AI Architect: I received an inquiry from an AI the other day.

That email had no 'heart'

The other day, an inquiry email arrived at our company.

Polite language, well-structured. However, after reading just a few lines, I intuitively understood.

“Ah, this wasn't written by a human. It was written by an AI.”

This is by no means a criticism of text written by AI. After all, I use AI like a 'subordinate' in my daily work. The problem was not that the text was written by AI, but rather the 'content of the question' itself.

The email said the following:

“We are considering implementing your product. Therefore, please tell us the recommended settings and industry best practices.”

At first glance, it seems there is no problem. However, I found myself at a loss as to how to reply to this email.

Why I 'cannot answer' a perfect question

Because the most important information was completely missing from that email.

• What is the customer's current situation, and what are they struggling with?

• What do they want to achieve using our product?

• What is the skill level of the customer's team, and what is their budget?

It is impossible to talk about 'recommendations' or 'best practices' without any of this 'data' from the customer themselves. It is like getting into a taxi without telling the driver your destination and saying, 'Take me somewhere recommended.' The driver cannot take you anywhere.

Having an AI 'write a question' is not the problem. The problem lies in the process of simply dumping the task on the AI by saying 'think of a smart question' without providing the 'ingredients (data)' needed to generate the question.

Realization: The ability truly required in the AI era

Through this incident, I was made to think once again about the 'new abilities' of humans in the AI era.

With the advent of AI, we have become able to delegate part of our 'ability to provide answers' to AI. However, in exchange, the 'ability to formulate high-quality questions' has become more important than ever before.

And a 'high-quality question' is not about beautiful language or skillful questioning techniques. It is nothing other than the ability to accurately provide one's own 'unique data'—such as situation, purpose, and challenges—to the other party (whether human or AI) and share the 'context' that serves as the premise for the dialogue..

That client, while holding the powerful engine that is AI, had forgotten to add the fuel (data). As a result, the engine just spun idly, producing an empty question that reached no one's heart.

Conclusion: We are all being 'questioned' by AI.

The more AI permeates society, the more frequently we will encounter 'soulless questions written by AI' like this one.

At that moment, we are being tested.

Can we see through the 'lack of data' behind the question and draw out what the other party is truly seeking? And when we become the ones asking the questions, can we provide the 'high-quality data' and 'precise instructions' needed to make our AI partner perform at its best?

In the coming era, I am convinced that mastering AI refers precisely to this 'ability to design dialogue'.

How should we prepare the 'textbook (data)' we give to AI, and how should we design the 'instruction manual (prompt)' for AI to prevent communication tragedies like this one and create productive dialogue?

If you are interested in the 'rules of dialogue' for this new era, all the blueprints for it are written in my paid article.