Entry-Level Jobs Singapore: PM Wong's Career Ladder Warning

Prime Minister Lawrence Wong closed the inaugural DBS Asia Future Forward Conference on Monday, 12 October 2026, in a dialogue with DBS Group CEO Tan Su Shan in front of more than 800 business people. The theme was intentional AI, but most of the hour circled one awkward question: when AI does the junior work, where do young people learn to do the job? The Prime Minister's Office published the full transcript that day, and CNA's report led on the same worry.

Hours earlier, President Tharman Shanmugaratnam made a matching point from Bangkok, at the IMF and World Bank annual meetings. Two of Singapore's most senior office-holders, one day, one warning about first jobs. That matters if you are graduating, hiring, or managing anyone under 30.

Every career has a first rung, and it is dull. Taking minutes. Drafting the first version of a memo. Pulling figures into slides. Nobody enjoys it, and almost nobody admits it, but that is where juniors learn how an industry actually thinks.

Generative AI now does much of that in seconds. Mr Wong said it plainly: the routine work is often mundane, yet a lot of tacit learning comes from it, and "AI will replace all of that routine work, but we should be careful not to automate away the learning that comes from it." He listed his own examples from his years as a young officer, taking minutes, drafting speeches, preparing PowerPoint slides for his bosses. Not glamorous, but the process taught him how the system worked.

His worry, in one sentence: if the first rung disappears, the ladder above it turns shaky. As he put it, "we are concerned that there may be that bottom rung of the career ladder that is broken, and young people may not get sufficient training."

This is not the first time a senior Singapore official has framed it this way. In September, Minister for Health Ong Ye Kung told students at NUS that AI mostly removes tasks rather than whole jobs, and that new roles keep forming around the work that remains. The bottom rung is being rebuilt, not simply deleted.

Mr Wong said the government is studying this sector by sector, because the answer for doctors in training will not be the answer for lawyers, bankers or engineers. It is working with tripartite partners and, in his words, "looking at whether there is a need for new programmes, new schemes." No scheme has been announced yet. For the wider jobs picture, our earlier piece covers what the Economic Strategy Review found about AI and jobs in Singapore. If you are waiting for a new grant to appear, read that as a direction of travel, not a date in your calendar.

The message to employers was blunter: hiring managers need to be more deliberate about training and mentoring. He also pushed back on the idea that AI literacy is the whole prize. Employers still want people who think on their own feet and exercise judgment, which is why schools should keep drilling the fundamentals. His comparison: we all learned maths with pen and paper, long after the calculator arrived.

On timing he was more hopeful than some headlines suggest. Factories took about 30 years to convert from steam to electricity. He does not think AI will take 30 years to be deployed at scale, but if the shift spreads over 10 years or more, Singapore gets room to retrain people. He was equally clear this is not a promise of zero retrenchments. Where they happen, they should come with advance notice so unions and government programmes can help.

President Tharman, on a panel about AI and jobs with IMF Managing Director Kristalina Georgieva and World Bank President Ajay Banga, said hiring demand for fresh graduates is weakening, most sharply in sectors exposed to AI. The knock-on effect is his real concern: without that first foothold, graduates wait longer, then compete with the next graduating cohort, and skills and confidence erode in the waiting.

His fix is joint. "There has to be a public-private collaboration, with the public sector incentivising the private sector to take in fresh graduates," he said, arguing for lasting measures rather than a short-term crisis response. He also wants universities to lean harder into applied learning through work-study programmes, attachments and internships, and told the room not to fixate on one model.

The dialogue also covered AI safety. Mr Wong said the gap is no longer awareness but control: the risk of rogue actors and rogue agents wielding powerful systems, and what the industry calls P(doom), the probability of a disastrous outcome for humanity.

In theory, AI safety would run on international rules the way aviation, finance and civilian nuclear power do. He was honest that this will not happen soon. His practical offer: start voluntary, then tighten. "You could have an international governing board that is independent, not selected by the AI labs themselves but independently appointed," he said, alongside an independent audit process, with requirements that stiffen over time as investors and customers start demanding the standard. Singapore's proposal for a UN framework convention on AI safeguards runs along the same track.

He closed that thread with three non-negotiables: safety and security, human responsibility with people in the loop for critical calls, and benefits spread broadly. The line worth keeping is this: "We are not trying to be an AI-driven society. We are trying to be a better society enabled by AI."

Two practical moves follow from the speech. First, treat AI as a supervisor rather than a ghostwriter, letting it critique your work instead of producing it, so you still build judgment. Second, if you manage juniors, audit their tasks now, before the routine work quietly disappears, and redesign the role around the thinking that remains.

There is money in that second skill. An Institute for Adult Learning commentary in The Business Times in July 2026 noted that AI supervision, a step above AI fluency, already earns a premium in Singapore, with some fresh hires in these roles starting between S$70,000 and S$90,000 a year.

If you want the formal route, the SkillsFuture AI Subscription gives six months of a premium AI tool after one eligible course.

Use the prompts below with ChatGPT, Gemini or whichever tool your workplace pays for. Swap the details in [brackets].

Click the copy button on any block and paste it into your AI tool.

💼 For fresh graduates and juniors

I am a fresh graduate in [your field] and my team uses AI for routine first drafts. Act as a strict but fair senior colleague. I will paste my own draft first. Critique it line by line, tell me what a reviewer would flag, and ask me 3 questions I should be able to answer before I submit it. Do not rewrite it for me.

📋 For managers of junior staff

Here is a list of tasks a first-year [role] in my team handles: [paste list]. For each task, tell me (a) whether AI can now do most of it, (b) what the person learns by doing it manually, and (c) how to redesign the task so AI handles the routine part while the junior still does the thinking. Then give me a one-page coaching plan for this quarter.

🎓 For anyone worried about their own role

I am a [job title] in Singapore with [number] years of experience. List 5 tasks in my job that AI can now produce a first draft of. For each one, write the review checklist I must apply before I sign off. Then suggest 3 skills that would make me harder to replace in the next two years, with a free or low-cost Singapore course or resource for each.