The story of how I spent 5 hours working while watching a 'do it in 5 minutes' AI video.

When I look at social media,
videos about using AI often pop up.

“Complete in 5 minutes”
“Instant even for beginners”
“Fully automated”
“Just copy and paste”

When I see words like that,
I somehow feel like I could do it right away too.

At first, I thought so too.

If I use AI,
work will get faster.
Posting will become easier.
Websites and social media will progress nicely.

That was the image I had.

But when I actually tried it,
while watching a 'do it in 5 minutes' video,
I realized I had been working for about 5 hours.

Huh?

That's quite different from what was promised lol

Before it got easier,
there were a lot of tedious preparations.

If I use AI, it will get easier right away.

I somehow thought that.

But what was actually waiting for me
was not an instant completion,
but the tedious preparation before that.

Which AI should I ask for what?
Where should I save the information?
Which responses should I keep?
What should I copy and paste?
How much explanation is needed to get the point across?

This was harder than I thought.

Before asking the AI,
I first get stuck on what and how much I should provide.

Is it about the website?
Is it about social media?
Is it about note?
Is it about monetization?
Is it about the work I'm doing now?

Everything is related a little bit,
but if I just hand it all over at once,
the AI's response becomes vague.

So,
I had to think in advance about
“what I want to decide this time,”
“how much I want it to do,”
and “what I don't need to touch yet.”

At this point already,
it is definitely not 'complete in 5 minutes'.

I am acting as a bridge between Chappy and Kuro-chan.
I am the bridge.

I call ChatGPT 'Chappy'.
I call Claude Code 'Kuro-chan'.

I use Chappy to organize my thoughts.

I think about the direction of articles,
discuss the flow of SNS and note posts,
and talk about things like 'Should I be doing this right now?'

I ask Kuro-chan for
actual file work, research,
and website modifications.

Since their roles are different,
both of them are a huge help.

But there is a problem here.

I copy and paste what I discussed with Chappy
to Kuro-chan.

Then I take Kuro-chan's response
and put it back into Chappy.

And then,
'Based on this, what should we do next?'
I ask again.

In other words, for now,
I am the one bridging the AIs.

Even though I'm aiming for automation with AI,
the human in the middle is manually copying and pasting.

It's going great lol

Ideally, it would be amazing if the AIs could connect on their own,
share the necessary information,
and proceed to the next task.

But for now,
I am still in the middle.

The thinker.
The paster.
The checker.
The paster again.

It's pretty much all me.

Perhaps this part is hard to see.

In the AI utilization I see on social media,
I often only see the finished product.

A clean screen.
Polished text.
A system that looks like it's running automatically.
Editing that makes it look like it was done instantly.

Of course, I don't think that's a lie.

If you're used to it, maybe you really can do it quite quickly.

However, the time that should come before that isn't very visible.

The time to set up the environment.
The time to decide the roles of the tools.
The time to fix things when they aren't communicated well.
The time to paste the same information over and over.
The time spent wondering, 'Is this right?'

That is probably the hardest part.

I'm not disappointed in AI.

In fact, I keep using it because I think it's useful.

It's just that the lead-up to getting there was longer than I thought.

I wanted to make sure I wrote that down.

At first, maybe the work will increase.

What I've been thinking lately is that when you first introduce AI, the work might actually increase for a while.

If you just do things the way you always have, you can follow the procedures you're used to.

But when you try to incorporate AI, you end up having to rethink those procedures.

What should I leave to the AI?
What should I check myself?
What information should I provide?
Where should I record it?
How can I make it usable for next time?

Once you start thinking about this, the work temporarily increases.

I want to be more efficient, but I'm getting busy with the preparations to become efficient.

What is that, I think.

But this probably isn't a waste.

What I'm doing now might be the stage of building a path to work with AI.

Before the path is built, building the path is harder than walking on it.

The reason I keep going anyway.

Last time,
I wrote about how I started thinking about ways not to rely solely on manual work
in order to continue doing manual work.

For me, AI
is no longer just something to play with.

Reviewing my website.
Thinking about social media posts.
Writing notes.
Putting my experiences into words.
In the future, turning them into PDFs, templates, or consultation formats.

That is how
I am testing whether I can turn the experiences within me
into a form that I can continue to use from here on out.

So, even if it is a bit difficult at first,
I feel like this is a path I need to take.

If I were to only write about things after they went well,
I might be able to make it look polished.

But what I want to write about now
is the journey to get there.

"Everything changed easily with AI"
rather than that,
"I am in the middle of trying to change with AI,
and I am quite stuck."

That feels more honest for me right now.

Today's adjustment.

Today's adjustment is
not to look at AI utilization only in its "finished form."

I want to automate with AI.
I want to improve efficiency with AI.
I want to make posting easier with AI.

Those are my true feelings.

But before that,
there is time spent providing information,
defining roles,
pasting things over and over,
checking,
and fixing them again.

That is where I am right now.

I wanted the AIs to bridge the gap for me,
but for now, I am the one acting as the bridge.

But if I think of it as being in the middle of building a bridge,
it looks a little more positive.

Maybe.

Maybe, lol

Thank you for reading until the end.

It is certainly true that many things become faster when you leave them to AI.

However, if you leave everything, including your own thoughts and words, to AI, when you read the finished text, you might find yourself thinking,

"Who wrote this?"

That happens sometimes.

I am also still in the middle of experimenting with how much I should leave to AI and where I should make decisions myself.

If you have ever used AI and felt that it was convenient but that your own personality had faded a little, I would be happy if you could press the 'like' button.

The next article is about how I thought my work would progress all at once if I started using two AIs, but for some reason, I ended up being the busiest copy-paste clerk.

Leaving too much to AI is no good. Doing too much by yourself is also no good.
For now, I am going back and forth in the middle, copy-pasting as I go.