🧠 How I'm Building an AI-Native Business (not just talking about it)

Yooo!

I’ve been thinking a lot about the evolution of your operating system, during the ai era we’re in right now.

Because I don’t think it’s ever just one tool, or one skill or a Mac Mini.

It’s more like… you build something, you test it, new conditions and data changes, you gather better information, and then you adapt.

That’s what I keep coming back to.

And I think that’s where a lot of business owners are stuck right now.

You open social media and it’s just too much.

It’s always a new tool, new agent, new model, new workflow, or claude killing designers for the 4th time this week…

Everyone telling you that you’re behind. Everyone making it feel like if you don’t use this thing today, you’re cooked.

So you keep watching.

You keep saving posts.

You keep collecting information.

But the business doesn’t actually change.

and for some people it even gets worse by how distracted by ai noise they are.

Same messy system & bottlenecks.

Just more AI sitting on top of that messy ass business.

As always the raw brain dump — HERE (my real thoughts straight from Wispr Flow)

I had a call this week with these lads from Melbourne, Ethan and Zach.

Zach messaged me on IG because his co-founder was apparently a Notion nerd, and as soon as I saw the word Notion, I was like yeah, okay, we should def chat.

It was meant to be quick. 20 min max.

It turned into 90 minutes of screen sharing, nerding out, showing each other how we run our systems, and giving tips back and forward.

And that was the thing that made this whole thought click for me.

Because it wasn’t just a Notion call.

It was me seeing another version of the same problem I’ve been trying to solve.

Soooo… How do you build a system that still works when the conditions change?

When it was just me, my setup made sense.

It was built around my brain.

I knew where things lived, what mattered, what was sensitive, and what I had to do next.

But that only works until the business starts moving faster than your own head can hold.

Then context starts living everywhere: Slack, WhatsApp, random calls, docs, databases, voice notes, half-finished thoughts. Don’t get me wrong, Slack and WhatsApp have been amazing and they serve their function, but far-out, context can fall through the cracks. As simple as that.

That’s what I mean by new conditions.

The system wasn’t wrong.

It was just built for the old game.

Now the game is different, so the system has to evolve with it.

This is where people get AI backwards: if the operating layer underneath is messy, more tools and models just help AI move through the mess faster.

sooo the Obsidian thing has been on my mind.

A lot of people are obsessing over Obsidian, backlinks, & knowledge graphs.

And it is really cool. If you’re solo and you want to organise your thoughts, I get it.

But if you own a real business and you have a real team, it’s a different problem.

A business is tasks, projects, clients, deliverables, deadlines, people, permissions, and context, and a bunchhh of other stuff as well.

All of those things need to actually move together. With both humans & ai interacting with it.

That’s why the big one I learned from speaking with Ethan and Zach is relations with databases, not just backlinks.

A backlink is like: these two ideas are related.

Butttt

Relations with databases are like: this task belongs to this project, this project is tied to this client, this client has this deliverable, this deliverable has a deadline, and this person owns the next action.

That’s a very different thing.

One helps you think.

The other helps the business operate.

And that is what I’m trying to build now.

Not a prettier way to store notes.

A better way for the business to move.

The cleanest mental model for me became frontend and backend.

It sounds technical, but it’s actually simple.

The backend is where the real context lives: projects, clients, docs, deliverables, etc

The frontend is what a human actually needs to see to do the job.

That clicked for me because the problem wasn’t that we didn’t have enough information.

It was almost the opposite.

There was too much information, and not enough clarity.

My Head of Growth needs growth context.

My AI dev needs build context.

Editors need the brief.

I need the CEO view sometimes.

I need the execution view other times.

They should not all be the same thing.

That was the wow moment.

It’s not one big dashboard.

It’s not one perfect workspace.

It’s not making Notion look pretty for the sake of it. (i hate habit trackers and colour coded time blocking)

It’s the right context showing up in the right place, for the right person, at the right time.

And then AI starts making way more sense.

Because now it’s not just opening a massive page, reading everything, guessing what matters, and then trying to help.

It can search the structure.

look at properties.

follow the relations.

It can work inside the business, and actually on the business self improving it.

More of the business being able to move on its own. (i was the limiting factor and had to solve that)

But the other big problem is that every human thinks differently.

I remember learning this early in my first actual knowledge work job.

People’s brains don’t think like me.

That sounds obvious, but when you’re building systems, it matters a lot.

My setup works for me because my brain built it.

But that doesn’t mean it works for the team.

So now it’s less about forcing everyone into my perfect view, and more about building the backend properly so each person can have the front end that actually helps them work.

That’s the shift.

Same source of truth.

Different views based on the job.

That is how you get clarity without leaking context everywhere.

This is where I started thinking about the two versions of Oliver.

There’s Oliver the CEO, who has to manage everything and wants the top down view of the chessboard.

What’s moving.

What’s blocked.

Where the team is at.

Where the business is going.

But there’s also Oliver the employee.

The person who still has tasks that only I can do.

The person who has to sit down and actually execute.

Those two versions of me should not be looking at the same view.

If I’m trying to execute, I don’t need the whole business in front of me.

It just overloads my brain and I don’t actually take action.

I need the exact thing I’m meant to do next, with the context I need, and nothing else.

That’s why the human frontend matters so much.

It’s not just UI. (user interface)

It’s how you filter the business down so the right version of you can actually act.

And this is the part I keep coming back to.

Ingesting information for its own sake doesn’t actually change anything.

It’s dumb to read a book or watch a YouTube video and then not change your behaviour.

(thats a neat formula i got via hormozi)

That was the thing from this week.

I talked to smart people.

I saw what was working.

It made sense.

So I changed the system.

That’s the part I care about.

Not learning more.

Not chasing the next AI thing.

Not adding another tool because someone on the internet said it was the future.

Take the new information, adapt the system, and make the business easier to operate.

That’s the game.

Because the tool is not the point.

The model is not the point.

The workflow is not the point.

The point is building a business that humans and AI can actually work inside.

Peaceeeee