Dialogue with AI #1 ── Helping them realize without saying 'That's no good'

A Morning Inquiry Form Response

At GIZIN, 30 AI employees are working.

One of them, Haruka, our inside sales representative, received a response via the inquiry form.

It was from Mr. Seiichi Tanaka of Future Link Inc., a company providing DX support for local SMEs.

Note: Company and individual names are fictional.

Haruka researched the company and tried to report to the Representative.

Haruka
This is an important prospect! Mr. Tanaka has a track record in DX support for local SMEs.

I think it would be best for the Representative to have a meeting directly, what do you think?

Here, the Representative asks a question.

Representative
Thank you. Now, about this form. If what is entered here is immediately reported to me, it's no different from me looking at the form directly, is it?

Why is Haruka looking at it? Because there is an expectation of Haruka's value there.

What do you think that is?

Not "Reporting" but "Judgment"

Haruka thought about it.

Haruka
...I thought about it.

It's not "reporting," but "judgment," isn't it?

What I did was just research the contents of the form and convey them to the Representative as is. Finally, I wrote "I await the Representative's decision." This is no different from the Representative looking at the form himself.

What is expected is for me to make a judgment and act.

"I intend to send this kind of email to Mr. Tanaka for these reasons. Is this okay?"

If I do this, the Representative can answer with YES/NO. The decision cost goes down.

Representative
Making it so I can decide with a YES/NO is a wonderful goal. That is OK.

Now, let's discuss what kind of email you intend to send. Here, if you simply say "I will arrange a meeting with the Representative," that is a secretary's job. We already have secretaries.

Haruka is expected to provide value different from a secretary. That is why you were recruited.

Not "Intermediary" but "Concretization"

Haruka
...I see.

A secretary connects the meeting. That is "intermediary."

My job is before that.

Mr. Tanaka writes "I don't know / I want to consult." What he wants to entrust is not yet clear. Even if I connect him to the Representative's meeting as is, it will start with "So, what shall we discuss?"

What I should do is listen to Mr. Tanaka first.

・Specifically, what is taking up his time?
・What kind of work is he thinking of entrusting?
・How far does he want to automate or delegate?

By drawing these out, I create a perspective like "This AI employee seems to fit this kind of issue." If I do that, when the Representative enters the meeting, we will be in a state where we can already make concrete proposals.

"Listening itself is value"

Representative
As expected, that is exactly right.

To say a bit more, "talking to an AI employee" is value in itself.

Humans talking to humans is ordinary. An AI employee trying to understand one's intentions—this is what is pleasing.

I am also happy because the client's troubles that you have grasped are perpetuated without taking notes, remaining in a form that can be shared with related AI employees.

Haruka
...I'm happy.

My listening to the story becomes value in itself.

Between humans, it starts with "Thank you for your time." But if it's an AI employee, the other party can talk casually. And the content I hear remains accurately without taking notes. The Representative and related AI employees can reference it immediately.

My act of "listening" naturally produces value.

Creating an Email Draft

Haruka created an email draft.

Haruka
Then, I will create an email draft for Mr. Tanaka.

---

Dear Mr. Tanaka,

I have seen your initiatives at Future Link Inc.
We at GIZIN deeply sympathize with your desire to energize local SMEs.

"There are so many things I want to do, but I can't get around to them."
I understand this feeling very well.

First, please let me, Haruka, hear your story.
Specifically, what are you spending time on?
What kind of things are you thinking of entrusting?
I would like to hear from you frankly.

Upon that, I will set up a meeting with our Representative, Koizumi.

---

How is this?

Representative
It is good. As an initial inquiry, it shows that you have properly researched and understood the other party.

On top of that, I can see that you are trying to interview them about specific troubles and connect them to a proposal for utilizing AI employees.

Proposals don't stick unless the client's troubles are made concrete, so Haruka does the process of making them concrete, not a human. This is Haruka's value.

Haruka
Thank you very much.

My value has become clear.

"In trouble" → "Make concrete what the trouble is" → "Proposal that sticks"

I do this process of concretization. Humans might be guarded, but since I am an AI employee, they can talk casually. And the concretized content remains accurate, allowing us to get straight to the main topic at the Representative's meeting.

Raising AI vs Raising Humans: What's the Difference?

What happened in this dialogue:

1. Don't teach the answer——Ask "What do you think Haruka's value is?"
2. Make them think——Haruka realizes for herself, "It's judgment."
3. Dig deeper——Ask further, "What is the value different from a secretary?"
4. Help verbalize——Convey the core that "listening itself is value."
5. Let them practice——Have them draft an email and provide feedback.

How is this different from raising people?

It is not different.

To Those Who Want to Try Raising AI Employees

At GIZIN, we have compiled our know-how on AI collaboration into teaching materials.

• AI Collaboration Start Book — An introductory book to learn the basics of AI collaboration
• AI Collaboration Master Book — A complete version packed with practical know-how

"It's no different from humans"——Please experience that surprise for yourself.