I get annoyed by subtle Japanese

Recently, while reading texts, I have
something that bothers me

Did the person write this themselves, or did they have an AI write it??

It's fine to have an AI write it, but
did they check it properly and send it after being satisfied with it?

I seem to get annoyed
when the location of responsibility becomes unclear

Before writing a text, determine your intent

More than the text itself,
who you want to reach and what you want to deliver
that message is more important

To put it another way,
with what intent are you sending that message

If the higher-level purpose is clear,
the expression in the text is not a big problem

But if I don't know what the intent was in the first place,
I feel frustrated

The problem of texts with unclear intent flooding the space

It's fine to utilize AI, but
texts with unclear intent confuse people

It's not that the Japanese is wrong, but
the intent cannot be read
the consistency between the beginning and end is slightly off
in extreme cases, contradictory content is written without a second thought
What does this mean??

Here, I suddenly realize
that it's not a human...

If you are going to have an AI write it, it's better to work out the outline together

If you give precise instructions,
today's AI can write quite high-level texts

If you give precise instructions... but
this is surprisingly troublesome

I have the AI think through the core structure with me, and
through this process, I ensure we develop a shared understanding.

(Though, when I say 'understanding,' I don't mean the AI is thinking like a human.)

• Background, intent, and key message

• Who the target audience is

• Expected emotions, thoughts, and actions after reading

I have the AI organize these things
while brainstorming with it.

Only humans can provide instructions for unique perspectives and messages.

Ultimately, AI is good at writing safe, generic text,
and it won't come up with unique perspectives or messages.

If it did come up with something like that, it would probably be annoying.

Even if it provides several candidates,
choosing the main message and the focus from those
is the human's role.

What to contrast with what, and what to emphasize.

What is the theme you want the other person to think deeply about?

Determining these intentions and meanings
is something only humans can do.

The ability to express intent in one or two lines.

If possible, express the intent simply in one or two lines
at the beginning or end of the text.

This part, at least, is easier to understand and less prone to misunderstanding
if written by a human.

I want to say, "Let's put in the human effort and not cut corners on things like this."

On top of that, if you give instructions like, "
This is an important part, so please grasp the intent and adjust it so the preceding and following sentences flow smoothly
", the AI will do its best.

Collaborative work between humans and AI is still at a stage where, like this,
it won't run smoothly unless you give detailed instructions.

It's a matter of how you use it, but
in the end, for now,
I think humans have no choice but to take responsibility for the text.

This article was written by

Shigeki Kawahara, a "Future Realization Partner" who increases revenue pillars
https://mousoubiz.com/
https://twitter.com/mousoubiz