?Introduction

The era of ChatGPT writing code has arrived.

In the past, it took days to create an EA by researching MQL syntax, finding sample code, and fixing errors. Now, if you ask AI, a template can be produced in minutes.

Looking only at this change, some people might think

“Are EA developers becoming unnecessary?”

and some may wonder.

But in reality, the opposite is true.

What becomes unnecessary in the AI era is simply developers who only write code.

On the other hand, developers who understand the market, design the structure of EAs, and verify AI outputs become more valuable.

EA development is shifting from coding tasks to using AI to design financial systems.

?Areas AI can replace

AI excels at routine implementations.

For example,

・Entry conditions using moving averages
・Overheat determination by RSI
・MACD cross determination
・Setting fixed SL/TP
・Simple backtest EA
・Adding input parameters
・Logging output

Such processes can be generated very quickly by AI.

The barrier of “just making a working EA” for beginners has been greatly simplified by AI.

In other words, developers who valued only simple implementation work will face a considerable challenge going forward.

?Areas AI struggles with

Conversely, AI also has clear weaknesses.

That is,

“Why does this logic have an edge?”

to determine.

AI can write code.

But it does not guarantee why it will win in the market.

For example,

・Which timeframes have an edge
・Is it for trending markets
・Or for ranging markets
・Aiming to capture volatility expansion
・Taking pullbacks
・Taking breaks
・In which situations should it stop

These judgments are not mere programming.

They are market understanding.

This cannot be filled by AI alone.

?The EA developers who will remain

Those who survive are not code artisans.

They are designers.

To AI,

“Create it under these conditions”

rather than,

“Translate this market structure into verifiable rules like this.”

are the kind of people who can instruct AI.

Furthermore, after AI outputs code, they can review it to verify that

・MQL4 and MQL5 are not mixed
・CopyBuffer usage is correct
・Handle is properly released
・ArraySetAsSeries is appropriate
・It is not overfitted to backtests
・It will not fail in real operation

and other checks.

In AI era, EA developers will be valued for their review capability more than implementation power.

?EA development roles move upstream

In the past, EA development involved humans thinking of logic, coding, and making fixes.

From now on, that will change.

Humans think about market structure.

Humans design.

AI implements.

Humans verify.

Humans decide improvements.

This is how it will be.

In other words, developers’ jobs do not disappear; they move upstream.

Judgment becomes more important than hands-on work.

?MAEP perspective

MAEP is built on this premise of change.

It is not enough to just have AI use MetaTrader.

AI needs to be educated about MetaTrader.

This is because AI does not fully understand MQL.

Mixing MQL4 and MQL5.

Mis-generation of OrderSend.

Misuse of CopyBuffer.

Handle leaks.

These problems demonstrate that AI only superficially understands MetaTrader.

Therefore, EA developers need the ability to supervise AI.

✅Summary

Can EA developers survive in the AI era?

The answer is yes, they can survive.

However, their roles will change.

Simply writing code will become tougher.

Nevertheless, understanding the market, designing EAs, mastering AI, and judging quality will increase in value.

AI will not erase EA developers.

It will push EA developers to upstream processes.

?CTA

The era of making EA with AI has begun.

But what is truly needed is to teach AI to understand MetaTrader correctly.

MAEP is a knowledge package to support MetaTrader development in the AI era.

Not only using AI, but educating AI.

There lies the next possibility for EA development.