Some questions sound technical until we realise they are really about people.

The question of whether we should still learn coding in the age of AI is one of them. It is not only about programming languages, commands or whether AI can now write code. It is also about access, understanding, responsibility and how ordinary people live with technology that is becoming easier to use but harder to fully see.

I was one of the invited penalists at the ASEAN Sustainable Ready Industries 5.0: AI, Research and Innovation Summit 2026 in Brunei held at Universiti Teknologi Brunei (UTB). The panel topic was framed around a question that, at first, may sound simple:

Is learning coding still necessary in the age of AI?

This is the kind of question that can become technical. It can lead people to start talking about programming languages, scripts, commands, software engineering, cybersecurity, app development and whether AI can now write code better or faster than humans.

And don’t get me wrong, those are important discussions. But I wasn’t there as a software engineer or a developer. That is not my background. I was more interested in education, AI literacy, responsible technology use and how ordinary people experience these changes in their daily lives.

Perhaps the deeper question is not only whether we should still learn coding.

Perhaps the deeper question is this:

What happens when technology becomes easier to use than it is to understand?

That, to me, is where the conversation becomes more human.

The easier interface

AI today is not only powerful because of what it can produce. It is powerful because of how easily people can talk to it.

You do not always need to know the exact command. You do not always need to know the correct formula. You do not always need to know where the function is hidden in the menu. You can simply describe what you want.

Write this.
Summarise that.
Explain this to me.
Create a table.
Generate an idea.
Fix this text.
Build a simple website.
Help me understand this code.

This is why conversational AI feels different. It does not only automate tasks. It changes the relationship between people and technology.

For many ordinary users, the barrier has become lower. The interface has become more human. The machine is no longer waiting for precise technical instructions. It is responding to natural language.

That opens up many possibilities.

A teacher can generate lesson ideas.
A student can ask for explanations.
A small business owner can draft content.
A public officer can summarise reports.
A parent can understand a school document.
A creative person can prototype something they used to need a developer for.

But did you know that there is another side to this?

When something becomes easier to use, we may start to assume that we understand it better than we actually do.

The skill does not disappear

During the panel, I mentioned that technology often does not remove complexity. It hides or abstracts some of it, so more people can participate.

We have seen this before.

There was a time when using computers meant typing commands into systems like DOS. If you wanted to open a file, move something or run a programme, you had to know what to type. Then graphical user interfaces came along. All of a sudden, users could click icons, open folders and use menus instead of remembering commands.

For most, DOS and typed commands slowly disappeared from everyday use. But that knowledge did not become useless. It moved to another layer, where technical users, IT professionals and developers still rely on it when they need deeper control.

The same happened with websites. When HTML became popular (for some), we bought thick books to learn how websites worked. I remember that period clearly. A friend and I spent money on those books because building websites felt like entering the future. Then tools like Adobe Dreamweaver came along, and people could create websites visually. Later came templates, content management systems and website builders.

Like command-lines, HTML did not disappear. It just became less visible to many users.

Excel gives another example. In the past, being good at Excel often meant knowing formulas. Today, many people can use templates, menus, pivot tables and now AI assistance. But formulas still matter, especially for advanced users. More importantly, understanding what you are calculating still matters.

Even media went through this. Television presenting used to require studios, broadcast institutions and production teams. Then YouTube and social media opened the door for creators, influencers and ordinary people to publish directly. But communication skills did not disappear. If anything, credibility, judgement, audience awareness and responsibility became even more important.

This is the pattern I see.

The skill does not disappear. It changes layer.

So when the first question at the panel was asked on whether coding is still necessary, the answer should not be a simple yes or no.

Coding may become less visible to everyday users, but it becomes more important for those responsible for building, checking and maintaining systems.

I was one of those parents too

Not too long ago, many of us were told that children should learn coding because coding was the future.

I understood that argument. I still do.

I was one of those parents too. My wife and I sent our children for coding lessons because we believed it was a useful future skill. At that time, it felt like one of the clearest ways to prepare them for the world ahead.

But after seeing how conversational AI has developed, our view has changed.

I do not think coding is no longer useful. I do not think children should stop learning it. But I no longer think the argument is as simple as “everyone must learn coding because everyone will need to become a coder.”

That is where we need to be more honest.

Some young people will need deep coding skills. If their pathway is software development, cybersecurity, AI engineering, digital product development or technical research, then coding is still essential. They will need to understand systems deeply. They will need to debug, test, secure and maintain what they build.

But most people in our society are not developers and they will not become developers.

For many, the more important need may be coding awareness rather than coding fluency. They need to understand logic, systems, data, automation and how digital tools behave. They need to know enough to ask better questions. They need to recognise when something may be wrong. They need to understand that digital outputs do not appear by magic.

In other words, they need to be digitally literate in a deeper way.

Brunei as users and builders

This question also feels different when we place it in the Brunei context.

Many of us experience technology mainly as consumers. We use apps, platforms, devices and services built elsewhere. We adapt to systems that were not designed around our local habits, languages, culture or scale.

That does not mean Brunei cannot build. It does not mean local innovation is impossible. But it does mean we need to be realistic about the pathways we are preparing young people for.

If Brunei wants to grow as a producer of digital products, AI tools, software, platforms and technology services, then we need people with deeper technical capability. We need developers, researchers, designers, cybersecurity specialists, data people, product thinkers and entrepreneurs.

But if most people will experience AI through everyday work, education, business, public services and personal life, then the conversation must also include ordinary users.

They may not need to become professional coders.

But they do need to understand the digital world they are increasingly depending on.

This is where I think education has to be careful. Coding should not be taught only as syntax. It should not only be about writing lines of code to produce the correct output. It should also be about problem-solving, logic, creativity, ethics, verification and responsibility.

Because AI may help us produce things faster, but it does not automatically help us understand them better.

Adoption is not readiness

This connects strongly with what I found in my independent report, Brunei and Generative AI.

One of the central messages in that report was this:
Adoption is rising, but readiness depends on literacy, ethics and access.

In the survey sample, many respondents had already used Generative AI. But far fewer clearly understood the difference between AI and Generative AI. That difference matters because it shows that people can be exposed to a tool, use a tool and even benefit from a tool without fully understanding what it is or how to judge it.

This is not a criticism of users. It is simply how many technologies enter our lives.

We often use first and understand later.

The problem is that Generative AI produces answers, explanations, summaries, images, ideas and code in a way that can feel confident and complete. This makes it harder for users to know when to trust it, when to question it and when to stop.

In the report, respondents were not simply rejecting AI or accepting it blindly. Many felt that the ethics of AI depends on how it is used. That is an important distinction.

It means the public is not only asking, “Is AI good or bad?”

They are asking:

What is it being used for?
Who is using it?
Who benefits?
Who is affected?
Is there human review?
Is it fair?
Is it accurate?
Is it appropriate?

Those questions apply to AI writing. They apply to AI-generated images. They apply to AI in education. They also apply to AI-generated code.

Whether AI can produce something is not just the issue.
The issue is whether we understand enough to use that output responsibly.

What the audience was really asking

It was interesting that during the session, some of the audience questions moved beyond coding. I would like to mention and share some here.

There were questions about commercialisation in Brunei. They asked whether local innovation can survive in a small market. They asked about profitability, procurement, regulations, online transactions, intellectual property, grants, research, industry readiness, job losses, security vulnerabilities, AI accuracy and opportunities for disadvantaged youth.

After having more time to think, I thought the audience was not just asking “should we learn coding?” but I think it is moving towards something like:

Are we ready for the kind of society AI is quietly creating?

I feel that is a very different conversation. These are things, for the past few years, have always been in my mind…

If AI makes it easier to build prototypes, who helps turn them into sustainable products?

If AI helps people create content, how do we protect local talent and creative identity?

If AI allows people to automate work, what happens to jobs and skills?

If AI can generate code, who checks whether it is secure?

If AI gives different answers to the same question, how do ordinary users learn to verify?

If AI becomes part of education, how do we ensure that disadvantaged youth are not left behind?

Look at those. They are not just technical questions.

They are questions about access, trust, opportunity and responsibility.

The new divide

For a long time, we talked about the digital divide as a question of access to devices and the internet.

Of course, that still matters.

But AI may create another kind of divide.

It may not only be between people who can code and people who cannot. It may be between those who are AI-ready and those who are only AI-exposed.

Being AI-exposed means you have seen the tools. You may have tried them. You may know they exist.

Being AI-ready means something deeper.

It means you know how to ask better questions.
It means you know how to check outputs.
It means you understand that AI can be wrong.
It means you know when not to use it.
It means you are aware of privacy, bias, fairness and local context.
It means you can use AI as support without surrendering your judgement.

This is why access alone is not enough.

Giving people tools without literacy may only create a more confident form of dependence.

What should we teach?

So should we still teach coding?
Yes, but not as if the world has not changed.

We should teach coding as one part of a wider literacy. It should sit alongside AI literacy, digital literacy, computational thinking, data awareness, creative problem-solving, ethics and critical thinking.

For some students, coding will become a professional pathway. For others, it will become a way to understand how digital systems work. For many ordinary users, it may help them become less passive when they use technology.

The aim should not be to make everyone a software engineer.

The aim should be to help people avoid becoming helpless in a world increasingly shaped by software, platforms and AI.

That difference matters.

If we only teach coding as a technical skill, students may eventually ask why they need to learn it when AI can generate code. But if we teach coding as a way of thinking, questioning, testing and understanding systems, it remains relevant.

Because the real skill is not only writing the code.

It is knowing what problem is worth solving.
It is knowing whether the output makes sense.
It is knowing what happens when a system fails.
It is knowing when a human should remain in the loop.

When digital life becomes normal

The more I think about it, the more I feel that conversational AI is accelerating something already happening around us.

Digital life is becoming normal before we fully understand what that means.

We are expected to use apps for services. We are expected to communicate through platforms. We are expected to submit, verify, book, pay, learn and work through digital systems. We are expected to adapt quickly.

Now AI adds another layer.

It does not only ask us to use digital systems. It asks us to trust systems that can respond, suggest, generate and decide with increasing confidence.

That’s why the human side matters.

It is not enough to ask whether people can use AI. It is not enough to ask whether students can code. It is not enough to ask whether a tool can produce an output.

We also need to ask whether people understand the systems shaping their choices.

Because when technology hides the hard parts, it can feel like everything has become easier.

But sometimes what has become easier is only the interface.

The responsibility remains.

What remains human

Leaving the panel, I kept thinking that perhaps we have been asking the question too narrowly. Coding is still necessary, but maybe not in the way we used to imagine.

For developers, it definitely remains a core skill.

For the rest of us, the need is different. We need enough understanding to live, learn and work in a world where more things are built through systems we may not see.

The future may not require everyone to become a coder.
But it does require more of us to become better questioners.

Better users.
Better judges of output.
Better protectors of our own data.
Better supporters of our children’s digital lives.
Better citizens in a society shaped by technology.

AI may make building feel easy.

But ease should not replace understanding.

And access should not replace readiness.