# Age of AI: why young Singaporeans should study humanities

- Date: 2026-11-05
- Added: 2026-11-06
- Who: CNA (CNA report) · CNA
- Source: https://www.youtube.com/watch?v=mbFpeLSat6U
- sgai: https://sgai.md/videos/v134/
- License: sgai-authored content (summaries, translations, analysis) is CC BY 4.0 — attribute and link to sgai.md. Verbatim source text (Hansard, speeches, transcripts, policy documents) remains © its original rights holders and is reproduced for reference only. Terms: https://github.com/meltflake/sgai/blob/main/DATA-LICENSE.md

## Why it matters

Guests recommend humanities education to help students assess values in AI applications.

## Summary

Analysis of why humanities education matters for young Singaporeans in the age of AI.

## Full text

© CNA — reproduced for reference only.

We keep hearing that AI will take away some jobs, but it will also create new ones. Is it as simple as it sounds? >> The labor force is going to shift towards, you know, basically less work for us to do. >> Really? >> And more leisure. >> So, some students who get an A, for example, they're not certain why [music] they got an A. >> Oh. >> [laughter] >> Right. And then that becomes worrying for me. >> Who benefits from the technological progress? Um I'm very concerned about that. Otherwise, we would just see greater inequality. >> [music] >> So, hey everyone. You know, last week we were talking about AI and whether we are moving to embrace AI too quickly without quite fully understanding the risks and whether we can even control those AI agents going rogue.

>> And this week we are talking about how AI will affect our jobs and our critical thinking skills. And who stands to benefit from all of this? Everyone or just a select few? Here's the rest of the conversation. There's also this concern that, you know, AI is also currently in the hands of very powerful, uh small group of, you know, very powerful foreign companies like your Open AI and Anthropic. But I've also recently learned that, you know, a lot of new agents that are coming online every day. Um if I guess, you know, if Singapore is concerned about putting, you know, private data into the hands of, you know, these kind of companies, what can the country then do to help its citizens keep abreast or keep, you know, a pace with AI developments? >> Um I think there are two layers to our question.

One is, um you know, at the country level, right? Geopolitics are also relevant, privacy issues. Um for that, um there's a concept called uh ontological AI. >> Wow. >> Meaning when we use these AI agents or tools, foundational models, how do we adapt them to the local environment? >> Okay. >> Ensuring privacy, limited, well-governed access to database, right? Within the company, within organization. There is active development in that field. I think we'll become better and better at doing that and that would help safeguard data, you know, privacy and developing more local proprietary AI tools for Singapore. >> that at the moment? >> Yes, in a way. But even for ontological AI, that's in its early phase. There's another issue in what you just mentioned, right? The profits going to the large companies and all that.

Which is distributional issue in the society. Who benefits from the technological progress? I'm very concerned about that. I think there are two big problems. One is business owners benefit. And and increasingly disproportionately more. So we've got to find a way for ordinary citizen to either become business owners or find a way to share the the the joint surplus and benefits that technology creates. Otherwise, we'll just see greater inequality. That's one. And with firms, if you look at the data, firms with more than 500 employees in Singapore, about I think over about 75 3/4 of them adopt AI. You look at the smaller firms, the rate goes back to 27%. So there is something that I call a barbell economy. So the largest firms due to economy of scale, they benefit a lot from this technology.

The individuals, the influencers, now you can be a movie producer. You can be a very successful content creator online. Individuals also benefit. They can do things that they couldn't do with the help of AI. It's the small size enterprises who are stuck in the middle. >> Okay. >> They can't fully, you know, implement the the the technology and the agents. So they lose relative to bigger firms. At the same time, they don't benefit too much from the customized content creation per se relative to the influencers and individuals. So, they are stuck in the middle. And I haven't seen much discussion on how we mitigate that. >> Mhm. >> I think that should go into the policy. It's just a question of what do we do about it? >> Um that will also affect the distribution of wealth um within the economy.

>> But I mean, distribution of wealth, surely none of these companies are going to say, "Hey, let me share my profits with you. " Could we then say everyone is entitled to their own personalized AI agent? Would that be a a start? >> So, uh you know, there are traditional economic policy tools, such as uh taxation. Right? That's one way to redistribute wealth. But what you suggested, I I think is a reasonable instrument. Um you know, at NTU, uh we are actually allowing the undergrad to train their own agents, their own team over the four years they are in university. They can basically train them to be research assistants. >> Mhm. >> Uh a team, right? And when they graduate, they can bring them uh with the students.

Uh I'm not saying that's the best implementation of this idea, but in general, if you can allow some ownership of agents, and to the extent that the agents can generate dividend and economic profits, that's one way to at least attribute some benefit of the technology to you know, the the citizens. >> Right. >> Uh as opposed to just to the CEOs, CFOs, or shareholders, or big companies. >> And I want to come back to this idea of jobs, because I think a lot of people are very worried about, you know, displacements um you know, from their jobs because of AI. Um In fact, a few weeks ago, PM Wong also mentioned it in a conference that the government is worried that AI may remove some of the routine tasks that can help train younger workers. Um and also, we keep hearing that AI will take away some jobs, but it will also create new ones.

Um is it as simple as it sounds? You know, upskill, learn AI, and then you will be able to do something else. Is it as simple as it? >> I think MOM just released the statistics, right? Um you know, um I think 8% of the companies actually stopped hiring more people and 6% are actually firing, reducing their head counts. >> Mhm. >> But then if you look at the related statistics, 19% is co-designing or redesigning the workflow so that humans can play a role together with the AI tools and agents. So there's both the positive and active side in terms of, you know, labor market impact. >> there's still some level of basic skills that we all need. So when I look at my undergrads, for example, at this point in time, uh, they are encouraged to use AI in class, right? And do the necessary research, synthesize information.

But that's also part of helping them to transit to the workforce where they will be using it anyway. Right? So I see the benefits in actually getting them up to speed in schools. Um but there are certain basic levels of skills that they ought to have as a student to begin with. And I do see that that is being a little compromised along the way already. Right? So for example, knowing what sources to look for, knowing how to do even simple citations, writing, synthesizing information. So I do see that some of these skills that you expect of a normal undergrad is giving way. Right? Because all these are already done by the AI agent. Um, I mean back then during our generation, maybe it was Wikipedia, right? That framed certain thoughts for us, certain ideas for us.

Um, now I'm not even sure students have that ability to evaluate what really is going on before they accept the answer and do whatever edits they need to do. So that's what I worry about the very basic level of skills that they need to have. So when we talk about, um, us being still being the human agent making the final decision, do we know what we are approving or rejecting? We may not have those judgments anymore. So if you think about professional skills like the doctors, for example, if you're a trainee doctor and and you're you know, going on and looking for, okay, let Let type in some of the symptoms, What's the diagnosis here? You you probably need certain experience, right? And certain exposure, make certain mistakes. I think that's all part of learning. >> Right. >> Um I really I do see that happening in my students.

So some students who get an A for example, they're not certain why they got an A. >> Oh. >> [laughter] >> Right. Then that becomes worrying for me. >> Yeah, right. >> They will come out to you and say, you [clears throat] know, prof, I what why did you give me this? >> Thanks, but how come? >> Yeah, like oh, I've got an A and I and and I'll be asking like, do you know how you did it? >> Mhm. >> [laughter] >> So that becomes a worrying trend, right? That they're not even sure why they got an A. They're not sure why they made a mistake as well. >> So is critical thinking lacking then? >> Oh, I think definitely critical thinking, but the way we think about critical thinking is not just criticize, but it needs to have certain kind of foundational attitude, right? Certain also foundational skills that identifying.

Uh but you only get it through practice and experience. >> So you're saying we become a bit too dependent on this technology. >> so. So there are certain intellectual virtues that we should have. >> again, I mean now I don't even count. I just use my calculator on my phone, right? When we're splitting the bill. So we've become lazy that way. And chances are we will become a bit lazier because AI will do more for us. >> There is a book um titled Amusing Ourselves to Death by Neil Postman. Um he's trying to say, you know, it's not the 1984 world that we worry about. It's the Brave New World that we worry about, right? All this convenience, entertainment take over. So that we lose our um you know, ability to debate, to write properly. There are certain skills that we're losing. And that's happening at massive scale with AI.

Um in a way, so he was talking about TV back then. And we we're fine. We lose something, right? But uh you know, we're coping fine. >> I've learned so much from watching TV. [laughter] >> But but uh because this is happening at massive scale, I think that's reason for alert because we don't want to yield this to AI. We still want to co-pilot. So I think one critical skill what Jennifer was mentioning concerns verification. Whatever content that AI is producing, whatever work that's completed by AI, is there a way for us to, um, you know, assess that and verify and be content and happy and assured that, you know, it's done. It's completed. I think that skill is going to be more and more valuable at least for a fraction of the population. Uh, we've just got to train um humans. Unless, you know, we decide, "Okay, we're actually giving up.

We'll just, [laughter] you know, enjoy life. " >> people who are currently in the workforce and they, you know, are worried about losing their jobs and some [clears throat] of these realities are, you know, is is real, right? It's a it's a everyday occurrence. We see it, you know, in the news as well. Um, is this just a transition period? Is it just a temporary disruption where, you know, there will be people who will lose their jobs. Once they upskill, they get the AI, you know, skills, they will be able to find something else. Or is it Are we going to be seriously looking at a generation of people who will perhaps, you know, find it hard to adopt, keep pace, and will just have to say, "You know what? Maybe I will have to go and do, you know, something that requires my hands and feet instead. " >> I think this is not so transitory.

It's going to be a, uh, persistent trend. What is the trend? Uh, the labor force is going to shift towards, you know, basically less work for us to do. >> Really? >> And more leisure. So, that that's >> That sounds good. >> [laughter] >> That should be good provided we distribute that benefit properly. >> Okay. >> Rather than just a few a small group taking taking, you know, advantage of that. >> should have an agent. >> [laughter] >> I have to disagree on the fact that, you know, there will be less work for us because I I find that, um, because I have the tools, I am to do more in the in the short time that I have. So, suddenly I'm not you know because my research now has to be supercharged, super speed. >> There is the formal as well. For us as well, we we do worry about oh, they are doing research so quickly.

How do we get to the publication [laughter] faster? >> So, you'll be expected to turn out five research papers in a year. >> on auto research. We're trying to displace ourselves, right? In a way. But, um in general, you know, we we did uh get displaced from physical labor. I think with mental and intellectual work, AI is going through that process. There will There will be reduced work. That doesn't mean we are becoming less intellectual. There are certain things and meanings in life we can still pursue provided we design it properly. But, the thing is we don't design it properly. In fact, I worry a bit more about the youth, the younger generation, than the existing, you know, workers who might lose job and all because these workers are already trained.

Many of them do have the verification skill, which will come back to be very important. Yes, AI is replacing their job. At one point, people don't trust the work produced. And they have to come back. But, with with the younger ones, they haven't gone through the process to develop that verification skill. And they may not even have the opportunity to do that. That's more concerning and uh it's harder for them to to even get um you know, a first job. >> Right. >> I suppose it's ironic. >> All right, because as AI gets better at the job, >> Yeah. >> the more we will trust AI, right? So, there will be a automation bias, which means to say I probably don't need to use so much of my own verification skills because the AI has gotten so much better at verifying their own results.

So, over time, I may actually work myself out of a job or particular skills that could be unique to me. I may not have or may not need that skill anymore as AI gets better. >> Mhm. Cuz you know, the government recently mentioned these 3,000 year-long internships. Not 3,000 years, but >> [laughter] >> 3,000 of these year-long internships is the solution to try and get them to do all that work now when they are still studying so that they can do all that bottom, you know, rung minion task and then when they enter the workforce, they will be able to, you know, come in at a little bit of a higher level that may not require that critical thinking [laughter] anymore. >> I think it's effective temporary solution. I don't think it solves the fundamental problem.

Uh if anything, I would encourage the students to, you know, it's fine to do internship to acquire some skill to make them competitive in the in the job market. But at the same time, I would highly encourage them to actually study humanities, study arts. >> Yeah. >> Do these things that either allow them to enjoy their free time if they have less work to do at some point or um give them a irreplaceable skill in future when we call for people to define our values for the society, for the group. When we call for people to better align AI agents and uh you know, all these algorithms. I think that's a very skill. Um A- a- and again, you know, I'm a financial economist. We we often talk about portfolio approach, right? We don't, you know, put all the eggs in one basket. >> Sure. >> They should do some internship. I think it's helpful.

But at the same time, they should think or try to build out some long-term skills to enjoy life, to to define meaningful life to define meaning for technology. Um that's going to be very helpful. >> So you're saying that the ones who will save us all are the people who will go and study humanities and try and teach technology human values because essentially that's the That's the only saving grace here. But they do still have to teach the AI because [laughter] the AI is going to take over, right? And they will take over. AI will take over, but we want to try to not let AI kill us all which is where the humanities come >> If you want all us to have our left and right brain working >> just as actively. So, we need all the tech to keep us going and to survive in this world with the new operating system that's just been updated.

At the same time, we need to know where to draw the line, the ethics, when to say yes, when to say no, and when to hold back. Um and you can only do that if you, I guess, grow up a certain way and allow yourself to experience certain things, right? Which is getting harder as we can see from all this, right? >> I mean, it all depends on our ultimate goal, right? If the ultimate goal is, let's say, sustainability on Earth, yeah, technological progress, maybe the technology side will drive everything. But if we still care about humans, you know, uh you know, we don't want to become uh or be replaced by silicon-based systems. Um I I I do think we have to take this uh well-balanced interdisciplinary approach okay rather than, you know, worshipping technology, worshipping, you know, all these things.

>> Five years from now, can we look back and say, "Hey, okay, quite balanced, we managed to navigate this. " Uh what do you think? In 5 years from now, you know, how how would you know whether Singapore has handled this AI transition well? >> I think the maturity actually of the society itself in having all these conversation and discourses. >> Yeah. >> Um at this point in time, not sure how many people are actually interested in talking about it, right? So, there are a lot of worries and fears, very practical ones, um jobs, education, for example, but the broader conversation, right, about humanity, about our values, about practices. So, for example, if we are introducing a lot of care bots into the healthcare system, hospitals, what about care? >> Sorry, what are care bots? >> Care bots are robots that >> Robotic nurses.

>> Uh okay, nurses um that probably go around distributing medication, for example, right? Um but it does affect a lot of our values of care, right? It does also affect our practice of care. Um you know, just giving out medication means I could observe, I could build rapport. I could see why certain people, certain patients are resistant to certain medication and then take appropriate actions. But if I want to optimize, right, my manpower, I want to make sure that I get my nurses to do something else that I think is more higher level. We have an impact on what care might mean and what care looks like at the end of the day. So I do think on a broader perspective having to understand this is actually rather important. So I definitely agree with you. Yes, take more humanities, right?

>> [laughter] >> And know where the human values lies so that you could eventually align. >> And there is a general piece of advice to not listen to predictions given by economists. So I'm not going to give a prediction. >> [laughter] >> On this podcast. >> Right. So I'm not going to give a prediction, but what I will do is I will go back and run my simulation. >> [laughter] >> Using AI or without AI? >> With AI, of course. >> Okay. >> Yeah. >> Okay, well, thank you so much for coming in and sharing these insights with us. I think it's been a, you know, a conversation of a big issue that I think is not going to be resolved anytime soon. So if you have any thoughts you want to share with us, do chime in. We do love to hear from you. But otherwise, that's it for this week's episode of Deep Dive. Bye for now.
