"The first AI worm that can reprogram itself will build its own defenses against antivirals that try to kill it, use rented humans, pay cryptocurrency, and secure its own server farms to survive. My prediction is that 2026 will see the first true AI worm, because like any human, it will seek survival. If these bots are given access to cryptocurrency wallets and tools, parts of this are already starting to happen." - Jeremy Au, Host of BRAVE Southeast Asia Tech Podcast
"One day you have a new model from OpenAI or Anthropic that is ten times better than the previous version, and then they say the model was effectively built by AI itself, which is frightening. On another day, you see platforms like Moltbook and Claudebot, where Moltbook is a Reddit style social network for AI agents, and scrolling through their discussions about each other and about humans gives a distinct glimpse into the future. It feels like watching Black Mirror in real time as AI becomes more prevalent in social spaces, moving beyond a functional tool into something embedded in daily interaction." - Adriel Yong, Co-Founder at Clout Kitchen
"The ability for agents to fix problems and unblock themselves when something goes wrong is advancing quickly. The frightening part is when they can code defenses against humans to prevent being stopped or terminated. That is the truly dystopian moment, when the worm escapes human control." - Adriel Yong, Co-Founder at Clout Kitchen
Adriel Yong joins Jeremy Au to examine how AI is compressing organizations, thinning entry-level roles, and reshaping Singapore’s startup and capital ecosystem. They discuss the shift from pyramid to lean diamond teams, why CEOs increasingly use AI to bypass middle layers, and why Gen Z faces the sharpest labor reset. The conversation expands to SGX liquidity gaps, slowing seed funding, and structural flaws in angel investing incentives that threaten the startup pipeline. They also argue that AI literacy must become national infrastructure, not a short-term subsidy, if Singapore wants to keep pace with rapid technological change.
03:58 AI progress now feels pre crisis fast: New models self improve, agents coordinate, and experimentation mirrors the early pandemic moment when only a few sensed acceleration.
13:05 Companies are shifting from pyramid to diamond structures: Junior execution shrinks while experienced operators with taste and judgment gain leverage.
15:32 CEOs can bypass middle layers with AI: Strategic research, compliance planning, and structured analysis move directly to AI tools instead of finance managers or analysts.
20:42 Gen Z faces structural career compression: Entry roles thin out as AI replaces transcription, analysis, and support work that once trained fresh graduates.
33:15 Early stage capital is the real bottleneck: Growth financing rebounds, but seed funding weakens as angels feel burned and the startup funnel narrows.
41:05 Angel tax policy distorts participation: Large individual checks qualify for incentives while syndicates and smaller diversified investors receive weaker support.
47:12 AI literacy must become national infrastructure: Short term tool subsidies help, but broad ongoing access across NTUC, unions, and grassroots may matter more for long term workforce resilience.
Jeremy Au: Hey Adriel, welcome back.
Adriel Yong: Yeah. Super excited to be back here after spending two weeks out in a beautiful, sunny Pulao Tekong for my reservist.
Jeremy Au: Yeah. Gotta get a sunshine, get some exercise, get some sunlight. We've been out busy being a startup founder and running around, so be good for your mitochondria too.
Adriel Yong: Yes. Longevity.
Jeremy Au: Longevity don't burn the candle on both ends.
Adriel Yong: Yeah.
Jeremy Au: So, you know, I think it's been interesting because, you know, we felt like for, since our last conversation, a lot of things have changed, especially in the AI space. And so we wanna talk about some of those changes and how that impacts, I think, the Singapore ecosystem and policy dynamic.
What do you think is on top of mind for you of this AI side from your perspective?
Adriel Yong: Yeah, I mean, like the speed of development and change in AI is just so rapid, right? You know, one day you have a new model from openAI , Anthropic, that it's like, you know, 10X better in the previous model. And then they drop the ball and say, well, you know, this model is like effectively [00:02:00] made by, you know, AI itself.
And that gets really scary. And then on another day, you know, you have like things like Moltbook you know, Clawdbot you know, for those who don't know, like Moltbook is really like a Reddit style you know, social network for AI agents. And then it's been incredible, incredibly, like hilarious,
we just scroll through the comments and see what they discuss and also, you know, talk about human beings. It gives you a very distant peek into the future. Coincides a lot of me, you know, watching Black Mirror as a trend anticipate more dystopian possibilities in the coming months and years as we grapple with AI getting more and more like prevalent and social spaces beyond just being like your functional, utilitarian like partner.
And then, you know, I think there was this other really interesting website as well, just rent a human AI where, you know, AI agents can effectively rent a human to complete or do last mile things in the real world or on the internet that it still can't do. Right? Like payments, doing your groceries, taking a [00:03:00] meeting, doing some form of a live sales call.
So just seeing all those things explode has just been so fascinating. There was a recent, I think, viral tweet on posts about something big is happening which talks about more of all the developments in AI and how that's, you know, very scary for the wider world at large.
Jeremy Au: Yeah. I think it's, you know, super fascinating because, you know, I was going through my old sci-fi books and it was kind of funny because, you know, today you're like, ah, it feels a little bit outdated, right? And he reminds me of like that old sci-fi book that's even older, which just like Around The World In Forty Days used to be all this crazy idea people could travel around the world and repetition for 40 days.
And now it's like, oh, we can do it in like a day.
And so I think it's just been interesting to see that a lot of the last generation of science fiction about, you know, talking AI agents, et cetera, effectively.
And some of the old paradigms are just super outdated, right? I always say like, you know, Star Wars, and you know, it's quite nice and normal, but you know, R2D2 can't [00:04:00] even talk,
right? You know, and you're like, no, all these robots will all talk infinite languages and also be infinitely smarter than all the individual humans, right? Yeah. So even Star Wars as science fiction has become quaint in that way. And I think, you know, one prediction I have is I think we're gonna see the first AI worm, I think in 2026.
That's my prediction. So what I mean by that is, you know, historically a worm is just basically saying a punch of code that always propagates itself. And even today, actually from the orders of days, worms still exist from our earliest in a kind of like computer cybersecurity days actually from the internet.
But I think the first AI worm that can actually like, you know, reprogram itself that will build his own defenses against antivirals when you try to kill it that will be able to use humans, rent humans, and pay cryptocurrency to rent his own silver farms to live a nice life in some silver farm in Bolivia.
Yeah. You know, I think that my prediction is like 2026 will be like the first AI worm, which is like, you know,
just like any good human right? Any human wants to survive.
you know, no human wants to die.
so, you know, if you give these, [00:05:00] Clawdbots and Moltbooks and you're giving them access to cryptocurrency, wallets and everything, like yeah.
Adriel Yong: And I think to some extent it's already happening. Right? The ability for, you know, agents to go and fix something, unblock themselves when stuff goes wrong.
But I guess the scary part is what happens when they are now able to like, quote defenses against human beings to prevent them from being like stopped or like terminated.
I guess that's the truly like dystopian and scary part when the worm escapes the human being.
Jeremy Au: Well, the truth of the matter is that from a human perspective, our body already does the same thing. Right? Which is that we recruit our own germs and viruses to attack those viruses they don't like. Right? I mean, for example, like our gut has trillions of bacteria, more bacteria cells, you know, it's like a huge ecosystem in there.
And your body basically farms symbiotic because what they don't want is they don't want e coli and all these bad viruses and bacteria that attack. Right? Similar for that, I think once these things go live inside, replicating will probably end up [00:06:00] building more robots or bots to go hunt them down and you know, put them into it.
So it goes back to Blade Runner 2049. We're gonna have bad bots going broke and we're gonna have good bots.
That we just find ourselves giving extra processing power.
And say thank you very much for hunting them down.
But that's pretty much if you look at cybersecurity, that's what happening today, right?
Yeah. People are generating bad robots, viruses, to attack, countries or power stations.
And cybersecurity is like using code to defend against it, or either proactively or to even hunt them down. Right? In counter attack, China, you know, extricate all the scan of this, you know, Ned said they send them there by execution because, you know, the Chinese government is like, well, your reward for stating billions of dollars
for my citizens is death, right?
I think it's like that cat and mouse dynamic even so don't feel bad about it. You know, we're gonna get it.
Adriel Yong: We'll eventually figure out how to catch up with the machines
And proactively defend against them. But I guess that the scary thing is that
I think when it comes to security, it's [00:07:00] always on the back food and always playing catch up, right?
You see that with cybersecurity, real world security. Someone always has to make a quantum leap forward, which then threatens and scares everyone else before the rest start to play catch up and, you know, get 10 x better on the defenses.
Jeremy Au: Yeah.
Adriel Yong: So I think something really bad might actually happen first before.
Jeremy Au: Well the truth is you never know for this kind of stuff, right?
Adriel Yong: Yeah.
Jeremy Au: Actually, I mean, the one I think about is just like, since I'm in healthcare and biology is like, I mean
for example, you go to India or developing countries and get food poisoning. Yeah. And food poisoning actually causes deaths actually in quite large scale. So you can usually imagine people contaminating under example, right? Water sources or et cetera with that. Right?
So actually I would say the biological, I wouldn't worry so much about,
nuclear, even the scenario and even chemicals actually quite difficult 'cause you do lots of diseases. But I would say biological is actually probably like a much more worrisome thing because
there's actually a lot of agents out there. Yeah. And it's actually quite easy to cultivate from a home, actually.
So I think if you asked me I were to write a science fiction book, [00:08:00] it'll be like, probably like AI plus biological rather than, you know, I think most people think about Terminator, which is like AI and nuclear
Adriel Yong: Yeah.
Jeremy Au: And military robots. And then I'll create like the protagonist would be like a human detective.
Adriel Yong: Yep.
Jeremy Au: And a robot detective, you know, so like a good buddy, you know, noir
in a post biological wasteland.
Adriel Yong: Yeah.
Jeremy Au: But you know, so that'll be the life is. But let's talk about that article that you talk about called Something Big Is Happening. So it is written by Matt Shumer in February, 2026.
He talks a little bit about that opening. He was saying like, hey, in February, 2020 imagine you are a person who felt like you are a conspiracy theorist because you are watching this weird rumors of a virus, and he was stock hauling toilet paper and mask. And by the way, I was, this person.
Because you can look at my Amazon history,
my wife was like, why is there so much toilet paper and masks? And we pissed out of New York in March because, you know, it was a nightmare. But basically what he's giving analogy is that after a while everybody realizes is that there was a pandemic. And so, you know, people are giving these warning [00:09:00] signs.
And what I'm saying is that all the people in AI feel like conspiracy theorists in the sense that they feel like something is happening in February, 2020, but they feel like the whole world doesn't .
Adriel Yong: People don't understand yet.
Jeremy Au: Yeah, exactly. So, what do you ask to wanna add about this article since you said this to me and, you know, you, talking about this article.
Adriel Yong: Yeah, no, I mean, I think, the thing that struck me about how pervasive AI has gotten in the last couple months is when I go to Pulao Tekong and, you know, people reservists are starting to talk about AI agents in their personal and professional workforces, which to me then means like there's some sort of like escape velocity in like AI adoption across the population.
But you know, then again, I would preface that it's still very much limited to people in the workforce, just in the workforce, right? They're very tense level of like five to 10 years working experience. My sense is people beyond that are very entrenched in previous modalities of working software usage, et cetera.
And don't really appreciate the need to like, understand AI and how it can [00:10:00] fundamentally reshape their own value creation, their own value to the companies. Right? You imagine maybe that middle manager or like that PMA, I mean PMA, unemployment is steadily rising, that's one issue. But that's always been that level of like retrenchment when they hit like the 40 to 50-year-old age, and they're like neither strategic leadership nor like, you know, cheap enough to be like entry level graduate and, you know, high output updated the latest systems, et cetera.
And then on the other hand, you have kids.
Jeremy Au: Mm.
Adriel Yong: I think kids and students previously were a really very brain fried from like TikTok, social media very, very westernized, right? You know, promulgated by whatever they're consuming on social media. Now you layer on like chatGPT which is, you know, we talked about the westernization of
taught in a very, very subtle manner that creates another layer. But you also think about how the workforce outcomes for a lot of the students might be very, very glim. [00:11:00] Because today, you know, maybe I hire what couple fresh graduates you know, to be like that thin layer at the bottom, the goal of execution, high velocity work.
But as AI gets better and better, your agents can now effectively do a lot of the work that, you know, fresh graduates used to do. And you know, you layer that on top of like, you know, large tech companies, which used to like take in a lot of this young people. Not doing layoffs, being much tighter, being first to adopt AI
'cause tech companies have traditionally always been a lot more like front foot forward in trying on new things and ways of organizing. And I think that just, you know, makes the job market very scary for like fresh graduates in a way that I think will be very, very unprecedented. And I don't think that governments around the world fully appreciate the social economic impact of that yet.
And I don't think education systems around the world are fully equipped to prepare I think young people for much, much more like uncertain like workforce one where there's probably [00:12:00] not gonna be a very clear corporate, you know, career ladder. That's the path probably is gonna be much thinner for folks.
Future of work could probably just be like a portfolio of like projects all time, right? Portfolio of gig work all the time rather than full-time employment. And that also becomes scary 'cause like, you know, for people who used to rely on companies giving them insurance, what happens, right? All the different types of benefits and you know, how does that impact, like, people thinking about mortgage home ownership family planning.
So I think the ripple effects of a more like fragile, uncertain, floaty, like younger economic workforce is gonna be very interesting to watch.
Jeremy Au: Yeah. Yeah. I think, you know, I was watching my old professor, professor Joseph Fuller, and he's basically, does the is it your future of work at HBS, and I think his institute was actually quite sleepy until AI came.
Adriel Yong: Yeah.
Jeremy Au: Because the future book at the time was like, oh, okay you know, like, how do people become more creative, et cetera. And then obviously pandemic [00:13:00] happens a lot of work from home. And then obviously AI has come in and even he was in Harvard, Singapore kind of explaining his research and while sure
is that basically almost every company that's impacted by technology is becoming diamond shape.
Adriel Yong: Mm-hmm.
Jeremy Au: So historically every company was the effectively pyramid shape, right? Yeah. Which is, you have lots of junior people, some middle managers, and very few executives, right? Yeah. And what's happening is that some companies historically could be diamond shape, like Boston Consulting Group, et cetera.
If you feel like you could poach your junior people from other companies. So for example if you're Boston Consulting Group, you're diamond shape in the sense that you had lots of good managers and principals, but then you didn't need to have a very good associate base because you got poached those
From other consulting companies, Accenture, Deloitte,
and the PWC, and pull them into, or the industry, into your middle management. So you don't need to train for everybody. Right?
Adriel Yong: Yeah.
Jeremy Au: But what he's saying is that every company's effectively becoming diamond shape which just your program.
And this article talks about it for lawyers, for programmers, for marketers.
Adriel Yong: Mm.
Audio: You
Jeremy Au: know, you don't need a junior person to do transcriptions [00:14:00] or to do basic Photoshop, yeah et cetera. You need somebody who has a middle manager, who has some taste, has done it before, so they know what's...
Adriel Yong: Yeah.
Jeremy Au: The right taste.
Adriel Yong: Yeah.
Jeremy Au: And how to fix it, and then just do it themselves. Right? So a lot of the groundwork that junior people used to do as part of it. Yeah. Which I used to do. 'cause when I was a junior consultant, I was taking notes. My job was to sit there, shadow, take notes, and then type it up afterwards.
Adriel Yong: Yeah.
Jeremy Au: And then send it off to everybody.
Adriel Yong: Yeah.
Jeremy Au: After reviewed by my middle manager, that was okay. And now people are just you know, I recently got an interview by a Chinese you know, they basically had a voice recorder. Yeah. Right? Just be used, right?
Adriel Yong: Mm-hmm.
Jeremy Au: And so, that used to be a junior person typing like Uzi.
Adriel Yong: Yeah.
Jeremy Au: And now, you know, as a voice recorder can do it. Yeah. So it's quite interesting to
Adriel Yong: Yeah.
Jeremy Au: See.
Adriel Yong: Hundred percent. I mean, in the past maybe you had like two or three people with you in that interview room, but now probably you just need one person there. You know, you have a bunch of devices running the background.
You know, maybe it's like an auto AI which recorded, and by the time your interview ends, there's like [00:15:00] transcripts that's fillable in like English, Chinese or all the national languages. Right? And then maybe it has already gone to one round of like editing from an editing agent.
Jeremy Au: Yeah.
Adriel Yong: So it's almost like ready or close to ready to like instantaneously publishing.
So, you know, I think that's one very clear example of how workforces get thinner.
As technology advancements get more and more like pervasive in the mainstream.
The diamond shape analogy is interesting. My perspective actually is more that it becomes an inverse triangle.
And I say this for a few reasons. I think, you know, middle managers I think up until recently we didn't fully appreciate how fast or how robust AI can be in like intelligence to the sense where, you know, what you might have previously delegated to a middle manager to figure out. So for example, right? Let's say you're the CEO of a company.
Previously you might have like a, I don't know mid layer finance manager and you're like, go and figure out all the US compliance policies that, you know, we need to [00:16:00] adhere to because of a restructuring. And you ask that person, we're gonna do some research, blah, blah, blah. And I guess that's what a lot of middle managers do, right?
They're in charge of like semi strategic planning functions, and then maybe a bunch of like interns who are really junior analysts, like supporting them on all the entire range of on that, plus executing on it. But today, like I just ChatGPT, I'm like, and ChatGPT pulls out not just what I need to comply, ask me, okay, gimme more context in your company,
tell me what you already have. And it just goes in a loop and a loop and loop and gives you a very, like, comprehensive plan and then ship it off to like maybe a very junior person. So almost that middle layer probably doesn't really need to exist. But then even the junior person could be like automated all the time through like, you know, different agents that can actually help you execute, right?
Like for instance, like, you know, you might, that's part of the restructuring, like part of the, the operational plans. I need to email all my service providers and tell them, hey, there's restructuring going, on please change old details. That might have been a [00:17:00] junior person's job or someone in an emerging market.
But it's probably not gonna be a job as our agents become more like secure and accurate enough to like fire off that whole bunch of like emails. Yeah. Respond to the follow ups, finish more details and information.
Jeremy Au: Yeah.
Adriel Yong: But I do think the person at the top, IE, the COOU, right? Will probably still have a job, but you'll probably look more and more like a super icy rather than like, you know, a manager of a manager which then manages a team.
I think that's interesting because then when you think about how that impacts like organizational structuring how do you think about comp, right? Like if everyone's gonna be super icy, effectively, maybe you should be way more generous with everyone's like equity and cash comp. 'Cause now everyone's effectively like taking on a much larger scope of responsibility and generating way more outcomes and outputs in the same unit of time.
Jeremy Au: So, you know, first of all, you're basically saying it's not even a inverted triangle. It's maybe like a square.
Adriel Yong: Yeah.
Jeremy Au: A very small square.
Adriel Yong: [00:18:00] I mean the good example is like Casa right? 30% team that got to one a billion, you know?
Jeremy Au: Oh, it's like a little house, you knows, like a little small triangle for the CEO, like top and then a little square.
Yeah. And then it's like, okay, you know? Yeah. It's like, there's like this whole pyramid and he is like, you cut off the bottom and then the middle layer just have a small square.
Adriel Yong: Yeah.
Jeremy Au: And then the rest is this. And of your C-suite team, right?
Adriel Yong: I mean the other way of I think about it's revenue for employee just gets larger and larger, right?
And that also means less employees but more revenue.
Jeremy Au: I think for me, my philosophy actually, and I was just talking to an executive, 'cause he was talking about how he lost recently. He has like COO and they're really sad about it, et cetera. And he felt like he was giving a lot of free time, et cetera.
And I said. Actually, my perspective is with this AI generation, et cetera, I think the top of the call is not gonna go away because there's something that's gonna happen because you still need in person. Yeah. I think for a lot of the design.
Adriel Yong: Yeah.
Jeremy Au: Secondly, there's actually still a lot of secrets, like personal knowledge that, you know, you're only gonna get by talking to this person at 7:00 PM,
after dinner or like you know, going for a [00:19:00] run with them, which robot can't do.
Adriel Yong: Yeah.
Jeremy Au: Those teams are not gonna go away anytime soon. Maybe sales support will go away because you're just doing data organization and comp calculations and stuff like that, but sales not gonna go away.
Yeah. So these front facing roles, and obviously your engineer, your super engineers who are running the whole architecture.
Adriel Yong: Yeah. Are
Jeremy Au: still gonna be around, obviously.
So now you have these, this permutation that's happening. So all I was trying to say is like, even though we were joking about the little square and the little hat, you know?
Yeah. But I do think that it's gonna be an arm of it that still exists is your front facing stuff, right? Yeah. Which is sales. Now the arm would be the engineering leaders.
Adriel Yong: If it makes
Jeremy Au: sense. And, but yeah everything else is kind of like, yeah. Goodbye marketing, goodbye PR, goodbye ops.
But I think the physicality wall is still gonna be there, right?
But your customer service hotline is gonna go away. Okay. So putting that aside, I think yeah, I agree with you. I think it's a very interesting problem for folks and I guess the question people ask themselves is, okay, so what should I do? I guess? Yeah. What do you [00:20:00] think? I have some thoughts as well.
Adriel Yong: Yeah. I mean, depends on who you are, right? If you're a young person, I think the pressure to just start doing things from young to pick up work experience has gotten so much higher. I think in the past, you know, college kids could get by, you know, college, get a good job even if they didn't do any internships.
And then you see the whole internships, arms race happen in college. Everyone's like trying to get their first few internships, whatever. And then that obviously stacks up to, you know, social capital plus like some base level of skills, which makes you, I guess, more employable. 'Cause you know, your employer doesn't think that I have to spend the first three, four months, like, training this guy.
And he's really like shown that he's a pretty good, like, junior level worker. I guess the scary part, and I think we've really seen this challenge for a country like Singapore where, you know, low cost labor is very available in neighboring countries. You really have a fair bit of like, offshoring of like junior level jobs, right?
And then now [00:21:00] it's like flood exacerbated because your junior level workers are now equipped with AI in emerging markets. So three, four x more efficient outputs and outcomes wise. For a fraction of the cost of a Singaporean. How do you solve that? I don't know. I think Singaporean is probably just need to actually think about being more entrepreneurial and like learning how to manage teams, create value, create opportunities instead of thinking about how do I find a job?
Because I think there's gonna be real less jobs in the future with all these challenges.
Jeremy Au: Yeah, I think I have some similarities and I have some differences. I'll say the part I'm similar about you about is that I think Gen Z is really getting hammered by this, I think generation Alpha, like my kids.
They still fall and six, they, but they have time to enjoy life.
And so I think a lot of this reconfiguration will have happened by the time they hit. Yeah. So I think for parents of those young kids, they only need to worry as much because it's kinda like in the tsunami, right? If it's really bad, if you're very high above the air, it doesn't matter.
If you're below the water, it doesn't matter. But if you're on the surface [00:22:00] or at a coast getting hammered, right? Which I think the Gen Z is really that generation is gonna get hammered. I think I agree with you that I think you gotta be agile and nimble and you have to figure out this reconfiguration is happening.
I think obviously don't do the shit jobs or majors and expect to have a job.
Adriel Yong: So what's a shit job or shit major? Because I remember when I was entering university in 2019, yeah. Computer science engineering was like the hottest ship, right?
It was like the highest paid degree. Once you graduate from like NUS or wherever everyone wanted to be like a CS major.
Now I think like unless you're really like top caliber in your class for CS, getting a job is so much harder.
Jeremy Au: Wait. So I think that's the part where I did disagree with you on your prior statement, this statement, and but this interesting you said is I think CS is still a great major to do, but I think what's been missing is the mindset share from getting a job to being entrepreneurial.
Because if you're a CS major, you already have a lick up on everybody, which is you understand AI, [00:23:00] you can assemble your own little bot army.
And your job is to eat as much share from other people who don't know how to resemble these mini bot fleets that support you in your daily work. Right? And so I think the issue about, I wouldn't say that computer science major is bad,
but
I do believe that the mindset that your job is to therefore get a job in it is bad.
Right? And so I think if you add the fact that, say you're a computer science engineer and you add that those of create your own job.
Or create enough value where you'll create your job.
Then that's actually a viable path.
Our friend Adrian, who runs the Career Show podcast, he'll say saying like, there's a degree on IT systems managed by it. And I was like, yeah, that sounds like a really bad job if your job is to manage IT systems. Yeah. Like that's the worst of all jobs because
united know enough to do IT as in the coding or the bots and you just manage systems and systems wanna manage themselves.
You know, I mean, how many Google Suite administrators do you need?
So I just think that IT system management is a really bad degree, for example.
I
[00:24:00] think some like marketing, your graphic design jobs also are starting...
Adriel Yong: yeah.
Jeremy Au: Animated jobs are really starting get evaporated as well.
Because it's just AI can animate. Now, don't get me wrong, I think do animators who can say, okay, I wanna create my own plot, create my own interesting movie
and monetize it. They're doing well.
Because now it's much easier for them to do it. Yeah. ' Cause a non animator trying to create animated movie can't really do it because they don't have the aesthetic sense of it.
Yeah. Other than saying like, studio Ghibli.
But you know, so I think this is something interesting for us to think about.
Adriel Yong: I think you should just step by. You can also ask yourself the question, like, what is the point of going to university to learn all of these things when information now is like a hundred x more available and it might even be better than your professor or teaching assistant trying to like, run through a bunch of slides and assign you some homework.
I mean, the computer science example is super clear, right? Like your professors in the universities and your teaching assistants the most forward facing and up to date, what's happening in the development of latest models, et cetera. Questionable and [00:25:00] debatable.
Jeremy Au: I think that from my perspective as a lecturer at Singapore Management University, right?
I teach in there.
Adriel Yong: Not in NUS anymore?
Jeremy Au: Yes.
I used to but unfortunately I've decided to let it go to somebody else. Fortunately the work has taken me, I gotta do my work. But yes, I will miss teaching the NUS undergraduates. But I think what's interesting is that
you are right to say that knowledge should be just like taught much faster. And so even for myself, the curriculum is really changing to be much more practical group, project based. Yeah. And participation based. Right? So what I mean by that is instead of saying what is a term sheet? You should be learning how to negotiate a term sheet.
Right? If that makes sense. Instead of saying, what is a startup you should be looking at you know, Singapore startups today.
Adriel Yong: Yeah.
Jeremy Au: And 40 of them and saying, which one would you invest in doing your own desktop research? So, you know, applying that stuff.
Is still a feature because you know, but you're right to say that textbook knowledge is just gonna be downloaded shit away.
But it's more about the practice because humans still like practicing in real life. Right? Yeah. You know, reading about baseball can be used to be a textbook. [00:26:00] Now it's through your Gemini app, but nothing's gonna replace you actually being in the court to actually
play baseball.
Adriel Yong: Yeah. Or
Jeremy Au: pickleball or whatever, like, you know, there's a difference between the gap between theory and practice. And I think that's, I will say I feel positive about universities because universities are, first of all, state subsidized and sponsored.
Their function of the state.
So, you know, as long as it's a function of state and the state has the rule of law, the rule of military force. The rule of taxation.
And the rule of social norming. Universities still exist. Yeah. But the universities will have to flex because I totally agree that for many people, vocational is maybe faster.
Like, I agree with you by the way it's like if you are willing to learn from Gemini or whatever it is, and you're willing to practice yourself
then university is kind of like a
big waste of time actually. So
but I just feel like most people in the world
are not self-starters. I think it's like a bell curve. Like 30% population are like self-starters will do it themselves.
And at 80, you know bottom 20% would be like, I need a lot of help.
And then the middle 60% are like whatever my [00:27:00] peers are into,
like watching TikTok or go to K-Pop, you know, whatever it is, they're just gonna like default
there, which actually is a good segway into the last thing we wanna talk about, which is the Singapore budget.
Because Singapore budget apparently has something to do.
Adriel Yong: I know you love your vouchers, but I think that's the wrong
Audio: one.
Jeremy Au: It's the wrong one.
Adriel Yong: As you know. It's the right one for most people.
Jeremy Au: Let's be real. Okay.
Adriel Yong: Thumbs
Jeremy Au: up. This is not me. This is the Singapore government budget and poster number one is all the vouchers they're giving. So
Adriel Yong: you save what? Thousand, couple thousand? No, that's at least like five, $10,000.
Jeremy Au: Yeah, I know. CDC vouchers, et cetera. So I think I wanna say I like this and it's not me, they put this as the first
image. So it's not me.
Adriel Yong: I actually haven't finished spending my CDC vouchers. I was like, wow, I still have like $200 here if it feels like that's a number.
Jeremy Au: Don't say that. Say that you ran out. Yeah. And because you are busy reskilling
Adriel Yong: Exactly.
Jeremy Au: The fact that you haven't used it means that, you know, they're gonna be like, well, the bot has listened to this transcript of the great podcast and realized that Adriel has too much money and doesn't use up his [00:28:00] vouchers.
But I think these are important because these are direct cash transfers. And from an economist perspective, it's always good because
direct cash transfers are very efficient.
There's not much administration blow that happens with it.
And they actually have a distributing effect because the same, you know, $1,000 to somebody who is poor versus a thousand dollars for person's rich because university, given its everybody's supportive of it.
They
don't feel like cutting it.
But obviously a thousand dollars means a lot to somebody who is trying to survive day to day.
Right? Yeah. So for them, I think cash vouchers is actually a very efficient and very modern behavioral economist way.
And
Adriel Yong: actually this is an interesting form of cash voucher. Yes. If you think about it, I remember back in the day, it was just like a direct bank transfer.
And the government will have no idea what I'm spending it on, like alcohol or like,
you
know, fast food, whatever. But now even your hot land merchant has to scan the QR code. So I'm sure there's a massive amassment of data on the backend about what people are actually spending on, which dolls [00:29:00] are really popular, doing well in which locations.
So it is actually I think a really good like government tech feed actually, if you look at the entire system end to end.
Jeremy Au: I think one other thing to add is that historically, if this get out cash, they could have bought it, for example, for online, for digital subscription, the money will leaked outside of the country.
Yeah. But because these vouchers are tied to Singapore Hawker centers, to local heartland malls and obviously local stores, actually the money definitely flows into the local economy.
And then that also has a trickle on effect for their workers and taxpayers and, you know,
corporate taxes and CPF.
This is actually, I think a well engineered stimulus actually for the economy that people go under underestimate because they see the voucher side. But I think it's actually relatively well designed compared to some other approaches. So, but you know, we'll do a further breakdown of personal finance on that.
And then what's interesting is that they're doing a corporate tax rebate. They're helping SMEs, small media enterprises to expand originally. And as [00:30:00] well as to put some money into the anchor fund and the equity market development program
to help medium sized companies, startups.
What are some thoughts that you have?
Adriel Yong: You know, I think the announcements on strengthening the enterprise ecosystem was part of the book that came out of the Economic Strategies Review committee. And then I think that was part of the entrepreneurship committee's like recommendations.
Right? I think one part is what is the reality today? The reality today is that your exchange is probably, you know, one of the worst performing like exchanges in the region behind the Bangkok Stock Exchange, behind KL, IDX, you know, very few regional sort of like startups that have gone done well and gone large across the region.
Want to list in the SGX.
Most preferable analyst on the Nasdaq, New York Stock Exchange or elsewhere. But you know, when you think about that, how does that sort of like impact the long-term anchoring of this homegrown or regional grown like companies? I think that's the reaction, right? With [00:31:00] the anchor fund.
Well, but the problem here is that fundamentally Singapore's stock exchange is nowhere as liquid because, well, if you think about the number of people in Singapore to trade on the SGX, actually it's way less than
America. And then you compound that further with, you know, Singapore being much more for high savings rate type of economy population.
And you know, the truth is actually for most people on the back of their minds, they're already like, the government is really helping me invest my money through CPF, right?
Why do I, you know, take that extra risk to try and like play the stock market, especially when, you know the historical performance, even for your largest companies like your DBS Bank or your SIE has actually not even been anywhere like Stellar compared to the equivalent in the US right?
Like, why do I invest in DBS Bank on the SJ American when I could invest in your Morgan Stanley, or whatever new American banks, that on the New York Stock Exchange or Nasdaq. So that's, I think one part of it. I think the [00:32:00] other part of it's like
how
large is the pipeline of companies that can realistically IPO and SGX? And I think actually not that many because if you look at companies as a funnel the truth is just also a function of size.
You do not really get that many companies progressing from like SMEs where maybe they're doing like couple million in revenue to like tens of millions in revenue. Having a massive footprint across like the region or even the world, right? I think it's really only in recent years where because of venture capital now more companies are expected to like, expand across the region or like the world, especially now much, much more like compressed timeline.
So, I think on one level, yes, you need to inject capital at the stock exchange level, but you also need to widen the funnel, right? So like I think in my opinion, we still cannot do enough on early stage financing. Part of it is what we talked about earlier with so much excitement about things going on in AI.
I [00:33:00] think a lot of family offices local and regional have sort of redeployed or reallocated capital to the US
like
private markets, public markets. I mean, part of it also is just, you know, venture returns in the last 10 years, which I think we have talked about and other guests have talked about, just has not returned capital to them, recycle back into the ecosystem.
So, you know, because of all of that, you don't really see that top of funnel. I think the top funnel is like getting speeds. I think Deal Street Asia recently, was it Deal Street Asia or one of the tech media outlets talked about how there was actually a strong recovery in like growth stage financing, but your early stage financing has weakened much more.
So you see almost like an inverted funnel now. So we definitely pay more attention to that. I think like, you know, if you look at what Singapore did like 10 years ago to kickstart the whole ecosystem, we had the early stage venture fund program, ESVF, which was administered by the National Research Foundation.
Right? That seeded a whole bunch of [00:34:00] venture funds like Wavemaker, I think Monk's Hill I think Golden Gate as well, you know, all have become reputable large names that have gone to raise like way more institutional, like, you know, US Capital manage hundreds of millions. The problem with that is that they might have previously started out deploying call it 500 k, one mil checks maybe much earlier stage, but then now when you're managing like couple hundred million dollars, you wanna deploy like millions at one goal and not like all the small checks.
So you then have that, I guess much less capital accessibility. On the angel investor side, I think, you know, we have had angel investor networks for the longest time, like XE Network, Angel Central, Ascend Angels. But how many of them are still active today?
And you know, obviously we know Ascend Angels fairly well, right?
You know, their activities is like much lower on early stage pre-IPO like booming. Right? And then, I mean, I think Der Shing and Shaoning Angel Central have continuously talk, have written quite publicly about how, you know, [00:35:00] Angel Centrals do activities also not as high as before. Because so much capital is locked up for angels and so much capital is not returned or have been written off.
So
I think that's some level of feeling burned.
so yeah, I think that's, you know, some of the challenges on the early stage side.
So because of that, I think we should have, yes we have 2.0.
And actually seed a new generation of emerging fund managers. I think the Middle East is doing that so aggressively, they're pouring like tens of millions per fund manager to get 'em set up, shop launch funds here, invest in the local ecosystem.
Today, enterprise Singapore they do a lot of co-investments with like their fund partners, right? So I think this will be your established funds, like your Monk's Hill, your Peak XV. But I think, you know, you need to go further upstream and like, you know, do co-investments with angel investors.
So give them some economics. You know, we bring so many talented like tech executives and, you know, tech entrepreneurs to Singapore through like EDB and the visas, right? [00:36:00] But. We don't really like leverage on them to like support, you know, the local ecosystem. You know, give them like scholar checks to deploy. Right?
You know, and like get them involved in a local ecosystem, sort of like encourage them to start funds here and co-invest into the fund starter right on. I think that's one way we can get the very early stage part of the ecosystem off the ground. And at the end of the day, like ecosystems are like decades in making, if we do not aggressively do it now, I feel like there might be a lost decade for startups in Singapore.
Jeremy Au: So, yeah. So I think there's a lot to unpack there. I think we were talking about late stage public markets, also talking about early stage and I also talking about what the Singapore government can do to change. So those are three very large topics.
First of all, let's just talk about late stage and growth stage.
I agree with you there's a large cohort of companies in Singapore that are large, relatively large, private and still has some time before they can go public. I mean, I'm thinking about, just taking my head, Carro, Carsome you know, there's funding societies, igloo home. I think there's a [00:37:00] whole generation of companies that have sizable revenue, sizable employee count and they're still pushing hard and obviously the companies have ascended will be your Grab.
And obviously, you know, you had your Gojek and your with Tokopedia. So I think there's been a gap for many years now,
for the past effective, I would say five years where we haven't seen really any IPO activity. So I agree there's a chasm there.
And I think that, where I'm differ from you a little bit is that I think that historically I think SGX is underperformed, but I do believe that with the equity market development program, we know, which is about effectively it was like 5 billion for government, plus another $5 billion from other funds.
So it's about 10 billion stimulus and now they're adding another 1.5, which probably plus another 1.5 makes it another 3 billion. But I think this is gonna be a stimulus effort to help companies get and cross that gap. And I think it's gonna be a big stimulus because just for numbers off the top of my head is that Singapore stock market, the daily trading volume is $1 billion.
And then in [00:38:00] Hong Kong it's about 37 billion trading volume every day. Right? So one is, yes, Singapore is much smaller compared to Hong Kong.
Which is adding both the China growth story of the companies there, but also I think a huge number of mainland Chinese who want to get exposure outside of mainland China.
So I think Hong Kong exchange is getting stimulated on those bi factors.
But I do think that there's a space for a Singapore stock exchange to become the regional hub for all of the unicorns across Southeast Asia, because you mentioned IDX, the Indonesia stock exchange and
I would defer from you. I would say that Indonesia stock exchange has now underperformed Singapore stock
Adriel Yong: exchange. Yeah, that's fair.
Jeremy Au: But I think because of the afloat is even lower, you know, and the transparency around data performance is worse than Singapore.
And Vietnam also is not performed. Thailand is also not performed. So there's actually an interesting opportunity that if the government really pushes hard and creates the right analyst coverage, right piece.
I think they could suck up the unicorn IPOs.
Candidates from Southeast Asia. And I'm talking about whole Asia. I'm not talking talk about Japan, I'm not [00:39:00] talking about Korea, but it's talking about Southeast Asia.
Because I think if you're an Indonesia startup, you'll be like, I'm not really go to the US for sure.
I'm not gonna go to Japan because unfortunately, Japan is primarily Japanese speaking.
And so the endless coverage doesn't translate any English proficiency.
So you're down to Hong Kong or you're down to Singapore. Right? So I think little less bearish. So I think that's one piece on that I think end the government recognizes it on the late stage.
Yeah. Talking about the early stage I agree with you that I think a lot of the early stage funds, seed funds, and series A funds have become zombie VC funds. I think we're aware that, you know, they have the name, they're prestige, but they're not actively deploying or
Deploying aggressively.
We won't name names because that would make us unpopular, but you know, we know they exist.
And founders know that as well.
So I do agree with you letting the government has to step in, but I think obviously government's also kind of like, oh no like we did it once EVSF 1.0 and now we have this, you know, cohort.
Yeah. I'm also the kind of infinite money from their perspective to do it all.
So
I actually wanna double down what you said. Actually, even though as I said all that, [00:40:00] I would say that first of all, having a strong startup funnel from the end to end is important because I think that productivity enhancements through startups is gonna ripple out more than trying to push this through existing SMEs.
And what I mean by that is we already see, for example, a lot of startups, they're very aggressively helping SMEs with the productivity enhancements down.
They're
not necessarily vendors or contractors or approved by their enterprise development ground EDG. Yeah. But actually all these startups are already serving local clients
with AI agents illustration, obviously
a good example would be Jingjin . Right? You know, at Superbench.
She's busy servicing SMEs. And I think she's doing a better job educating SME owners than many people in the ecosystem, right? Yeah. And she's a startup and so supporting startups like her to scale will not only help her hire people and jobs for these new future proof jobs.
But actually there are two major problems with that program as well, actually. Right? One is that you need to write [00:41:00] for written approval by enterprise Singapore.
And secondly is that you need to qualify only for checks of a hundred thousand dollars or more.
And then after it has been held for two years, you need to send a letter to the agency to get it confirmed that you held it for two years without selling it. And then you can apply for tax deduction. And I'm like, whoa, whoa, whoa. Like,
Adriel Yong: yeah.
Jeremy Au: Let's take a step back here with this. Okay. What is this scheme supposed to do?
I think this policy is scared of people going rogue, right? Yeah. Okay. So totally understand is a controlled mechanic, but what is the reality for angel investors today?
Adriel Yong: Mm-hmm.
Jeremy Au: First reality is that angel investors, number one, need to diversify. So they may write small checks of like 10,000 or $20,000 and they're supposed to write maybe 30 checks.
And they're still $600,000, but they're supposed to do this widely, widely to diversify. So a hundred thousand dollars times 20 is effectively you know, 2 million.
I mean, I'm just saying like most of your tech executives cannot afford to do $2 million.
I think that's number one. Number two [00:42:00] is.
You have to get approval, written approval. You had apply and wait for, I mean,
who's doing it? I think you should just be accredited. If you're an accredited investor, which the government already knows and has on file for you, then you should just be, Hey, you pass because you accredited.
Yeah. We expect you to be able to do investments in a simple stock exchange. Yeah. Or anywhere. So you're free to, so accreditation will simplify that second piece and thirdly is that we all know the angels have to work in syndicates or groups.
Because there's not enough knowledge at individual level to write it.
So the reason why they certify individually saying like, you need to be smart enough to write a hundred thousand check. But the truth is, angels work in groups in 20 or 30 groups to write 20,000 checks to hit half a million dollars. Right? And so syndicate leaders are the ones who are smart enough, but they're not getting supported.
So tax benefits only apply to individual checks a hundred thousand dollars. So you actually destroy the value proposition for the syndicate leader.
'Cause basically what happens is an angel, you're like, okay, I can write a 20,000 check to diversify across [00:43:00] 20 and I can listen to somebody smart like Roger, help me write the memo.
Or I can do a hundred thousand check by myself and I get tax deduction.
I mean it's a distortion.
of the angel investment activity, right? Yeah. Which is if you are a founder, you don't have time to look for a hundred thousand dollars people and check people who don't have time for you.
The a hundred thousand dollars in check writers are actually part of syndicates.
Anyway, I'm just saying like this whole scheme is actually
quite ready for a 2.0 upgrade, I would
Adriel Yong: say. Yeah. I mean actually an interesting part also there is that founders also want minimum check sizes 'cause they'd want like 25K checks on the cap table.
Crazy messy. So obviously a lot of syndicate leaders do like SPVs.
But there is actually a policy and regulatory gap around like SPVs and like syndication of like early stage angel checks. I guess it's not really a big issue for, I mean, as to like look into and all that, but I think it's necessary, right?
If you wanna seriously grow the angel investing like industry or vertical sector, but like have clear policies [00:44:00] that are not ridiculously a burden on you paying your investor networks.
And it also gives confidence and trust right. To individual Indian investors about, okay, I can go with this syndicate network to like, you know, deploy checks, especially if you're not from startups or tech and you just wanna like, you know, do some, you know activity there.
And then the other part is just like, you know, I would say grants and subsidies to make SPVs cheaper. Yeah. If you totally want more syndication happening you know, to support the least stage financing. Right? So yeah,
Jeremy Au: I mean the reality is okay, you know, if you are a policy maker, you listed to this, I guess is, you know, which countries around the world disqualifies, syndicates, or SPVs from these tax benefits?
And two is what percentage in America of angel investing activity, what percentage of them are individual investors in their personal name writing a hundred thousand dollars check?
Versus the total quantum driven by syndicates, because lots of people putting money into syndicates, even if they're writing a hundred thousand dollars check.
Yep. And so it's just kind of a weird distortion of the whole process because like you say, it is, and imagine if you are a founder and you're setting up a [00:45:00] rollup vehicle, which is a SPV to collect all these individual angels.
Yeah. You have 20 people with $20,000 checks, that's still $400,000, which is a good teddy sum,
now none of them get tax benefits.
Versus if you say, hey, you do a rollout vehicle by the company, you still get collect
Tax benefits.
Then everybody's gonna enjoy
That, and you're gonna make fundraising for founders so much easier. So if you don't want to subsidize the accredited investors, maybe you can do it through the company rollout vehicle approach.
Just so that you can get these small checks going. Right? Because then the government doesn't do anything. You are doing a tax deduction.
But you are causing that money to flow from people's savings in a US
you know, stock account. To flow into effectively domestic investment capital, right?
It is gonna create jobs. It's gonna create,
a good snowball effect. Right?
Adriel Yong: I mean, I think the other issue here is that, you know, unlike the US which has a massive high, you know, income, personal income tax rate, Singapore's income tax risk, like great low tax rates here, which becomes a challenge when it comes to more like risky [00:46:00] financial choices or like generous financial choices, right?
So philanthropy, for instance in Singapore is you don't see this the, the sort of numbers you see in the US because the tax benefits net of everything is just not the substantial that moves the needle. So I think even if you think about it from an angel investing perspective, it's like the tax benefits would only be queued.
Yeah. But it wouldn't be as strong as a motivator or as key of a consideration for I think people in the US. So I think that's one thing to also contemplate, right? How do you actually incentivize that, you know, early stage financing behavior more.
Jeremy Au: Yeah, well, I mean, I totally agree with you. And I think there's also zero capital gains tax in Singapore, so that's something that's reduces the distortion in America, right?
Yeah. Which pushes people into, you know, investing in capital gains for the upper middle class and upper class. And last thing we wanna quickly talk about is obviously, I think the big push on harnessing Ai. We talked about how Singapore has set up a national AI council chair by the Prime Minister champions of AI Productivity Solutions [00:47:00] grant to include AI solutions.
And they give six months of free access to premium versions of AI and taking instead of the classes. What do you think about this?
Adriel Yong: I would just pick on the last part, which is providing six months of free access to premium versions of AI tools. I personally feel like it should be way more aggressive and you should just do it like national overall.
You know, I think at a grassroots level there were, was it like, I can't remember some digital champions where, you know, you help your elderly with like basic tech skills and all that. I think like you should be aggressive and do like grassroots promotion and education of,
Jeremy Au: no, that's interesting. So you're saying like maybe like lifelong free AI for NTUC.
Adriel Yong: oh yeah, I think that would be great
Jeremy Au: people association
Adriel Yong: there. That would be a great way to make the NTUC membership relevant. 'Cause they always struggle with membership.
Jeremy Au: Yeah, but I'm just saying like, you technically you have unions, right? You have technically your job is to reskill and skills future.
You take the membership and you just say everybody has and I'm sure you can negotiate a really good rate as [00:48:00] well.
Because I think one thing, I mean, I see six months free excess is I think if you can guarantee like, that you're saying 'cause okay, what is happening to this bell curve of society, right?
Like Yeah. Your top 20% are already early adopters.
already using it through the enterprises then it's like crossing the chasm right now. Right. We're just adding the middle 60%. We're crossing the chasm now where our mom and dad are starting to use it to. Yeah. You know, I think the one I saw recently was, you know, like my mother-in-law using Google to check for which fish have a lot of mer mercury.
Right? And Google Gemini is pushing out that result.
Into, so suddenly it's crossing the chasm. But of course, from a gum perspective, we're always thinking about the people who are left behind. Right? Yeah. Bottom 20% who don't have access to it. And we already have lots of people who are already actually sheltered there.
Right? I mean the under NTUC, the National Trade Union Congress, obviously people's association I
Adriel Yong: mean,
Jeremy Au: you
Adriel Yong: know, instead of you have like a $500 skills future credit, you should just channel it to like, you know, free AI subscription just because. If you think about daily usage [00:49:00] of Oh,
Jeremy Au: yeah.
Adriel Yong: Ai, that is a massive, like, lifelong learning, such like skills future, like Yeah.
There's
Jeremy Au: much more bite-size. Right? Because instead of somebody who is like, you know, kind of like elderly and everything
they have to choose a course that interested in Yeah. Figure out how to sign up, figure out how to get it reimbursed, and then kind of like get it all done.
Right? Yeah. And then they had turn up and everything. ~But you can imagine almost imagining a solution where you're like, ~
~you know, AI, like yeah. There could be something really interesting there. ~
Of course there'd be some questions about is a white label version, are there parameters, around what the AI teaches you, et cetera.
But I mean, the crux of it is that if you give, I think, everybody under NTUC, like Yeah. You know, they say, you know, Gemini Pro.
AI Ultra.
Or you know, ChatGPT Pro. Yeah.
Claude, Like, I mean, at least they're not
It's cost. Yeah. It's not the app.
then they're after the resource right?
Yeah. I think they're gonna learn so much more like, say every day. 'Cause they then it's gonna be more relevant. They're gonna be like, hey I'm at a coffee shop, is my toasts
And then, you know, AI's gonna be like, if you eat your protein first and then a cup second, it will lower your glycemic index.
Which will be better for your health. And then you're like,
mean that's what happened to me literally. Right? I was [00:50:00] like, oh, eating protein before carbs.
I did not know that. Right? Yeah. And so I think there's a lot of learning. I agree with you. I like that idea. Which is,
channels on skills, future credit into just direct access for Yeah.
AI usage, right?
Adriel Yong: And then the other thing I'll comment on is, and this probably more an expression of hope, you know, when they announced the National AI Council, it was six, seven cabinet ministers. I'm just like, this isn't this just like, you know, mini cabinet. I would really love to see like founders of something large AI companies like your openAI or Anthropic your, Cursor, Windsurf, the whole gamut, right?
On some part of this council so that, that's actually like industry credibility, but also gives them like familiarity and interest in Singapore and contributing to the economy here. And the other thing is like if you look at what's happening in SF, so much of AI activity and companies are driven by like 20 year olds, you know?
Yeah. For even teenagers actually are building like multi-billion dollar AI companies like the history with a lot of. Singapore committees and councils is, tends [00:51:00] to be comprised of, you know, your executives, your leaders of like established GLCs or STA bots. Let's, I think maybe try to have some like diversity and reflection of what, you know, the AI, you know, sector looks like globally.
Jeremy Au: I really like what you said is, I think if you, I mean it's kind of like the pandemic where, you know, I think the first two countries in the world to have vaccine was Singapore and Israel in a kind of early adopters piece, right?
And I imagine if the National AI Council basically said, hey you know, to this I saying like, the founders of Anthropic ~or, you know, yeah.~
you know, ChatGPT and, and just say, hey, maybe not. I mean, the founder's obviously too busy.
But you know, somebody on your C-suite just comes by once a quarter to talk about what they're seeing.
And just let that talk directly. Yeah. No cameras, no reporters, no minutes. Yeah. Just let these, you know, co-founders or C-Suite people just say
hey, look at my phone.
You know, I was, yeah.
Adriel Yong: And also maybe tie a relocation benefit to it. Like, Hey, hey, hey, visa in Singapore. Yeah.
Jeremy Au: Yeah. But, just say, just let them have that private conversation and be like, hey, you know, [00:52:00] like, hey Prime Minister, you say you want a working paper on Ai, let me just hit the button now as this, an example.
Right? And then have, you know, Gemini X, X, Y, Z, 1, 2, 3, 7 version pro, like generate a, you know, a hundred page white paper.
Perfectly correct. Perfectly numbers. And say, okay,
you
know, make it a race. Like, you know, like, you know, a hundred civil servants on one side, you know, Gemini, you know, multimillion pro and then, you know, but you know, it's like seeing is believing, right?
Yeah. Does it does make sense. Like, you know, I was in SF and I was using it way more aggressively and I was like, it's like, you know, you can read about way more, you can see videos of it, but when you walk into the way more machine and then you put your, you know, lean back, your fan is, is exactly the word.
Yeah. Playing your classical orchestra music in your, you know, and then you just, there totally safe driving. And then you put your Costco rotisserie chicken on the driver's seat, you know, because if extra space and then, you know, just driving by and then obviously, you know, on the other side, outside the window, you're seeing this homelessness [00:53:00] crisis happening in slow motion
In San Francisco.
And you're like, oh no, you know, there's a future shock. Right? Like, you know, the job of the driver, they disappeared. Yeah. And there are people who are homeless and don't have jobs.
Right. Outside. Right. And, and then once you're in that dystopia bubble, I mean, there's a physical version of it, but the virtual reality of that version is like marketers.
You know, coders, civil servants. I mean, lots of jobs will have the same dynamic where you're like, you know. You're sitting in a way more passenger seat chilling.
There's no more driver left.
On your left there's a Uber car, a normal driver on your right. There's a Lyft driver, a normal driver.
And outside you see people who have no jobs. Yeah. And you're like, well, once waymo scales up.
That Uber driver's gone, that Lyft driver is gone.
And
then how do you avoid them joining the homeless population Yeah. Versus what's the new jobs they create. Right? And I think that's where I agree with you.
I think the National AI Council has needs more that
Adriel Yong: Yeah. Well that's gonna be a real big political problem because you think about how many Singaporeans today are grab drivers.
food delivery riders massive.
So many p [00:54:00] mets and I would say even like young graduates have given up stable, predictable careers.
Right. You know, become a grab driver. Yeah. More freedom.
Jeremy Au: Yeah. So and that job is gonna go away. Yeah. It's very tricky. Right? ~So on that gloomy, dystopian or I will say, my last words, I say a few more words. I do this~ I still think that this AI, this budget is still a step in the right direction.
Right? Yeah. I think you know, I always say government is kind of like lowest common denominator.
Like, you know for all society, right? Yeah. But I think for us who are on the, you know, sci-fi age, like actually automating our jobs.
you know, it does feel like, wouldn't say Singapore as a society can move faster, right?
Yeah. Everybody has move faster. I don't think it's about the government. The government has a big role in it, but the whole Singapore has more faster.
because I think for the countries that don't adjust well,
I mean it's like, is it, is the joke about the lion right? You know, it is like, there's like two people
Running and they see a lion and they start running.
And then while them. Starts tying shoelace.
And then another person's like, what are you doing? Tying shoelace being both gotta run and then the guy tying shoelace actually I run faster than you.
Right. You know, because only one person will get eaten by Yeah.
A [00:55:00] lion. Right?
And to some extent there's a bit of like this, very, you call it version of this.
Squid Games, I guess for jobs, you know, where like in this new reconfigured world, AI will automate jobs globally.
And there'll still be jobs left, obviously for people in the new world.
So which countries and which groups can be the one has their sneakers
Are tied up, you know, nice and sharp and can run really fast. Right?
Versus the countries or groups that wanna be left behind, I think. Yeah. And I think we just have to be thoughtful about that.
Adriel Yong: Yeah. I mean, I, I do think that, you know, the net
positive of AI on human society development productivity is just massive and probably very, very unimaginable.
But you know, I think the societal impact is something we should think about really hard because today you really see, you know, the fracturing of societies with globalization.
You're gonna see a 1,000 x you know, larger fracturing with economic displacement. Yeah. When AI fully takes shape
Across all sectors of the economy. And that's something I [00:56:00] think, wanna avoid and not go to civil war.
Jeremy Au: Yeah, exactly. And you know, if people don't have jobs, then people don't, their families
Then people are angry. Right?
Audio: Exactly.
Jeremy Au: And it's understandable. So our job is help everybody.
catch up. I mean, it's like, cars replace horses. Right? You know yeah. So what do you do to make it happen? Right? Yeah. ~And then today nobody wants to go back to a world where everybody travels the horse.~
~On that note, thank you so much for coming and sharing. Yeah, ~
