Malaysia shouldn’t think of AI as just another chatbot

From Ian Ku

For many Malaysians, artificial intelligence still means ChatGPT – asking a question, writing an email, improving a resume, generating an image or summarising an article.

Those are useful applications. But working in San Francisco’s AI ecosystem, I see AI moving beyond something you chat with to something that can carry out entire processes.

Think about a typical salesperson. Their day is spent researching potential customers, writing personalised messages, sending follow-ups, answering questions, booking meetings and updating customer records. That is not one task but an entire workflow.

Now imagine an AI agent carrying out that same process, not for one person at a time but across hundreds or even thousands of prospects simultaneously. That is where AI starts to feel different.

A chatbot answers; an AI agent executes. Instead of responding to a single prompt, it can move through an entire workflow with very little human involvement.

A traditional chatbot might tell you a refund is being processed. An AI agent can check whether it is allowed, process the return, update company records, and involve a human only if something unusual happens.

In the United States, companies are already building AI agents that research sales leads, handle customer support, process returns and book appointments automatically. This pattern is emerging across industries, whether in healthcare, banking, logistics, education or construction.

This matters for Malaysia because many of our companies depend on exactly these kinds of repetitive processes.

Imagine a private medical centre where AI schedules appointments, confirms insurance details, sends reminders and prepares patient records before the doctor walks into the consultation room. These are not futuristic ideas but business workflows AI companies overseas are already building.

If businesses elsewhere begin automating these workflows while we limit AI to writing captions, emails, presentations or resumes, the gap will gradually widen. The real risk is not simply that other countries build better AI tools – it is that they use AI to run their businesses more efficiently than we do.

Malaysia is equipped to compete by applying AI to the industries we understand best. Local companies know their customers, languages, regulations and business challenges far better than overseas developers ever will.

Humans will still matter for judgement, empathy, trust and accountability. But as repetitive work becomes increasingly automated, the opportunity is to design better workflows, supervise AI effectively, and remain responsible for the final outcome.

AI will not wait for every company to feel ready. For Malaysia, the question is no longer whether we should adopt AI, but whether we use it simply to work faster or to rethink how work itself gets done.

That difference may determine which companies stay competitive.

Ian Ku is a founding engineer of ArchiBoost AI and an FMT reader.

The views expressed are those of the writer and do not necessarily reflect those of FMT.