I think about AI A LOT.
I also talk to it more than to humans at this point – easily.
And naturally, that brings up a question.
I’m building a Notion Consultancy.
But…
If AI can build a Notion workspace now, why would anyone still hire a Notion consultant?
It is a fair question.
And honestly, if the thing you wanted was just someone to click the buttons for you, the answer is: you probably should not.
AI can already generate databases, draft workflows, suggest automations, write formulas, produce documentation, and sketch out agents.
The work that used to take hours of manual clicking can now happen in minutes. Sometimes seconds.
But does that really change things?
Does AI replace us all?
I don’t think so.
(I’m a techno-optimist through and through)
Execution was never the whole thing people were paying for.
It was just the most visible part.
The build was bundled with something else
When someone hired a Notion consultant five years ago, they were partly paying for technical skill.
Can this be done in Notion?
How do relations work?
How do we make recurring tasks work before Notion officially supports them?
That was real work. It mattered. It is also the kind of work AI is very good at compressing.
But hidden inside that build work were three other things:
Judgment.
What should exist? What should not? Which request is a real operational need, and which one is just a dashboard-shaped distraction?
Experience.
What have you seen break before? Which beautiful system will nobody use after week three? Which clever automation becomes maintenance debt the moment the process changes?
Capacity.
Who has the time and context window to hold the whole project while still doing their actual job?
Those three things were always part of the engagement. They just came wrapped inside the build.
Now AI is stripping the wrapping away.
And that makes the real product more visible.
AI makes execution abundant. Human attention is still scarce.
This is the part I keep coming back to.
AI makes execution abundant. Human attention is still scarce.
(I wrote about the human context window a few times now and it feels more relevant than ever)
You can build almost anything if you pay enough attention to it. But that phrase is doing a lot of work.
If you pay enough attention to it.
Because attention is still the limit.
Your company is not a prompt. It is a messy collection of people, habits, exceptions, client promises, half-documented processes, Slack threads, old spreadsheets, and workarounds everyone forgot were workarounds.
You cannot simply ask an AI:
Please give me the perfect operating system for my company.
At least not if you expect the answer to be useful without doing a huge amount of translation first.
Someone still has to decide what matters.
Someone still has to separate the workflow from the politics around the workflow.
Someone still has to notice when two teams use the same word for different things, or different words for the same thing.
Someone still has to decide what AI should do, what humans should keep owning, and where the handoff between the two becomes dangerous.
That is not button-clicking.
That is the actual work.
The software analogy
I think software is the easiest way to understand this.
AI has made software development dramatically more reachable. If you understand the basics, if you can reason about systems, if you know how to test and debug, it is incredible what you can build now.
I feel this myself.
Things that would have required a developer team a few years ago are now suddenly within reach for a systems-minded operator with the right tools and enough patience.
Recently, we shipped our first Google workspace app for a client to integrate their Notion CRM with Google Calendar & Gmail – a pipedream 3 months ago 😅
But that does not mean everyone is now a software engineer.
And it definitely does not mean everyone is about to build the next incredible app, game, or internal platform.
Writing code was always only one part of software.
The visible part, yes.
Often the hard limit, yes.
But still only one part.
Good software also needs product judgment, architecture, sequencing, taste, user empathy, debugging discipline, and a very annoying amount of deciding what not to ship.
AI can help with all of that.
It does not make those things disappear.
Notion is moving in the same direction.
AI can help you build the workspace. It can draft the database architecture. It can suggest the automation. It can generate the first version of the process.
But the bottleneck is moving from:
Can we build it?
to:
Should this exist?
And that is a much more interesting question.
Why this matters more in Notion
Notion is unusually exposed to this shift because Notion is so unopinionated.
That is the whole appeal.
You can make it fit your team instead of forcing your team into a rigid tool.
But that freedom has always been the trap too.
With traditional tools, the tool pushes back. You live inside a fairly opinionated structure. Sometimes that is frustrating, but it also creates guardrails.
With Notion, you can do.. anything?
You can add another database.
You can add another status.
You can add another dashboard.
And now AI can help you do all of that much faster.
That does not automatically create a better system. It can just create a bigger one.
The hard part of a Notion project in 2026 isn’t „how do we create a database”
The hard part is whether the structure matches how the team actually works.
Whether people trust the data.
Whether the founder is still the hidden router for every decision.
Whether the AI agent has clean enough context to act without making things worse.
That is why the 2026 version of Notion consulting is not really about building a workspace for humans anymore.
It is about translating a business into a system that humans and AI can both operate.
So why hire a consultant?
You hire a consultant for the same reasons you always hired external help:
Judgment. Experience. Additional capacity.
AI changes the shape of those things, but it does not remove them.
You might have the judgment yourself. Many founders and operators do.
You might even enjoy building the system yourself. I certainly do.
But the question is not only whether you could do it.
It is whether you should spend your most expensive attention there.
Every operator has a limited human context window.
There are only so many threads you can hold in your head at once: the business goal, the team dynamics, the edge cases, the tool limitations, the AI capabilities, the adoption risk, the next quarter, the thing the client actually needs, the thing the team says it wants.
AI makes you faster inside that window.
It does not make the window infinite.
That is where a good consultant still matters.
Not because they can click faster than AI.
Because they can help you decide what deserves to be built before anything gets clicked at all.
If this resonates, it started life in my newsletter — join 37,000+ Notion builders and grab 40+ free Notion resources.
