Your Degree Is Not Enough Any More: Surviving the AI Job Market

The work juniors used to learn on is the work AI does best

Think about what fresh graduates have been hired to do for the last decade. You are not making million-ringgit decisions. You research, prepare slides, write reports, do basic coding or scripting, maintain spreadsheets, reply to emails, prepare documents, and generally do whatever your manager does not want to do. You learn along the way and, over a few years, become ready for harder work. That is how careers have worked everywhere.

The problem is that this simple work is exactly where AI is already better than a fresh graduate, and where it is improving fastest. If I used to need three junior staff and now one employee who is good with AI can do most of it, why would I hire three?

I run a company, so I am telling you this from the employer's side. When a tool saves my team five hours, I use it. I am not going to make my company less productive so I can hire more people, and neither will most companies. We cover this shift in more depth in AI took the bottom rung.

Stop asking whether your career is AI-proof

I do not think an AI-proof career exists. Lawyers, accountants, marketers, developers, HR, finance and operations will all use AI. It will be part and parcel of running a business. The better question is: what kind of person becomes more valuable as AI gets better? Within the same profession you can already see two very different types emerging.

Everyone has the ticket now, so the queue is long

This is why "just get a degree and you will be fine" is more outdated than ever. Our parents grew up in a different economy. A decent degree had real signalling value: it was your ticket into a job, and the queue was short. Today everyone has a ticket, so the queue is very long. And everyone has the same access to AI, so simply using AI does not make you stand out either.

Picture yourself as the employer interviewing two business graduates with similar grades. One says: "During my internship, my team spent hours every week preparing a report for managers. I automated it into a dashboard that pulls from several data sources, and now it takes five minutes. Here is how I check the data is accurate, and here is how errors get flagged." Who do you think I am more interested in?

Notice I do not care whether that person is an "AI expert". Too many people are trying to become AI gurus, and half of LinkedIn is now an AI strategist. You do not need that. You need to be good at your actual job and know how to use AI inside it. A marketer who uses AI well. A programmer who uses AI well, not one who leans on it as a crutch.

What do you do with the three hours AI saves you?

If AI makes you twice as productive and you spend the saved time relaxing, that gives you no leverage and no protection. The person who becomes valuable thinks: this report used to take four hours, now it takes four minutes, so what bigger problem can I solve? Can I talk to customers? Work out why revenue dropped? Improve the website? Automate another process? Help sales? Build something?

This is the shift young people need to make: stop thinking about tasks and start thinking about outcomes. The slide is a task. The decision it enables is the outcome. AI makes tasks cheap, so the person who understands the outcome they were hired for becomes the most valuable person in the room. Your next best competitor is not thinking "great, I can chill now".

The soft skills people laughed at

Communication, sales, leadership, building connections, reading a room, knowing when something feels wrong. Schools barely teach these, and people used to dismiss them. They matter far more now.

Imagine two people with the same AI. Both generate a decent report. One sends it to the boss: "Done." The other reads it, notices a number that looks off, checks the data, realises the AI misunderstood something, fixes it, then walks into the meeting, explains clearly and convinces everyone what to do next. I would pay that second person double. And when one capable person can do the work of ten, I can afford to.

Things that sound impressive today, like using AI to build a dashboard faster, will become entry level within a few years, because every new graduate will know how. The better report is not the moat. The judgement after reading the report is.

Be careful: AI can make you stupid

I use AI every single day and I love these tools. But if every time you do not understand something you ask AI to do it, you become extremely good at producing work you do not understand. It is most obvious in coding, but I see it in video editing, copywriting and proposals too. There is a lot of slop, and to anyone who works with AI daily it is painfully obvious when someone just pasted a prompt's output.

Someone can one-shot an app with AI. Impressive, but a low-level skill. And what happens when it breaks? Fixing it requires fundamentals: enough understanding to debug, question the AI and notice when it is going in the wrong direction. In a world where everyone gets an answer in minutes, the important skill is knowing whether the answer is rubbish. If this resonates, read don't become dependent on AI.

That is why one of the most valuable skills right now is learning how to learn. How fast can you enter a topic you know nothing about, grasp the principles, ask good questions, practise and become useful? Your exact skills today may not be your skills in five years.

Depth, connected dots, and people knowing you exist

Careers used to reward being narrowly specialised. I think the most valuable people now connect fields: a marketer who can analyse data with AI, an engineer who understands the whole sales funnel, a designer who understands psychology, business and analytics. But I hate the advice to learn 15 skills and be average at all of them. Be T-shaped: go deep on one thing you are genuinely good at, then use AI to expand around it. With no depth anywhere, you have no judgement.

The last thing young Malaysians underestimate is distribution: do people even know you exist? AI increases the volume of everything: content, software, freelancers, applicants. Being good may not be enough; people need to know you are good. You do not need 100,000 followers. Two thousand people in your industry who know your name and what you do is already powerful. Post your projects, share what you learned, write about obstacles you overcame.

And do not write "experienced with AI" on your CV. Everyone will. Show it: I automated this, I built this, I cut this process from X to Y, I increased this number. AI makes everyone sound smart, so you need proof to stand out. Our guide to portfolio projects that get interviews is a good place to start.

What I would do at 18, 21 or 25

Pick one field I actually want to be good at. Learn its fundamentals properly. Use AI aggressively inside it, automate the boring work, and spend the saved time on harder problems and new skills. Deliberately work on communication, judgement and understanding people. Put my work somewhere others can see it.

I would not spend three years panicking every time someone on X says software engineering, marketing or design is dead. Nobody knows exactly which jobs will exist in 2035. Everyone is guessing. But one thing is obvious: if your entire value is predictable tasks AI keeps getting better at, you should be worried. If you understand your field deeply, use AI competently, make good decisions, work well with people and own outcomes, you will be fine. You probably have more leverage than any generation before you. Our parents went to the library for information. You have it at your fingertips.

The goal is not to find a role AI cannot touch. It is to become the kind of person who gets more valuable every time AI gets better. If you want to build that kind of depth in software, with AI as a multiplier rather than a crutch, try our free trial or look at the full AI-Native Software Development Programme.