How Much English Do Engineers Need in the Vibe Coding Era?

Recently, I have been seeing the term “vibe coding” quite often.

To be honest, I am still not very used to the phrase.

But regardless of the wording, I use AI in my daily development work, so I understand what it refers to.

Compared with the past, I spend much more time giving instructions to AI:

“I want this kind of feature.”

“Proceed with this design.”

“Investigate the cause of this error.”

Instead of writing all the code myself, development is increasingly becoming a job where I assign work to AI.

That has made me think about something recently.

How much English does an engineer need in the AI era?

If You Work Only in Japan, It Is Honestly Not That Big of a Problem

First, I should say this clearly.

If you are working on projects for the Japanese domestic market, I do not think a lack of English causes a major problem.

AI’s Japanese ability is already quite strong.

Specifications are in Japanese.

Meetings are in Japanese.

Communication with customers is in Japanese.

Instructions to AI are also in Japanese.

That is enough to keep work moving.

In fact, business domain knowledge is often more important.

For logistics systems, you need logistics knowledge.

For accounting systems, you need accounting knowledge.

For medical systems, you need medical knowledge.

There are plenty of things to learn before English.

Global Projects Change the Situation

The situation changes when the project itself becomes global.

In older global projects, there were times when you could somehow get by if you could write code.

Even if meetings were difficult, technical ability could cover part of the gap.

You could read documents with translation tools.

Engineers also tended to have a mindset that code was everything.

There was a feeling that even if your words were not enough, it was fine as long as you could write code.

But I think the AI era changes this.

The center of development is gradually shifting from writing code to explaining specifications.

And the other party is not only AI.

It may be overseas engineers.

It may be overseas project managers.

It may be overseas customers.

And they are also using AI in the same way.

In other words, the quality of the final output is increasingly tied not only to the ability to write code, but to the ability to verbalize requirements.

Competing Globally Raises the Required Level of English

Here, something slightly ironic happens.

As I mentioned above, when the focus shifts from code to specifications and explanations, the required level of English rises.

You need to organize requirements in meetings.

You need to explain design intent.

You need to clarify ambiguous requirements.

You need to persuade stakeholders.

You need to construct precise instructions for AI.

In other words, what matters is not just English for reading.

It is English for thinking.

And that requires a fairly high level.

That Said

Vibe coding has only recently started becoming a popular term.

Also, I am not currently participating in a global project.

So this article is based entirely on a feeling, or perhaps speculation. Please take that into account.

With that said, my conclusion is this.

If you work only in Japan, you probably do not need English very much.

But if you are looking at the global market or overseas projects, I think you will need it at a high level.

In the past, there were many engineers who could say:

“As long as I can write code, I can manage even if I am bad at English.”

There were many stories like:

“I work in Silicon Valley. My English is terrible, but they hire me because I can write code.”

But now AI has started writing code.

Then the important question becomes:

“Please explain what you want to build.”

In other words, as AI starts writing code, engineers are pulled back into the work of language.

Even though some of us became programmers because we were bad at talking to people.