Lý Hồng Nghị: Thủ tướng Google trở thành Lãnh đạo GovTech, Mở rộng Cơ sở hạ tầng Kỹ thuật số & Chống lừa đảo bằng Hệ thống - E559
“You don't have to have management structures be purely control systems—they can be focused on empowerment systems as well. I think the assumption that most people make is that management structures are purely control structures, which is that the CEO, or the minister, or the Perm Sec, or whoever has a thing they want, and they tell the director, who tells people, and everyone's job is just to unlock. It's a massive control structure in order to do what the boss wants you to do. But as you know for yourself in VCs, yes, you need a little bit of that. But sometimes what you actually want is an empowerment structure.” - Li Hongyi, director of Open Government Products
“The concept of ‘put in where you're parking and pay money’ is a very simple concept. It's not a big idea. In fact, I think everybody who's ever used a paper coupon is like, ‘Why you make me do this? Why can't I just use my phone?’ The idea is not hard. The difficulty was getting all the stakeholders to want to do this and to integrate with all of the enforcement systems to actually make it happen. Just to give you a sense of scale—building the actual software for the app probably took two, at most three months, at least for the first build. We obviously iterated on it since then. But the real challenge was realizing that for most of the parking lots, we didn't actually have digital records of how much they cost.” - Li Hongyi, director of Open Government Products
“My first experience with a really good boss—not that my other bosses were bad—but the first time I really felt like, ‘Wow, I learned a lot,’ was with my manager on the search team, Peter Lindsley. He was the PM on Image Search when I was there. I remember when I was looking around within Google and talking to different people, I spoke to him and he was just really excited about all the things we could do. When you talk to other people, they’d say, ‘Oh, why are you interested in this?’ and you’d give some ideas, share some thoughts, try to solve the problem like a student putting on their puzzle-solving brain. But with Peter, I remember talking to him and saying things like, ‘Oh yeah, for image search, we could put a big map at the top of the page when someone searches for a place.’” - Li Hongyi, director of Open Government Products
Jeremy Au sits down with Li Hongyi, director of Open Government Products, to explore his journey from aspiring physicist to building digital tools for public service. They discuss agency, leadership, and the realities of driving change in government—from the impact of a Google internship to lessons in management and building systems that protect against fraud.
1. Dreaming in equations: As a kid, Hongyi wanted to be a physicist—he loved thinking in systems and solving puzzles like the resistor cube challenge in secondary school.
2. Systems over subjects: He saw physics, economics, and computer science as different languages for modeling how things work and how people behave.
3. Google made it real: His internship showed him that his work could help real users, shifting his focus from theory to practical impact and leading him to computer science.
4. Agency isn't given—it's taken: He realized in university that waiting for the “next step” wasn’t enough. A friend pushed him to apply to Google, which changed his mindset.
5. Becoming a manager meant unlearning: Early on, he micromanaged engineers. Over time, he learned that great managers remove roadblocks instead of redoing others' work.
6. Empowerment beats control: Inspired by his Google boss, he now sees leadership as creating the right conditions for others to thrive—not just setting direction.
7. Parking.sg was the easy part: Coding the app took 3 months, but aligning agencies, digitizing data, and syncing with enforcement took 8–9 months.
Jeremy Au: (00:00) Hey Hongyi, good to see you!
Li Hongyi: Good to see you too, Jeremy.
Jeremy Au: Yeah, I've always been following you since our secondary school days and we've had some parallel tracks in technology, California, Boston. And now glad to be here.
Li Hongyi: Yeah, no, thanks for having me. It's definitely been a long time.
It's funny how even after school actually the world isn't that big and you end up going in very similar loops.
Jeremy Au: I was going to the tech loop and entrepreneurship, and trying to make a difference. I would love for you to introduce yourself.
Li Hongyi: Sure. So yeah, my name's Hongyi. I run a team in the government called Open Government Products. We are a software development team. We built form.sg. We built the vaccination portal during COVID. We got 40 something different products. It's a team of engineers, designers, and product managers mostly.
Before that, I was at Google for a couple years. Worked as a PM, worked on the search team. Worked on the infrastructure team. And before that, I did computer science and economics.
Jeremy Au: Yeah. Amazing. So I know we were in secondary school together, the same classroom, and we went to the same JC together, which is crazy in retrospect.
Who did you want to be?
Li Hongyi: When I was in secondary school, I really wanted to be a physicist, (01:00) for most of my childhood, I wanted to be a physicist. Because all the topics in school physics was the one I was good at. And you read these books about astrophysics and quasars and black holes and all these things, and they look pretty cool. And yeah that was my aspiration in life was to unlock mysteries of the universe.
Vâng.
I actually even up until applying for Uni, I was still thinking about applying for physics more than anything else.
Li Hongyi: And I only really changed my majors once I started Uni.
Jeremy Au: Yeah. Now, I remember in my secondary school, very clear memory was we had a physics teacher, Mr. Ng.
And I think he laid together this challenge where he was like, this is the electrical diagram from A to B.
Oh, I remember that. He was like, this is a really difficult program. Go find me later during the break. When you get the answer and then you straight away raise your hand, and then you gave the answer straight away.
Li Hongyi: Oh yeah. It was a resistor cube. I remember this.
It was a cube of resistors. And with each length of the cube, there was a one Ohm resistor and you were trying to figure out what there was a net resistance of the entire cube.
Jeremy Au: And then you gave the answer Correct, just mind blown.
Li Hongyi: I remember the puzzle, It was quite a fun puzzle.
Jeremy Au: Yeah. All I remember it was like flabbergasted, (02:00) and then you were like, no, it's very obviously the answer. And then I was like, mind blown.
Li Hongyi: Because the edges of the cube, 'cause cubes have multiple faces and then you have the branching things and essentially the solution was you have to recognize that after the first bit, the three points
had to be ISO potential and therefore you could normalize it into a simple three series and split into parallel. So, the first one splits, the second one splits into three again, and then you just normalize things that way.
Jeremy Au: I know. So you see, so to you obviously, it's very clear to you— that problem. To me,
I just, I was just flabbergasted. I was like, I don't even understand the problem. I don't even know where to begin. I was more on the bio side than the physics.
Li Hongyi: Yeah, I know. I'm terrible at bio. Not a good bio student at all.
Jeremy Au: So, we all have our, we all have things we suck. I have things. Yeah. No, it's amazing. I remember you are always really good at math, you're really good at physics. Okay. Yeah. Then you're doing quite a lot of that in junior college and after that you went off to university and then, there was a time when you were debating between economics and computer science.
Li Hongyi: Yes.
Jeremy Au: So to me it was actually even more confusing, because this guy's going to math for physics.
Li Hongyi: In JC, my dad lent me this book, it's called "The Armchair (03:00) Economist".
And it was like a kind of casual reading book but it was laying out like how to think like an economist in very sort of layman terms and as a casual reader. And I remember reading it and thinking this just makes sense to me.
Yeah. It's like the kind of thing where like, you have been thinking this way your entire life and someone puts it down, and I'm not very much of a reader. Yeah. Like I get very tired like after 10, 20 pages. But that is the first book I have ever read. I think the only book I've ever read cover to cover in one sitting.
Yeah. And after that it's not that like I like math or I like physics, or I like talking about the economy or whatever. It's that, I like systems. Right. And for me, physics economics is the same thing where it's okay, imagine you have whether it is mass, force, acceleration, or supply demand curves.
You have these properties of the system and here are the rules by which they interact. How does this system evolve? And whether you're calculating the trajectory of a ball or calculating what the equilibrium price is for something,
see (04:00) it as the same thing just like with different formula. And I think that's the part that made me really switch over. And even when I was in Union, my still plan was to do economics and physics. So I did a bunch of economics. I did a bunch of physics classes.
But I think the thing that really turned me onto computer science was when I did my internship at Google. Because up until then, 'you're just a kid, you're a student, you are writing papers.' The biggest piece of work you've done is like a paper that you wrote that your teacher and maybe your parents read, and you're not even sure if your parents read it.
But when I did that internship, even though I was just an intern, they gave me work to do. And it's not major work. It's like small feature works on small things like that. And I realized that I was doing things that people were actually using, like actual humans were getting value from.
I was working on what later became the Google Play Store and what later would become Google Notes or Google Keep, so to speak. It was a similar form of thinking whereas you're trying to figure out how to design and figure out the system of how people interact with it.
It felt very self-satisfying to know that your work was going somewhere other than just showing that you could do it. And I think to me, that was the biggest gap like, there are very good (05:00) physicists and there are very good mathematicians and are people who are way better at that than me—
and so, I don't feel any need to prove how good I am, like how capable I am. I think the thing that I found really satisfying was doing things that people found useful.
Jeremy Au: Right.
Li Hongyi: And so, I switched over from physics to computer science and that's where we're today. I still finished my econ classes, so I technically have an econ degree still, but I don't really use it all that much. I will say that I was a much better economist than I was a computer scientist. So I feel like this is weird things that end up in life of the things that you end up doing.
Jeremy Au: It's interesting 'cause I had a parallel pathway. I was like in a computer club and I was never a really good programmer. And so, economics was something that clued me in I went in that direction. I'm just curious what was it like for you, because, we went through like secondary school, JC, Uni—
what do you think you discovered about yourself during that timeframe?
Li Hongyi: I think the biggest thing that shifted for me was just the, like, how much agency you have. I don't know whether it is true for everyone, but growing up in Singapore school system, you like, just (06:00) follow the path laid out for you. Okay, you have this exam coming up, study for the exam. You have this homework, you do, I never did my homework, but whatever.
The next step is laid out for you. And I think it was only in, I wasn't a very good student throughout secondary school, but it was only like towards the end of secondary school that like, I did not enjoy ACS. I know, I know a lot of people do, and it's fine and it's great.
I did not enjoy ACS. And I think it was only towards the end of secondary school that I was like, I want to be able to go where I want to go. It's not a bad school, but it was not my first choice. And I wanted to, I decided for myself that I wasn't gonna be stuck anywhere. I remember I asked the teachers to throw string school. I was a terrible student, but for all levels I studied like crazy and I was very determined, and I was very focused. And I did well.
So, I got to RJ and I think this is, I know I'm conflating like the difference between JC and secondary school, and RJ and AC and all the things, but I found that like, when I got to JC and there was a lot more autonomy in terms of what you could do with your life; like you, you choose your classes, you spend time—
it felt like there was a lot more (07:00) trust in you being able to. I was so much happier. The teachers had a lot more trust in you to ' okay, you figure this out. They won't chase you over anything.' And as long as you got there at the end of the day, I think they were fine
conflicting multiple factors. But that was, I think the next big shift for me was when I went to Uni, I was a government scholar doing that and I realized now, very unhealthily, I was, I was just 'oh the track's kind of laid out for me.'
I got the scholarship. I will come here, get decent grades. I'll go home, we'll start doing work, and I don't really need to try that hard for a career because again, all my friends were applying for jobs and internships and stuff, and I was just like, 'ah, just take the day off and go play video games or do whatever.'
It was only like, I think, a third-year of Uni that a friend of mine made me apply for the Google internship. I remember this because I had this very kind of unhealthy attitude being like, 'oh, what's the point? There are so many other people, so much smarter than me, like doing this.
And they've had so much more experience in.' I remember she was saying like, 'no, just stop (08:00) it. Just try. If you don't get it, that's fine, but just try.' Yeah. And so, I did. I applied for it, and for some reason, I did reasonably well enough at the interviews to get an internship.
And I think that changed a lot. This idea that rather than waiting for what is laid out for you, you can take action to do it and you can determine what do you want to work on and how you work on it and you can choose things. And I really liked that experience. And it was very, I felt so empowered that summer, that internship, not because I was doing like, the most mind blowing work or whatever, but just that people trusted me to do things that other people would use.
I think from there, it shifted my attitude a lot. So rather than just be like, okay, what's the system? What's data? Okay, let's go with that. You asking yourself, no, but what do you want? What do you want to do? What should we be doing?
And then making that happen. So, I started being a lot more proactive, which is a very big part of who I am today.
Jeremy Au: Yeah.
Li Hongyi: Yeah.
Jeremy Au: Did you have any good bosses or managers that changed how you saw agency or work?
Li Hongyi: Yeah. I think my first experience with a really good boss,
not that other bosses were bad. But my first experience with a really good boss— who I really felt like wow, I learned a lot (09:00) from —was my manager on the search team, Peter Linsley. He was the PM on image search when I was there.
And I remember when I was like looking for within Google and I was talking to different people and I spoke to him, and he was just really excited about all the things we could do. Because we talk to other people, they're like, oh, why are you interested in this? And you give some ideas, and you give some thoughts, and how would you solve this problem?
You put on your student brain and try to solve the puzzle. But with Peter, I remember like talking to him and being like, ' oh yeah, for image search we could put a big map at the top of image search when you search for a place, or like when you search for shopping results then instead of just showing you random pictures, we specifically focus on things that people could buy because that's actually what you wanna do.
Or when you search for celebrities, rather than just showing you a hundred pictures of their face, you could show them at a concert, them out and about.' And when you talked to him about it, he was like, 'oh yeah, absolutely!'
Yeah. This is how you do it. Like imagination of the possibility that excitement of there are all these things we could do. Let's just figure out which ones we do first and not. I think that really changed a lot. That was a really good thing. For me, the thing was just that perspective of let's try to, like he really wanted. He was (10:00) excited about seeing the things that we thought about and like we really wanted to just try to help me see if we could get those things done.
It wasn't about oh, because you're just a junior PM. It wasn't about oh yeah, here's this agenda and everything. He passed it to me and tell me things do. It was more about yeah, let's get you to think about all this stuff and,
and how can I as a manager help you make that stuff happen? It really shifted my mindset which even to today I think is a very core part about how I run my team. So just that, you don't have to have management structures be purely controlled systems. They can be focused on empowerment systems as well.
I think the assumption that most people make is that management structures are purely control structures. Which is that, the CEO or the Minister, or the Perm-Sec or whoever has a thing they want and then they tell the director
who tells people. And everyone's job is just to allow, is a massive control structure in order to do what the boss wants you to do. But as you know for yourself in VCs, yes you need a little bit of that, but sometimes what you actually want is an empowerment structure.
Sometimes, what you want is to find good people, especially when it comes to like, more (11:00) abstract knowledge work and things like that. It's not that I already know the answer and I need you to do it, it's that I need to somehow find the people who I know can figure out things that I can't figure out and support them in doing so.
And you are trying to empower people and I think that was my first experience like yeah, with a really good manager.
Jeremy Au: Oh, that is amazing. And that sounds really inspiring. And I think, what's interesting is that you're starting to talk about two things, right? Which is like the organizational cherry and also, your own personal experience as a manager.
How do you think you've evolved as a manager over time? Because once upon a time you're an intern, they became an individual contributor, they became a manager; now you're a manager of managers. And also you have a lot of experience now, so you know, across different products.
How have you evolved as a person?
Li Hongyi: I think I've recognized a lot more about what people need to do well. And that seems very abstract, so maybe let me elaborate.
When you first start work, your primary challenge is how do you do this job well? Your boss gives you task, you figure out how to execute. You build your skills, whether programming or talking to stakeholders. When I (12:00) first became a manager, I had this mental model of 'I know how to do well, I'm going to get to do the things that I would have done.' When I had my first couple of engineers reporting to me, I was very micromanaging. I would ask them write me a code. And after, I would read through every line of code, I would format this way, and I would go and get it to be exactly as the I would have done it.
And I think there are industries and there are jobs where that is what you want. But I found that I wasn't actually very productive. Because then, we ended up just doing the work more than twice because they would have to do it. Then I would have to essentially look through it and do it as if I was doing it.
And then we have to like dingdong back and forth together and it wasn't very productive. So, I had to try and figure out what was I trying to do? And firstly, I think I to learn to let go of a lot of the details.
Effect, like good use of brain power. Yes, you are right. You could spend your time doing this and maybe you do a little bit of it more at the beginning so that they get the vibe of what you're trying to accomplish. But over time, actually what you want is you are not trying to get them here to do exactly what you would do.
You are trying to get 'em there to do things because you don't have time to think about it. And so you get your staff to, I think that's the (13:00) first thing I learned, which is like, what things to let go of. I think the next stage I had to learn was, as the team got a bit bigger and going, is I think I would say 70% of good management is just trying to be a good manager. A lot of managers don't recognize that's a goal. They see management as a position rather than as a responsibility.
It's oh, I am the manager now and I get to. Therefore, I get to tell you what to do. But I think for me, the big shift was like, actually reading just even a couple of books on management and things like that, and really understanding that management was a thing that you are like, really, what are you trying to do as a manager?
What are your responsibilities? You have, your job is to make sure everyone understands what the goals are. Your job is to figure out where people are blocked and unblock them. Your job is just to try and help people.
And don't get me wrong, there's a lot of very deep expertise in this. But I have found that the top determining difference between really crap managers and really good ones whether or not they want to be a good manager, once you just start having one-on-ones and keeping notes with (14:00) your reports and trying to talk to them and understand what the challenges are, actually you get really fine. There's still another 30%, which is really hard to get through, but 70% of it is just wanting to be a good manager.
Jeremy Au: Yeah.
Li Hongyi: I think where I've grown to now,
I mean there's a lot of things I've learned, but my most recent thing that I've sort figured out is at a kind of meta level. What are the conditions you need to set up for people to work together well? Because everyone knows some teams work very well and some teams don't. And it's very easy to chalk it up to culture or just personality or whatever.
And you're not wrong, but as someone whose job it is to run an organization made up of literally dozens of teams, you can't just hope that you get the right personality. You need to understand what allows some teams to have very good culture. What allows some teams to have very good personalities working together?
Why they work and why, and what needs do they have? it is not just skill. I think that was one of the mistakes that I made. I used to think that if I just hired really capable people and put them together, they would figure things out. You need to make sure that they have clear goals of what they're trying to work towards.
You need to teach them some basic human interaction skills because funnily enough, you think all of us of university, you think you're fully (15:00) formed adults, but actually in retrospect, you realize how much you're missing a lot of social skills. You need to figure out the right kind of stress and pressure to put on the team. No stress or pressure results in the team not even having to be there because there's no reason for anyone to try and accomplish anything. You put stress or pressure in the wrong areas. The team starts turning on each other because anyone who's played board games or video games knows this. You put the right stress or pressure, the team comes together and really rallies and works towards something.
Simple things like just having the clarity of the goal and putting pressure on 'Hey, let's get this outcome.' Rather than trying to be punitive on particular behaviors. Be clear like, I care about this, therefore I am trying to get you this.
Just having that one extra sentence of 'yes, I'm blocking this, but because of this, I think it gives people psychologically, a lot of alignments where they're going as opposed to just trying to dodge, getting zapped. I don't know. This is something I could talk about forever, but hopefully I am a better manager today than I was when I started.
Jeremy Au: And I think what's interesting is that, obviously, it is always a challenge for people to learn to become managers. And what's interesting is that you're doing this in the context of the (16:00) government, right? Yeah. And public service. Yeah.
Which is another thing altogether because for me oh, you pull up an MBA, there's a manager in a corporate setting, and targets are very clear. But I think government is also a different, like animal, right? So, I'm just curious, like, how do you feel like some of those learnings have been in that sphere as well?
Is there more stakeholders? Is it more organizations? Is there a different set of metrics to go after? How do you think about that? How is the art of management different within government, from your perspective?
Li Hongyi: I know people say in private sector, profit is a very clear driver where government, where public sector doesn't have that. And that is true, but that's only true for the most senior positions. For the vast majority of people in, let's say Grab or Netflix for example, they're not responsible for the profit of the whole company.
They're not even responsible for the profit of their division. Most of the time, you have the same kind of proximate anchor which is bosses want this, how do I deliver that and get the team together to deliver that? And you get a good feature out and get people happy and using it.
I don't think that on a day-to-day basis, the designers in Grab and the designers in OGP are like dramatically different in terms of oh, this one I'm trying to (17:00) squeeze. I think they're trying to do good design for what the users will care about.
That is true if your goals are correct. I think, it depends at the level where you're going at. So I think, if the goals are already set up fairly clearly, the work is actually quite similar. I would say, that maybe one of the big differences between government and private sector is that in government
you have to do quite a bit more work to figure out what your stakeholder wants from you. In private sector, generally speaking, if you show that this will bring them revenue and profit and stuff at a high level, yeah, you go. And because of that, when you talk to the junior level staff, they know and the whole org knows that if you give this goal, if you help them do that, they will be able to bring that to their boss and their bosses. What I found in government is that the different departments can have completely different goals which have nothing to do with each other.
And when you go around talking to people about what is good for the public? People were like, nod and technically yes, but what does that mean for me? And so I think a lot more of the work. I have to do is to (18:00) try and figure out okay, what is it you care about? What are your goals?
How are we going to clarify that angle? 'Cause government tends to be, it's a difference between optimization and kind of like, caretaking. I guess house sitting and so to speak. The analogy I give people is, let's say your friend asks you to house sit for them. You go live in their house and you make sure everything's clean and the trash gets taken out and things like that. If a contractor or a salesman came to you and be like, 'Hey, I can make your house better.
I can install these fans, I can change these slides. I can do this, and the house will be better.' You would look at them and be like, dude, you're crazy. I'm just here to take care of the place. What the hell are you doing?
That is a very similar perspective to have where you are, whether you're the owner versus the caretaker. Yeah, and I think in government, we rightly have fairly heavy caretaker mindset where it's okay, look, a lot of people have built this place up before me. I am now here and new in this role, and it doesn't matter whether you show me that this house can have triple it square footage for basically no cost and increases risk of value, whatever, it's not my house.
Jeremy Au: Right.
Li Hongyi: I am just here to make sure that it is clean and in the order that it was left in.
Jeremy Au: Right.
Li Hongyi: And if you start calling your friend, be like, 'yo, this contractor (19:00) came in.' You're like, you're crazy. So, when you want to get something done in that framework, you need to find the people who have the kind of owner mindset rather than the caretaker mindset.
Because for them they're like, oh yeah, great. Increase the resale warehouse. We'll do, we improve this? I'm gonna live here. And that I think is the biggest challenge. In private sector, it tends not to be a problem because almost everybody there is clear that if you were to increase the revenue of the company and increase the profit or increase efficiency, someone would benefit somewhere.
And you can find that. Whereas government it's a, it's not wrong, it's just a very different mindset.
Jeremy Au: Maybe we can use that example like, Parking.sg, right? So, my mom loves this app, right? So she, one day she just messaged me all of a sudden she was like, 'wow, this app is so amazing!
And I love this app.' It's so much better than whatever. And I was like wait, this is home. Yeah, baby, that you the product. So I laughed and I said my mama message is oh, actually my classmate did this.
Li Hongyi: So, the idea behind it was very simple. And to be clear, we were'nt the first people to have this idea. I think, in almost every hackathon you have in Singapore,
Jeremy Au: Yeah.
Li Hongyi: there was a bunch of private sector people who were trying to get the government to do this.
There was a bunch of hackathons (20:00) demonstrating it. And the concept of put in where you are parking and pay money is a very simple concept. It's not a big idea. And in fact, I think everybody who's ever used a paper Google was like, why you make me do this? Why can't I just use my phone?
The idea is not hard. The difficulty was getting all the stakeholders to want to do this and to integrate with all of the enforcement systems to actually make it happen. To just give you an idea like just a sense of scale. Building the actual software for the app probably took two and most, two, three months about that, at least for the first build.
It'll be obviously iterated in it since then. But like, the difficulty was realizing that like, most of the parking lots, we didn't actually have digital records of how much they cost. Oh, interesting. So, the record was the sign. And so we had to go through HDB and URA and LTA and dig up all their archive records of the database of what fees they put for different things.
And then we digitized that, and then we put it into the app, then we realized that what was set in the record and what the sign set on the ground were different things. And then you have to fix that. And then you have to figure out, okay, then how do you communicate to the enforcement officer that like,
this guy has paid. Do they check your records? Do they look it up versus the (21:00) price? If I put the right amount of money, but I put the wrong parking lot, do you find the guy? Like, that was the challenge. The concept of paying for parking using your phone is,
other countries in the world have done it. If anyone who's gone traveling and had to pay for street parking, somewhere has done it. The difficulty was getting it integrated into our data and our enforcement, and letting the URA officers update the parking fees and when have to change prices or change number of lot or change the boundaries of the lots and which one was where and things like that, the big concern that the government officers had was that if we did this system.
Would it dramatically slow down enforcement? Because, you literally have humans on the street going around, checking everyone's things. And if you don't make it smooth for them— Yeah. And now, yeah, congrats, you've got it. Very convenient for customers, but like, your enforcement operations is now slowed to a crawl.
That was a problem. So, we had to like, take videos of us, like showing okay, this is what we take. You take a picture of the license plate, or you just keyed in on the keyboard or something and we were timing it to show that the rate of progression was the same.
So, you can see how the product isn't so much like the software itself. The product is like the (22:00) software connected to all the different touch points that it needs in order for it to work smoothly in our system.
Jeremy Au: Because the UX is very clean, very simple. Yeah. Very consumer friendly.
Li Hongyi: Yeah, the app itself is relatively simple. It is the operations and integrations that really, that I would say. So, at building about three months getting everything else done was like eight, nine months, something like that.
Jeremy Au: So, it's a little bit like an iceberg, and I think we see like different versions of this, right? You've also done forms, right? Yes. For the government. I think, there's also some anti-scam stuff that you're working on as well. That's a big part of, so maybe we'll talk about a scam site, for example.
It's a crazy problem, right? Singapore has rich people; scammers exist, criminals exist.
Li Hongyi: We are clearly very rich and very stupid, and therefore we get targeted.
Jeremy Au: Yeah, exactly right. So it's like a piñata where they keep hitting us and money still keep flowing out.
So I feel like everybody's all rallying, DBS bank and moving towards fixing it. All of us who have elderly parents are also like warning them all the time.
Li Hongyi: This is a problem I've been trying to work on for probably eight years now.
I remember when I first looked at it, I think this was (23:00) 2017, and it was a lot smaller then, but you saw that year-on-year, the amount of money we've lost to scams has been increasing exponentially at least 20% year-on-year .
I think, last year was 60% or something insane like that.
Jeremy Au: That's the best stock market performance
Li Hongyi: ever.
Jeremy Au: That's crazy.
Li Hongyi: And the thing you have to understand that, I think the mindset shift that we need to have is that you are going to lecture people out of this. Yeah. I think it's quite clear at this point, despite all the PSAs, and the posters, and the like warning messages and notification, it does jack all. You cannot think of this as a like nagging thing.
You have to think about this systematically. Because I think, the mental model people have about scammers is that, ah, someone offer you like, investment opportunity to good to be true, you'll surely take it. Yeah. You, any of you get scammed, it's, it is morally right that you lost your money 'cause you're stupid.
But if you look at them nowadays, they are really sophisticated. Right. The message that you get from the scammer and your bank are exactly the same. The only difference is that the link is like a couple of characters different or something like that. You cannot expect people to
parse all that and make that (24:00) call. In fact, there was a problem at some point where the scammers could literally change the sender ID of the message and send it in the same thread of conversation as your actual bank messages. So it'll be legitimate DBS messages, and then there'll be a message from a scammer, literally in the thread, and you're supposed to know that this one message out of it and that they could compromise and call themselves DBS.
That's not reasonable. Right. You are not gonna solve this by lecturing people harder. You need to figure out systemic solves. Right. You need to understand, like from a system perspective, what were the authentication mechanisms? What were the identity on that authentication infrastructure that you were supposed to be able to do, and why were they not working?
Because rightly it, again, this is just showing the weaknesses we have as a high trust society, right? The government would randomly call you from random phone numbers and tell you to pay fines or pay things. They would. Like, part of the modus operandi was just like, oh yeah, there's a fine
someone needs to pay, or there's a issue that we need to call them up for. Just pick up a phone, call them, declare that you are from the government and because we're Singapore, they will comply.
Jeremy Au: Yeah.
Li Hongyi: That is clearly (25:00) a vulnerability. And when we weren't as connected to the rest of the world, that vulnerability went unexploited, but now that we are clearly irrevocably connected to the rest of the world, that vulnerability gets exploited.
You need things that we are taking for granted. You need to be explicit about. You need firstly, proper digital identity. You need not just identity, but authentication mechanisms. You need to be like, how can you prove that you are this person?
Whether it is something like Singpass, whether it is some other mechanism— you need an authentication mechanism. Once you've got identity authentication mechanisms, then you need to figure out like, checkpoints and safeguards for these things. If I try to sign into my World of Warcraft account from another computer at a weird time, I get an email being like, yo, something's weird here. Can you please double confirm? But if at 3:00 AM I decide I wanna transfer like a million dollars from my bank account to some random guy in Vietnam, you're like, yeah, totally
cool. Go ahead, do that. That doesn't make sense. How can it be that your World of Warcraft account has all these anomaly detection and checks and all this, but when you're transferring huge amounts of money at random times to random people, (26:00) they're like, yeah, go for it, man. What do you want me to say?
Don't get me wrong. There are always gonna be like the more sophisticated scam victims where they've really captured the guy, they really calmed the guy that you know, and the news. The police will go and like really try to convince them, and no matter what the police tell them, they're not gonna believe it because they've been capture it.
But there are so many intervention points before that you could have gone in and been like, 'Hey, this is weird. This person isn't actually a policeman. Brad Pitt is not actually in love with you.' And so many mechanisms downstream from that where you're like, okay, but this bank account clearly has been receiving like a lot of weird payments in the last 24 hours.
Maybe we should stop that or lock that down or something. This is a entire class of problems. Just like crime is an entire class of problems. You are not going to solve crime with just one app.
You need a justice system. You need a law enforcement system. You need investigation officers. You need prisons, you need judges. You need legislators and lawyers. You need that whole thing to get the physical criminal justice system to work.
You are going to need that same level of industry in the digital crime space. You're going to need law enforcement mechanisms, you're going to (27:00) need training mechanisms. You're gonna need a way of chasing these things down.
You're going into ways of blocking and censoring avenues of attack that the people come in from. And it's going to be continuous just like physical crime is continuous. I think, hoping that there'll be a quick fix to this, is the fastest way for us to basically fail.
Vâng.
Jeremy Au: No, I love what you said. I think a total mindset shift on this whole scam side. And I, myself had credit card fraud from Singapore and I'm just like, I just have to physically check my statements. And I work very hard not to be defrauded, but it's really,
Li Hongyi: It's really hard. And lecturing people and expecting the mass of everybody to do this and ah if you don't it's your fault, is just abdicating our responsibility. Yeah, exactly. We need to build systems to protect people. And that's how Singapore got here. There are other countries basically say the same thing to physical crime, which is that there are some neighborhoods where it's just not safe to walk, buy a gun and take care of yourself.
But in Singapore we decided, we don't want people to have to be worried about their lives when they walk around the country and we don't want people to be worried about their digital lives as they do anything online.
Vâng.
Jeremy Au: I agree with you a hundred percent. I think that's a very good point. I think, what's interesting as well is just (28:00) there's also like some sort of geographic, dynamic to that as well, right? Because we're talking about a defense perspective. But also, they're known, these scam centers are, in obvious Asia, this whole industry.
And I think China's so frustrated to the point that they're basically bankrolling, military and other interventions to just extract the nationals from call centers out in the Golden Triangle.
Li Hongyi: I am not informed enough or smart enough about this space to tell you what the geopolitical solution to financial fraud is.
What I can tell you is that for us at least, rather than focusing on what people should not do, focus on what people should do. You saw this with the banks for example, previously, right? We keep telling people, ' don't pick up these phone calls. Don't fall for the scams.' But what should you do, literally? 'Cause DBS will message you to login things and send you pin codes, the scammers will message you. LTA will call you with phone calls, the scammers will call you pretending to be SPF. And I think, if you are trying to patch all the holes in the ship, you're going to fall over.
So, instead, I think the strategy we (29:00) need to have is what are the protocols we use for working with government and working with banks and working with hospitals. It's gonna be a lot of work to get people there, a lot of work to train people in these protocols. But figure out what you should do rather than telling people what you should not do, and then move things onto there so that you are at least protecting your particular important parts.
But then the people rightly have the question, if I want to do a transaction digitally with my bank, what should I do then?
Jeremy Au: Yeah.
Li Hongyi: How do I make sure that this is secure? And don't tell me, look for the plus six five, because that was a stupid message we put out. And immediately they compromised like in no time flat.
Jeremy Au: I remember that actually because it was like, if the call has plus six five,
Li Hongyi: It was such garbage like, because anybody was like, like it took like literally like less than a couple of months for the scam be like, okay, they're looking plus six five, we just put plus six five for our numbers though it failed to understand like,
like what the actual vulnerability was. It was not a wrong heuristic, but it wasn't. Yeah. My personal view, I think just as the government provides NRIC and like (30:00) infrastructure and trust. We need to provide that same level of infrastructure and trust around digital transactions. So we need to get our digital identity system working well and ubiquitously.
You cannot just be like, oh yeah, but don't use this, but don't also reuse your passwords and don't fall like, we need to be, we need to provide a reliable mechanism, actual people in Singapore to verify each other digitally. Now this could be singpass, it could be something else.
I don't particularly care, but this is just infrastructure that government needs to provide. Because the alternative is for the private sector to do this and the only alternative that was like world coin or something like that. And they had a whole bunch of fraud and associated stuff with it, because of course it would.
Yeah. Once you have that, then you need to have law enforcement mechanisms looped into this data fairly real time. And apart from just providing the authentication mechanisms, you need to be seeing let's put it this way, right? The scammers move money within minutes.
You could imagine putting in mechanisms where like safety valves, right? No one, I don't know anybody who really needs to move $10,000 within a fraction of a second.
But actual consumers who don't have that money to lose, (31:00) no one needs it in a fraction of a second. And it does, it's fine to check slow, authenticate, make sure it can't just disappear like that. You could imagine a whole bunch of verification mechanisms you have in place. You could imagine for example, making sure that bank accounts are tied to specific people, but like, when the pattern of anomaly like spikes, you need that person to show up in person to my bank account.
And yes, it'll cause some inconvenience to some people, but you need to do that to prevent money. I think we need to go target this from the supply chain as well. There needs to be some liability on media platforms and communication platforms who push this stuff out because otherwise they're just not gonna do anything about it.
Facebook, for example, still has a lot of the, you see these sponsored ads, which are clearly scams. It's like fake deep, fake whatever. And they are getting paid for these ads, and obviously it's gonna continue if you keep getting paid. So you must take some action around that.
There must be some responsibility around that. I wish I could tell you I had an entire plan in place of how to fix this. I don't, I just have a vague sense of the different components that we are going to need to start working on and it's not gonna be a quick fix, but I do know we need (32:00) identity and authentication.
I do know we need some detection and intelligence around it. One of the things we're doing with Scam Shoot for example, is we're trying to build some kind of collective intelligence. Because previously, people would just get targeted individually, but with something like scam shield, if 200 people get a scam message all at once, the government should know about it instantly. They shouldn't have to wait for a few months later where people report the crime.
Jeremy Au: Yeah. I think another part that's interesting is I think a lot of people, That's how mindful a lot of folks is. People are very anxious about the vote today. There's the US, there's China, there's decoupling, there's a lot of stuff. And I think, what's interesting is that, it is less about that global stuff, there's sometimes the sentiment I talk to people and people are like, oh Singapore's doomed.
Like Singapore's very small, Singapore's very fragile. We can't defend ourselves and the world is going sideways. And it's kinda interesting because somebody's come back and I'm like, yeah, I'm helping. I'm just curious Yeah.
Li Hongyi: Look, I can't convince you to give a shit, basically. My view on this is that, if your main takeaway from consuming all this is that the world is doomed and we should just lie down and die, that is something that if you believe there is no logical argument to persuade you (33:00) otherwise.
Jeremy Au: It's fear.
Li Hongyi: Because every part of our existence is just a matter of whether or not we want to. Singapore as a country, only exists because we decided we wanted to. I think that's something people need to realize. Like, it is not that ' oh, this was such an optimal place' and you know how like oh yes, we had a deep harbor.
Yes, that's true. There are some advantages to us being here, but basically there were, if you look at any of our history, a hundred points in which we as a country could have collected, you just said. 'Ah, screw it. Not worth the trouble.' And just either merge into the British Empire or merge with Malaysia or merge into part of Indonesia, or just have everyone dispersed and just become a village and a footnote of history.
The only reason this country exists is because we decided that actually we could do something here. In defiance of everybody else in the world who basically were like, whatever, who cares about his backwater crappy village? Because no one was going to care about us but us. And because we did, we worked together, we created a vision and it requires us to coordinate. And if only a small fraction of us believe this, but enough (34:00) Singaporeans, enough people who are here believe that we could do something better, that we said, we'll put our legs down and we'll try to push this forward.
And we did. And that's where we are today. It is, what if it wasn't for that, we would just be like a random patch of dirt in the ocean somewhere where someone thought they saw a lion once, but probably didn't. That's what Singapore is. That's all we are. So, if you don't think that this is worth doing, okay,
I can't argue with that. That's up to you and good for you. I think that if you look at what we've built and where we've come and the people we have achieved something here that no one else in the world has been able to achieve.
A good place. A place where things work, and there are a lot of problems and a lot of issues, but things work and we have a chance of making it better
But we have found a small place in the world where we have more or less gotten things right, where people can go outside and they're not worried about getting shot or killed, and we have a place where healthcare more or less works and it is worth something.
No one's gonna convince you otherwise. But I'm hoping that enough (35:00) people understand that this is worth something and know that enough other people also think it's worth something that we can do something together on this. Because yeah, like
if you're right then you build something great. And if you're wrong, then it doesn't matter anyway. If we're all doomed. You might as well go out trying. What's your argument? Your argument is we are all doomed so I just wanna lie down on the sofa and cry until the apocalypse comes.
Yeah. If we're all doomed, then there's no harm in trying to get something built. And I think that, and got something good here. And if you feel like you are too scared and too worried and you feel like your better option's, okay. But we are only here because we decided otherwise.
Jeremy Au: I think interesting part is, what you've reminded me of is when we disaggregated, I think one part of is fear. One part of it is the perception of helplessness, right? Singapore small, so we can't change anything. Then, also there's a bit of the flight response, right?
It's like either flee or fight, or freeze, I guess is the other response dynamic. And the last part of dynamic is also like how do I even, where do I even get started? You know what I mean?
Li Hongyi: one is to be honest.
Basically, you need to decide for yourself. (36:00) Like, it's a very weird result of, that is quite replicable and obvious through modern social media, that if you talk to someone enough, you can convince them to lay down and die. Which is the attitude that you have a lot on social media, which is that if I spam you with enough gloom and doom and frustration annoyance, you get so wrapped up in your head that your decision is just, ah, fuck it.
You need to decide. You don't want to do that. I think that's the first point. beyond everything else, there are a lot of places you can start. You can volunteer with people, you can work for social enterprises, you can try in your small ways, in small places to make things better.
There are a lot of people who need help. I think that's the attitude that I have. That's the framing that I take, which is that, if you believe there's nothing to be done, then okay. But if you look around you, that's obviously not true. There are so many charities and government organizations and people who are trying to do things, and all of them have problems. And none of them are going to solve this in one step.
If you think you're going to come in, do a bunch of work, and then the world is safe tomorrow, no. But every, it's going to be not one step, not two, it's not even gonna be 10. It's gonna be like a hundred different steps of a hundred different (37:00) people trying to push this forward and get something through.
And I think that's the attitude that I have towards this, which is just pick a place and start. It doesn't really matter where. It's you come back to a house and it's messy and cluttered and go ' oh, where should I start cleaning?'
But actually, the most effective thing to do is just pick up and start. And as you pick up and start, you will see what works, what doesn't work, and you will learn a hundred things which don't work before you find something, which does.
You will see a bunch of projects filled. But this is getting some traction. You'll put some weight on there. Then you will build those skills and you go and you get somewhere. And if that sounds like a lot of work to you and you'd rather lay down and die, okay. But for the people who don't, this is what you do.
Yeah. Bit by bit.
Jeremy Au: And wrapping things up, when have you seen yourself been brave? Could you share a personal story about a time that you've been brave?
Li Hongyi: I don't know whether I would consider myself brave for any of these things, because truth be told, like compared to a lot of other people in the world, our lives here, especially in Singapore, are very safe and very sheltered. And nothing you do, whether it is, arguing with a colleague or going is particularly brave. And (38:00) I think that to me is my reference point. Now, the way I describe it is that you look at the Ukrainians, for example. For them, they are literally deciding that
we're gonna do this. They could very easily at a hundred points were just like, fuck it our country doesn't anymore, but they literally fight for their lives while going to work. If you are in a lot of other countries in the world fighting against corruption and government inefficiency, literally gets you killed.
Literally. They don't argue with you. They don't dock your career. They don't give a bad remark in performance or your thing. They literally, shoot you and kill you, sometimes, worse. So in Singapore, I think it's very presumptuous for me to call anything I've done particularly brave.
What I try to say is, if other people elsewhere in the world can deal with so much risk and so much harm to make their country just a little bit better surely in Singapore where your risk is like, someone like saying something snide about you, surely you can do the same. And that's the way I think about it.
I don't think anything I do is particularly brave, but at least on a day-to-day basis on these things, when I started my team, I think it would've been very easy for me to say, ah, this is really hot. It's really (39:00) frustrating, really stupid. And I think there were times where I wanted to do that and just say, screw it.
I'll go back to California. You can drive around in the California sunshine, it's really nice. But if other people in the world can deal with so much to push for things and we have to deal with so little,
I think we should do it. And I try to make that choice every time.
Jeremy Au: Amazing. Thank you so much for sharing and sharing on a podcast. It's good seeing you, Jeremy. I love to summarize the treat takeaways. First of all, thanks so much for sharing about
your childhood dreams of being a physicist and how it's shaped up in terms of economics, in terms of computer science, in terms of who you are today. Secondly, thanks so much for sharing about, your reflections on who were your good managers and how you've improved as a professional over time and how you've become a manager, and how you see that coming together as your own philosophy.
And lastly, thanks so much for sharing. I think, people who are thinking about joining government or working in public service or working in industries that have a social impact element. How to think about the world, the place of Singapore, and how people can get started on a tiny corner of their house, as you said.
Thank you so much
Li Hongyi: it was good seeing you, (40:00) Jeremy.