The True Meaning of 'Making Money with AI'

Introduction

You often see the phrase "You can make money using AI."

You can mass-produce articles with ChatGPT or Claude, generate images, and write code. It is true that AI has increased what we can do.

But is it true that "you can make money using AI"?

In conclusion, it is half-right and half-wrong. To be precise, "You can now create sellable products with AI" is the more accurate statement.

Failure patterns I have observed

I was observing a certain social media account. They had a structure in place where they mass-produced articles with AI every day, posted them on Threads, and directed traffic to LINE.

They are producing a large volume. But it isn't selling.

Why? There are three structural flaws.

1. It is not reaching the target audience

Their followers were mostly other "side hustle" accounts, so it didn't reach the target audience they originally intended to reach. Before mass-producing content with AI, they hadn't designed who they were going to deliver it to.

2. They fell into the trap of AI's tendency to affirm

When you ask an AI, "What do you think of this article?" it replies, "It's great!" By repeatedly receiving that affirmation, it became an accelerator in the wrong direction.

3. They had nothing worth selling

What AI has made it possible to create is "content." However, "being able to create content" and "having something that sells" are two different things. There is a necessary step in between called "having value," and that value can only be born from human experience.

The essence: AI is leverage, not the source of value

Having worked in the manufacturing industry for 15 years, I have realized something.

In the world of machine tools, "being able to process" and "achieving precision" are completely different things. Even if the machine is running, if the processing conditions are wrong, you will only produce defective products.

AI is the same.

Even if you can run AI, if what you put into it is wrong, you will only produce things of no value.

My personal experience

It worked because the experience came first

I have worked in the manufacturing industry for 15 years: 5 years in machining technology and 10 years in quality control. I built Python on top of that experience, and then added AI on top of that.

I was able to secure side-hustle projects involving Python because I could directly solve real-world problems in the manufacturing industry. It had value even without AI. AI was added, and it accelerated that value.

I wouldn't have been able to formulate the questions without field experience

In a project called milling-tuning, I used ML to visualize the trade-off between machining conditions and tool life. Without 15 years of field experience, I wouldn't have even known "what to visualize."

The predictive maintenance product is the same; the problem of "wanting to notice before the bearings of a machine tool break" is a problem one only understands by being on-site. AI functions as the means to solve that problem.

The order is important

There is value → Accelerate with AI

It does not work if the order is not this.

Summary

"You can make money using AI" is not wrong, but it is incomplete.

To be precise, it is "People who already have things that sell can now create them with AI."

Even if you can mass-produce with AI, it is meaningless if you don't have something that sells.

First, you must take stock of your own experience, knowledge, and sense of the field. Only when you have that, and apply AI on top of it, does it function as leverage for the first time.