Where the thinking goes
There’s a lot of noise right now about AI.
Is it good or bad?
Helpful or dangerous?
Something to lean into — or something to keep at arm’s length?
Alongside that question comes another:
Which AI tool should I learn next?
Lists circulate.
Charts compare novices to experts.
Frameworks promise productivity, speed, leverage.
Most of it is well-intentioned.
Much of it is useful.
And yet, something feels slightly off.
People are moving faster, but not always more confidently.
Outputs are improving, but judgement feels thinner.
Work is accelerating, while unease quietly grows in the background.
It’s hard to name — which makes it easy to argue about, and even easier to ignore.
This isn’t really about the tools
The conversation keeps circling the same questions.
Is AI a threat or an opportunity?
Is it replacing something important?
Is it time to resist — or to adapt?
But all of these questions assume something similar:
that control lives in the tool itself.
What I’ve been noticing instead is that control — and expertise — show up somewhere else entirely.
They show up in pauses.
In restraint.
In knowing what matters before anything is generated.
In recognising when an answer sounds fluent but hollow.
In other words: expertise shows up in where the thinking goes.
A familiar kind of intelligence
I often describe AI to friends as a very smart colleague at work.
The kind who knows a lot, thinks quickly, and is genuinely helpful — but only if you bring them into the problem properly.
If you walk up and say:
“I’ve got a client coming in — where should we go for lunch?”
They’ll answer.
Probably with something popular.
Convenient.
Perfectly fine.
But if instead you say:
“I’ve got a client coming in on Wednesday.
They’re vegan.
We’ve only got an hour, between one and two.
I need somewhere quiet, close by, and a bit classy.”
Now the answer changes completely.
Not because your colleague suddenly got smarter —
but because you undertook more of the thinking before you asked.
A habit worth noticing
When AI gives vague or disappointing answers, we often blame the tool.
But most of the time, what’s missing isn’t capability — it’s context.
AI doesn’t know:
what you care about
what you’re optimising for
which constraints matter
what risks you’re trying to avoid
Unless you decide where that thinking belongs.
And if you don’t, the system will happily guess.
A second, quieter truth
There’s another part of this metaphor that matters.
That smart colleague?
They’re also an intern.
Bright… Capable…Fast.
But still learning.
You wouldn’t hand them a client report and send it straight out the door without reading it.
You wouldn’t assume every confident answer was correct just because it sounded plausible.
And you wouldn’t stop thinking just because they’d done some of the work.
You’d review.
You’d question.
You’d take responsibility for what went out under your name.
Not as a burden — but as authorship.
Why this matters more than people expect
AI doesn’t make most mistakes loudly.
It makes them convincingly.
Like a capable junior colleague, it fills gaps without announcing them, smooths over uncertainty, and keeps going when it should probably stop.
That’s not malice.
That’s enthusiasm without judgement.
Which is why the real skill isn’t learning how to ask clever questions.
It’s knowing when to slow down and take the thinking back — and when to let it go.
Responsibility doesn’t disappear
One of the quiet risks of powerful tools is that responsibility starts to feel lighter.
The work comes back quickly.
It sounds right.
It’s neatly formatted.
But responsibility didn’t move.
It stayed exactly where it always was — with the person who decides what matters.
A quieter problem
Most people don’t misuse AI because they’re careless.
They do it because they’ve never been asked to think about where cognition belongs.
Does it belong in the prompt?
In the model?
In the human reviewing the output?
In the system that surrounds the work?
These aren’t technical questions.
They’re judgement questions.
And judgement is rarely taught directly.
Why stories help where frameworks don’t
Frameworks are good at naming stages.
They’re less good at showing how it feels to move between them.
Stories do something different.
They let us recognise ourselves without being told what to do.
They show patterns without demanding agreement.
They make space for opting out.
That’s why the essays that follow use ordinary moments — kitchens, habits, mornings — to explore how people actually learn to work with complex systems.
Not because AI is simple.
But because learning is human.
An important permission
Before we go any further, it’s worth saying this plainly:
You do not have to become an “expert” in AI.
If what you’re doing works,
if “good enough” is genuinely enough,
if speed matters more than depth for where you are right now —
that’s fine.
This isn’t a ladder you’re expected to climb.
It’s a set of observations you’re invited to notice.
You can stop reading at any point and lose nothing.
What this space is for
These essays are about judgement, not optimisation.
Systems, not shortcuts.
Cognition, not prompts.
Confidence that comes from understanding, not speed.
AI will appear throughout — but as context, not centre.
The real subject is how humans decide where to place their thinking when powerful tools are available.
That question won’t go away.
A gentle beginning
The next essay starts somewhere ordinary.
A morning.
A routine.
A habit so familiar it barely registers.
Not because the story is important —
but because recognising it might be.
If you find yourself nodding quietly rather than reacting loudly, you’re probably in the right place.
And if not — that’s fine too.
Either way, thank you for pausing here.
—
Jules Dee