What Was Agile? - Things I Finally Understood in the Age of AI

About 10 years ago, I was constantly confused.
Every time I read a technical article,

these three always appeared together.

That was all they were. Back then, I always had the question,
"What is the difference between these three?"

However, now that it is 2026, when I organize technology alongside generative AI, the answer is surprisingly simple.

They were simply explaining things that exist on different "layers" all at once.

The trigger for looking back at Ansible was this article byYuimaru - Born in 2005, using AI extensively outside the IT circle.

You can understand it instantly if you think about making curry

For example, let's say you are making curry.
Here,

Agile

This is the "way of proceeding with cooking."

In other words,the way of working itself.

DevOps

This is the teamwork between the cooks.

レシピを考える人と、
実際に調理する人が、
ケンカせずに協力する仕組み。

In short, it is organizational culture.

Ansible

This is not a kitchen knife.

面倒な作業を全部自動でやってくれる道具。

もっと言えば、フードプロセッサーです。

In short, it is a concrete automation tool.
It is automation that has existed normally for years before AI.

These three belong to completely different layers:
· Policy
· Organization
· Tools
—so they are not actually things that should be compared.

Why were they all discussed together back then?

The reason is simple.
The business side started saying, "We want you to improve services faster."
Then developers would say,

"We built it quickly using Agile."
However, the operations side would respond,
"It would be an accident if you touched the production server that frequently."

That is why DevOps gained attention as a culture where development and operations cooperate.
But work cannot proceed with culture alone.
In the end,

tools like Ansible, which automates server configuration, became necessary.
In other words,


as a result of wanting to realize Agile, DevOps became necessary, and Ansible was used as the means to implement it.

That was the flow of things.

Back then, there was an atmosphere of "learn everything"

The problem was marketing.
Tech magazines and events
under the banner of
"Latest Technology!"
introduced them all together.

As a result,
from a beginner's perspective,
they all looked like technologies in the same category.
In reality,
they were not even comparable.

The turbulent period of 2014–2018

These few years were an interesting era.
Ansible spread Infrastructure as Code,
Node-RED was open-sourced,
the ESP32 appeared,
and the IoT boom began. If you opened Qiita,

Raspberry Pi
ESP32
Node-RED
DevOps
Docker
Ansible

such keywords were lined up almost every day.
They were all technologies that felt like the future, but
because they all appeared at the same time,
I think the whole world was a bit confused.

2026

Now, AI
naturally connects this fragmented world.

AI bridges the world that used to require crossing through many specialized books to understand, all within a single conversation.
Now,
it has dawned on me that "back then, they were just explaining things with the layers all mixed up."

Technology evolves. But the structure remains the same.

The tools that are popular will change.

will eventually pass their roles on to the next generation. However,

the structure itself remains unchanged.

When you encounter new technology,
first think about "which layer this belongs to."
Just doing that will significantly reduce the chances of getting lost in your learning.
After 10 years, I have finally understood that.

#Agile #DevOps #Ansible #GenerativeAI #HistoryOfIT