's the whole thing. And

this is why like I think sometimes why I've been so heavy on this for so long.

I've been like documenting my life. I've

been uh you know timeline of all these events and like this whole identity stuff. Again, go watch that other video

stuff. Again, go watch that other video because I went deep into how I did that.

I'm going to have to like re re orchestrate a little because it wasn't perfect then.

But this is the most important part and let me break it down a little bit. So

again, it's literally just like markdown code like non-proprietary stuff like you know yours forever like this is all about you or your business or your client's business whatever. So like

context the con and I and again I kind of mental modeled this but maybe it maybe it changes maybe there's better ways but like again context is the interpretations the decision like the current state like the actual I thought

of that as like literally like words like what does your business do um like who you are things that you've done current like current state of that stuff the skills are like the methods and the

processes the playbooks and how to do things so those are again skills from like a MD perspective So context would be uh you know we are a roofing company.

Skills would be okay. How do we obviously not how to how to put on a [ __ ] roof but like how we take in leads or something right? And that would like be a process. And then the execution would be the like this like

the actual code and the workflows and the connections. We'll talk about that

the connections. We'll talk about that in a second. But those are the things that actually execute that stuff. It

would take context. It would take the process and it would actually do the stuff. Now obviously from it would be

stuff. Now obviously from it would be already written so it would kind of just like it kind of took a little bit of that before and then you and then you're running it through the code. But

obviously if you need to infer a little bit more if you need to like if you have an executing like a script that calls out to an LLM you could pass more information in from those skills or from that context. That's the kind of the

that context. That's the kind of the idea. Now again in my case specifically

idea. Now again in my case specifically here like this is how I would compartmentalize all of this context though right because you'd have maybe one big folder called AIOS. I'm I'm just

you know as an example and then but you still want this stuff to be organized because organizi there's got to be a better way to say this but organized but also um related is the best way I can

say it right now. Like for me personal Samansky media and motive 11 personal is just like my life me who I am identity timeline maybe seamansky media is my

content advising business the community like modern bu modern builder society all that stuff and then motive 11 is everything that I do for clients and service provider stuff there three kind of distinct things they all have their

own context they might have all their own skills and execution but at any time this is the [ __ ] key okay at any time if I'm if I'm operating in all of these things like kind of like in fluid

fluidity in a way, right? Maybe you

don't want personal on yours, whatever.

Okay. But like if I I operate literally I'm in one second I'm doing something kind of like personal whatever maybe that has a part of that or needs part of that. Then I'm doing like some content

that. Then I'm doing like some content stuff then I'm doing um uh you know client stuff. If I'm doing all of that,

client stuff. If I'm doing all of that, I need to have access to all of that context, all those skills, all that execution all at all times because at any time I am doing something there, I

need it to be able to execute the way that I wanted to do with the with the right things. And this is exactly my

right things. And this is exactly my problem that I've ran into. And even me again, somebody that's like been doing this now for like very heavily for the last like several months, I still have

these problems where like [ __ ] like man that's that's over in Claude for some stupid reason because I just wanted to spin up a quick thing. I don't want to be in chat interfaces honestly ever again. I don't want to be in claude or

again. I don't want to be in claude or in chat GBT. I want to be in my environment that I've crafted that's ultimately is leveraging chat interfaces

in some way kind of but it's not like the version one simple thing anymore. It

has evolved into something that is highly contextual, highly agentic, and is incredibly efficient like in all these fronts because I don't I don't

need to like switch context or anything and it and it constantly updates. And

again, the reason I say that real quick is because that was a that's a real big problem is like, okay, okay, let's say Mark, you you put in all this context today. Well, what happens in two months?

today. Well, what happens in two months?

You have to update it all. That would

[ __ ] suck. So I hope that that's not like Hermes I know is like one of the agents that uh is is more focused on that like self self-improving self-updating. So that's I guess just

self-updating. So that's I guess just got to look into it I guess more. But

still though the the concepts here are all things you have to think about if you want to really move forward in this I don't know I feel like in this world in this line of thinking with all this

right okay so last thing I want to wrap up here is sources right because what haven't we talked about and again this is personal to me but you would you would do whatever you want here I'll give you a resource in the description

um for to to kind of like kind of create something similar and audit yourself if you want to play around with it just throw it in AI and answer the questions preferably with your the sources. What are these? Well,

the sources. What are these? Well,

here's the problem. Okay, what's the last piece of this puzzle? The last

piece of this puzzle is that we have stuff to do work. We have the inference, we have the harness, we have all that stuff. We have the context. Now, we took

stuff. We have the context. Now, we took a lot of time to put all that together.

Well, we don't but we're still going to need certain tools. Like, you could probably replace some of these tools potentially, and I talked about that in LA, the very last part here, but like just imagine like do a do an audit real

quick in your mind of like all these tools. Like, look at these tools on the

tools. Like, look at these tools on the screen. Like, do you use any of these?

screen. Like, do you use any of these?

I'm sure you do. How often are you like clicking into like your bookmarks or wherever and you're going to these different things and you're manually doing [ __ ] in there? Probably pretty

often. I don't want to do that anymore.

I you don't really have to. Like you you you could agentically do probably at least 50% of the [ __ ] that you're already doing if you had all the context, you had the skills, you had you

had the execution of all that stuff.

It's not it's really not rocket science.

It sounds confusing and I'm oversimplifying some of it, but it's not. It's very possible. So via like MCP

not. It's very possible. So via like MCP or just direct API because everybody shits on MCP. The point is you connect these these things and if you even if you've played around with like simple cloud or simple chat GPT at this point

that's just the chat interfaces. It

already exists. The concept again already exists. It's just not bundled

already exists. It's just not bundled all together.

It's these connectors and these apps that you could connect in there. And

then if you deal with you know OpenClaw Hermes I'm sure I again I don't have the experience but like I'm sure that it's again simple. It's like, hey, we're

again simple. It's like, hey, we're going to go connect to the API for Gmail. Okay, let's get a Google API.

Gmail. Okay, let's get a Google API.

We're going to go connect to the API for, I don't know, Aonic. Let's get an API key. Right? And then now you can

API key. Right? And then now you can push these things through and you can do these different things from all the way down the chain because now the only thing left is like the only thing left to decide here, right? Imagine if you

have all these things connected now to your whole system. Okay. Okay. Well, you

still need to like post stuff and do things and and edit stuff and like, you know, you know, communicate and check like again like Stripe as an example.

It's like you don't have the updated context and you shouldn't have it in your like folder so to speak of like your revenue, but you want rec you want like thoughts. You want like updated

like thoughts. You want like updated information every morning on how you're doing or something like that or on your YouTube like you know I want to know like hey how do I have any comments I need to reply to? Do I have all this stuff? Like that's what I'm talking

stuff? Like that's what I'm talking about there. The last part of this is,

about there. The last part of this is, you know, this is where you get a little a little deeper here and you got to make a decision. Is is this good enough for

a decision. Is is this good enough for you or are there things and this is again I would say I don't know take this with a grain of salt but like there's like three modes for like tools that you

already use. So all of those things up

already use. So all of those things up there in that thing it's like you got to make a decision and I would start with just connecting to most of them but like are you going to connect to them? Right?

There's things that I think that I don't want to say we're never going to get rid of, but this whole time has been crazy, so I'm not going to make any crazy predictions there, but I don't think we're going to get rid of like Gmail or email like anytime necessarily soon. I

mean, obviously some people have already gotten rid of email, but like if you use it, you can't like really replace it.

You know what I mean?

YouTube, I'm not going to get rid of YouTube. I'm not going to get rid of

YouTube. I'm not going to get rid of Stripe. I'm probably not gonna get rid

Stripe. I'm probably not gonna get rid of Circle, right? I'm not going to recreate those things to to do that.

there's like a risk there's like a you know costbenefit analysis that carve out so like pull out mechanical operations out of the tool and and put that into the execution layer. Well there's

certain things that like I might not actually have to utilize that heavily anymore. I'll give you perfect instance

anymore. I'll give you perfect instance is this one right here. I haven't made a video on it yet but I've talked about it. I used to use Da Vinci Resolve all

it. I used to use Da Vinci Resolve all the time to edit my videos. I still do but I've already made a flow again in cloud code without actually writing any of the stuff just orchestrating it. I've

already made it so I could give it like an hourong video and it'll it'll know using ffmpeg and other libraries where the silences are and it'll it'll you know it'll transcribe and it'll do all this stuff and it'll it'll know where

the silences are and it'll cut it out and it'll give me a 32minute video, right? So, you see what I'm saying? Like

right? So, you see what I'm saying? Like

there's certain things that if you're not using 95% of the tool, right? Like

down here, right? If if almost nobody these SAS products are created, this is when people say SAS is dead. It's it's

that's too simplistic of a take, but this is kind of an example, right?

Almost nobody uses 100% of any tool. The

interesting work is identifying which two to 15% again maybe more depending on what you're doing you actually use and then deciding which model here it fits into. So it's like are you just going to

into. So it's like are you just going to connect the whole thing because it's kind of like a necessity which there will be some stuff or are you going to try to get rid of you know some of it you know maybe carve out certain things of it right? So like maybe maybe certain

things again like I don't have to use Da Vinci as much as I did before maybe to clean up some things. um bento like I could send things to there and then just clean them up for through there and then send some stuff. It all also depends on

what you can do through APIs and MCPS and all that or the last one absorb it like do I really need notion like do I you might maybe you have a big team or

something like that it's a it's a good platform I saw Theo T3G like talking about it the other day or whatever like it's a good platform cool whatever but do you really need it? I I don't know if you really need it. It depends on your

situation, but for me, I'll talk for me, I don't think I need it because all the information that I have in there, it doesn't get like it's kind of a just like a cold storage for information that

like I would love to have just in my my context and that I could just constantly be updating with my voice or with any sort of, you know, any sort of harness that I choose

or any sort of interface, right? So,

that's the kind of the idea, right? And

then close is like a CRM thing that's kind of like meh. decent platform again, but like it's like it's something that like doesn't really need to be there.

It's not providing like a ton of value.

And I'm not picking on those two. I'm

just saying like that's that's the idea, right? So, I think with this framework,

right? So, I think with this framework, okay, I hope you understand a little bit more about like when you hear the term AIOS, this is it's mostly kind of what is kind of going on. Anytime you hear

any sort of like again agentic stuff, you have to just realize that there's a lot of people that started like grabbing this. This is a little bit of an aside,

this. This is a little bit of an aside, but there's a lot of people that started doing this open claw thing and I don't think they really understood like kind of like all the pieces that go go through it. And I'm just one of those

through it. And I'm just one of those people like to really break this down so I understand it so I can make some good decisions. And I think if you understand

decisions. And I think if you understand this in this way, and I'm not saying this is perfect, but this is like, you know, largely I feel like how how things are connected here, then you can move in

a direction where like you make like just better decisions in certain in certain areas because you understand like kind of what's going on. I made

another video like this. If this was too deep, hopefully it wasn't. Uh, I made another video like this uh on the channel. It's called AI. Like, if you

channel. It's called AI. Like, if you understand, if you watch this, you'll know more about AI than 99% of people.

So, go watch that one, too, if you don't know all the stuff. I went down deep rabbit hole and all the terms and everything. But, I'm going to make this

everything. But, I'm going to make this one kind of like a a mustwatch on this next evolution of like an AIOS and a whatever you want to call it, your your AI assistant, agentic thought process

and and diagram and all that sort of stuff. I'll figure out the packaging.

stuff. I'll figure out the packaging.

But the thing I want to leave you with is I wanted to reverse engineer something because I think that one thing that could be valuable here is if you think about this and you have this, I'll

give you this as well. But if you uh if you go in the description, there'll be a link to grab this. I'll give you this framework, this this diagram and everything like that so you can look at it. The other thing is a skill that you

it. The other thing is a skill that you could load into. Again, a skill is a markdown file. Okay? So realistically,

markdown file. Okay? So realistically,

there's nothing super special about skills like inherently. It's the the information in there that will then drive the instructions so to speak for like the like whatever LLM you're giving

it to. Right? So, I'm going to give you

it to. Right? So, I'm going to give you this markdown file. Okay. When you

download it, take it into if you want to take it into chatbt or cloud, it'll work. You could also take it into like

work. You could also take it into like your coding environment in a way and pull it in, you know, via the path of the URL or whatever. It's not important.

Somehow get it into whatever AI you want. I would recommend doing it or you

want. I would recommend doing it or you could just read it and just like think about it yourself. You just want to like, you know, go outside under a tree and just read it or something. uh go

back to you know before this time. But

what it'll do for you is when you if you do it with AI, you'll be able to answer all the questions and then you'll be able to kind of like audit your own business or your client's business or change it and modify it however you

want. And it'll kind of give you

want. And it'll kind of give you something similar to this in a sense of mainly around the sources and kind of like what context you should focus on, you know, bringing to the surface and

kind of feeling all that out because that this is the I cannot underestim under understate how important it's going to be to have the context of everything because if this stuff gets

really good as it it's already good, but as it gets more and more if like the power gets better, you are going to need to prevent against slop even more. And I

don't even like that term because I feel like most of the stuff comes out half decent now. But still though, the the

decent now. But still though, the the differentiator is you. It's the

information that you're putting in. It's

if you if you are still kind of blackpilled on like AI, it's incredible at synthesis and analysis when you give it the information. when you don't give the information, it's got to it's got to

be it's even it's it's always probabilistic, but it's even more probabilistic and it just starts making guesses like even like even more guesses than it than it does even when you give it [ __ ] right? But every single time

that I've ever given it a lot of context about these things that I'm trying to do and it has access to everything, it's a way better experience and the product is just unbelievably better that that comes out of it, right? The output. So, um go

down there, grab that link in the description if you like this type of content. If you're a business owner and

content. If you're a business owner and you watch this, appreciate you. Thank

you for being here. Um, there's a link down there to work with me. We can talk about this. We can talk about other

about this. We can talk about other implementing other things like this in your business. If you want to navigate

your business. If you want to navigate this whole transition and this whole AI thing with other people, I've got a community. It's called Modern Builder

community. It's called Modern Builder Society. There's a link in the

Society. There's a link in the description for it. Uh, we got bi-weekly meetings in there. Uh, great

conversation. Uh, there's exclusive like resources and content in there as well.

So, um, I appreciate you as always if you give me even a second of your attention for any of these videos that I make. I thank you so much. Um, keep

make. I thank you so much. Um, keep

going. This is uh just going to continue to get uh crazier and crazier. But

honestly, again, I think that something like this is like really the next step for me. I think it's the next step for

for me. I think it's the next step for you as well if you're kind of dabbling in this because this is where it starts to actually get very productive and impactful. When you have all these

impactful. When you have all these things interconnected, everything else has been kind of cute to this point and it's been helpful. But this is this is the next level. This type of thinking, I'm not saying it's going to stay like

this forever. It's going to continue to

this forever. It's going to continue to evolve. But if you can think like this

evolve. But if you can think like this in this video and you can put these pieces together, uh it's it's going to be good for you and your business. So, I

appreciate you. Thank you so much. I'll

talk to you in the next one.