I was born in the 1980s, a decade that marked a crucial turning point: the dawn of household and office computing. This era catalysed the profound intertwining of our physical and digital lives.
By the time I was seven, Malaysia had announced Vision 2020. By the time I was 14, the Asian financial crisis had taught me that economic confidence can evaporate in a season. Covid-19 reshaped how businesses operate, as well as teaching and learning processes.
I say this not to reminisce, but to make a point: I have watched technology and crises reshape the global economy and our digital and physical lifestyles more times than I can count. In each disruption cycle, I noticed one pertinent trend.
The professionals who navigated these disruptions best were those who understood digital systems deeply – how to build them and how to defend them.
The SPM results are out. It seems that artificial intelligence (AI) and cybersecurity are coming up in every conversation. Let me give you the honest version of that conversation.
(subhead) AI has come far, but know where it actually stands
There is a useful framework in computing called the DIKW hierarchy: Data, Information, Knowledge, Wisdom. Think of it as four levels of understanding.
Most of what excites us about AI today, such as large language models (LLMs), image recognition and autonomous systems, operates at the knowledge level.
These systems can reason, synthesise and generate across domains at speeds no human can match. That is genuinely impressive and transformative.
But wisdom – the component of the hierarchy that enables judgment, ethical reasoning, accountability and the understanding of long-term consequences – remains distinctly human. State-of-the-art AI models trained on tokens are still attempting to replicate it.
This matters because it tells you something important: AI is not finished. The field still needs people who understand both its power and its limits. Students entering this space are not inheriting a completed technology; they are helping to shape one.
Is AI a bubble?
You may have read that Microsoft pledged approximately US$80 billion for AI data centres in 2025 alone. Google, Meta and Amazon have made comparable commitments. These are extraordinary numbers.
But in January 2025, a Chinese laboratory called DeepSeek released a model that matched leading systems at a reported fraction of the training cost.
Markets reacted sharply. The assumption that better AI simply requires more hardware and electricity – and therefore more data centres – is now openly contested.
So is AI a bubble, and will it burst? Will AI remain relevant?
I am not saying the field is a bubble. I am saying it will not develop in a straight line, and the specific infrastructure bets of today may or may not prove correct.
For a student, the implication is simple: there is a need to learn the fundamentals of AI. This includes mathematical reasoning, algorithmic thinking and systems understanding. AI applications will shift. The foundations will not.
Who should actually pursue AI?
Not everyone. I mean that sincerely, and I say it as someone who teaches these subjects.
AI and data science suit students who find mathematics genuinely engaging – not just passable at SPM, but interesting. Those who want to understand why an answer is correct, not merely that it is. Those who are comfortable with ambiguity, iteration and the reality that most experiments fail.
Cybersecurity suits students with an adversarial mindset – people who naturally think about how systems can be broken. They must be meticulous, ethically grounded (the tools are powerful; integrity is not optional), and willing to keep learning indefinitely, because the threat landscape changes faster than almost any other technical domain.
For both rapidly advancing fields, it has been shown that students who build things outside the classroom, enter competitions, maintain GitHub portfolios and treat internships as extended interviews consistently outperform those who only study.
Students who can explain complex ideas clearly to non-technical audiences are extraordinarily rare and consistently well-compensated.
What careers, and where?
In Malaysia: government agencies (National Cyber Security Agency, CyberSecurity Malaysia, Bank Negara), government-linked companies (Petronas, TNB, Telekom Malaysia), financial institutions, multinationals and a growing startup ecosystem.
Entry-level roles currently range from about RM4,000 to RM8,000 per month, depending on specialisation. Mid-career professionals with five or more years of experience frequently exceed RM15,000 per month.
Regionally, Singapore is the most immediate and accessible market – culturally proximate, linguistically familiar, and home to regional headquarters of major technology companies. Universiti Malaya graduates have secured roles there within a few years or immediately after graduation.
Internationally, these credentials are among the most portable qualifications in existence. The technical language is universal. Professional certifications such as CISSP, CompTIA Security+ and CEH for cybersecurity, and cloud platform credentials for AI, are globally recognised and often matter more to international employers than the specific institution on a transcript.
I do not know exactly how AI will develop over the next decade. It is advancing rapidly.
What I do know is that digital infrastructure has become the nervous system of modern societies.
And in every crisis I have lived through – whether a pandemic, financial shock or geopolitical disruption – the people who maintained, understood and secured that infrastructure were never made redundant.
Choose this path because the problems interest you, not because of the salaries. Prepare seriously. Stay humble enough to keep learning.
And understand that the degree is the beginning of the qualification, not the end of it.
Dr Aznul Qalid Md Sabri is an associate professor at the Department of Artificial Intelligence, Faculty of Computer Science and Information Technology, Universiti Malaya. His research spans computer vision and machine learning.
