Shiyan Koh, Managing Partner of Hustle Fund, and Jeremy Au discussed:
1. Trump’s Economic Policies, Tariffs & Crypto Initiatives: They examined the economic impact of Trump’s 2025 return, including a 10% tariff hike on Chinese imports and new tariffs on Canada and Mexico. The administration also ended the de minimis exemption for low-cost e-commerce imports, affecting platforms like Temu and Shein, which had relied on duty-free shipping to US consumers. These changes disproportionately impacted lower-income Americans, despite cost-of-living concerns being a key election issue. Canada and Mexico secured a 30-day tariff delay, while Trump also launched Trump Coin and proposed a US Bitcoin reserve, signaling a pro-crypto stance that could draw US crypto firms back onshore.
2. DeepSeek & US-China AI Dynamics: They discussed the launch of DeepSeek-V3, a Chinese AI model matching GPT-4 but with lower training costs, which Jeremy called a “Sputnik moment”. The model’s success exposed the limits of US chip export bans, as Chinese engineers developed efficient AI training methods despite NVIDIA H100 restrictions. DeepSeek’s open-source availability via Hugging Face complicated regulatory enforcement, leading US Senator Josh Hawley to propose severe penalties, including 20-year prison terms for users and $100M fines for corporations. Meta’s AI lead, Yann LeCun, framed the issue as a debate between open-source and closed-source AI rather than a purely U.S.-China rivalry.
3. Grab-GoTo Potential Merger: They revisited ongoing Grab-GoTo (Gojek) merger talks, noting that Grab’s stronger financial position made it the likely acquirer. While Singapore’s regulators were expected to approve the deal, Indonesian authorities might impose conditions such as fare caps or job guarantees to prevent monopolistic practices. Reduced competition could push ride-hailing fares higher, with some Singaporeans already shifting back to public transport as Grab’s peak-hour prices reached $40. SoftBank, a major investor in both companies, had long pushed for consolidation, and with Gojek’s founding team no longer involved, negotiations had become more financially driven.
Jeremy and Shiyan also discussed Waymo’s self-driving taxis and their potential impact in Southeast Asia, Singapore’s emphasis on “future-proofing” careers versus the US culture of embracing disruption, and how US trade and AI restrictions are accelerating Chinese firms’ shift towards Southeast Asia and the EU.
Jeremy Au: Hey Shiyan, how's [00:01:00] life? Happy New Year, Jeremy. Oh, the Lunar New Year. I don't think Chinese New Year is no longer in vogue, I think, for 2025.
Shiyan Koh: None of your PC stuff, okay? We're Chinese people, we can't say Chinese New Year.
Jeremy Au: Thank you for the red packets for my kids. It was fun to receive them. And now I have to get into the process where I'm like, got to figure out to open it up, see how much, for equivalent amounts. I don't know. That's the classic parents stuff. So we want to talk about this year for 2025 has been a huge start, right?
When we did our last recording a month ago, it was pretty much like the December timeline and I hope that was just for the year. And we talked about the eFishery as well. Yeah. And now it seems, "Whoa"
Shiyan Koh: I guess Trump came into office and then things just happened really fast.
Jeremy Au: So I think when we predicted that there was this acceleration, I frankly found that it accelerated faster than I thought, right?
Because I felt like when I thought acceleration, I don't know. You wouldn't have been a week one, right? But it hit it all in day one, actually.
Shiyan Koh: I mean, they, I feel like they came in [00:02:00] with a set of things they wanted to get done and they just went boom. I have to say taking over the, what is it?
Office of Personal Management and then the Treasury Payment System, it makes sense, right?
Jeremy Au: Yeah. Are you thinking over the HR and the finance of any company you're effectively lifeblood, right? Because HR, you handle all people and finance, you handle all my payments. So everything runs on those two things, right?
Yeah. So I think the velocity is much higher because during this timeframe, I would say on the economic front, like he proposed tariffs on Canada and Mexico, which obviously is huge impact on the manufacturing base. And obviously, He also increased tariffs on China by effectively 50 percent because of 10 percent tariff on top of an existing 20 percent tariff.
Yeah. That went up even further. Yep. He closed the e commerce loophole or exemption, depending on how you look at it, for e commerce goods of China going to low income and middle class Americans. Yeah. So a lot of things have happened really fast.
Shiyan Koh: Yeah. It is a little bit odd, right?
Because one of the big reasons for people voting for Trump was, [00:03:00] cost of living and prices. And I read an analysis that said, Temu and Shein. The postcodes that they send to are actually disproportionately low and middle income, which makes sense, right? Because they're cheap. And I think he's actually turning the screws on people who are suffering most from cost increases.
Yeah. But he came out and was like, some pain is going to happen, but I'm not really sure that the pain of that is actually going to lead to more manufacturing in the US
Jeremy Au: Yeah, I think it's going to be an interesting debate that happens. For sure. I think we all know the impact, like you said about the goods and so forth. If you're going to Costco and you buy maple syrup, a 20% tariff will increase.
Although they have a
Shiyan Koh: 30 day reprieve.
Yeah, they did.
So they were supposed to go into effect last Thursday. Yeah. And both Canada and Mexico were able to. Get him to hold off for 30 days by essentially promising to do things that they were already doing.
Jeremy Au: Yeah, and I think that's where [00:04:00] the markets are trying to figure out, right? Is this part of a negotiating ploy? Or is this actually a systematic increase of tariff and tax structures, right? And I think the market is like swinging between those two points of view right now.
Because if it's a structural tax increase by tariffs, obviously it does have a very strong impact on the value flows between importers the American economy, right? Versus if it's a negotiating ploy, then, fine. Then it doesn't structurally change the economics.
Shiyan Koh: Yeah. But, I think if you look at the difference between the first Trump administration and today, right?
The trade surplus with China has gone down, but the trade deficit with Mexico and Canada has actually gone up, right? So it's going to be an even bigger impact on the U S economy.
Jeremy Au: Yeah. I think it's going to be interesting to see how it plays out. And I think just today they announced that Donald Trump was saying, perhaps there can be terrorists on Japan, as well.
So I think basically the entire world is basically going, line by line. And I think EU is waiting for the shooter to drop as well.
Shiyan Koh: Yeah.
Jeremy Au: Yeah. So what does that mean about technology?
Shiyan Koh: I think the question on whether some of these [00:05:00] restrictions have worked or not has been blown up by the deep sea canoes, right?
So the idea was like, Hey, we're gonna stop China from getting access to the best Nvidia chips and halt their progress. And then, the news basically of DeepSeek's V3 reasoning model coming out, and matching state of the art at reportedly a much lower training cost, collective trillion dollar freak out on the markets, right? I don't know, what does it mean for technology? I feel like the trade war tension is only going to get ratcheted up, right? Separate from tariffs, what other restrictions are going to be put in place? And, yeah, the climate is just not, I don't know, man. The whole thing feels very insane.
Jeremy Au: It's definitely not an easy set because there's so many moving pieces that's happening, right? So if I was to go through that list, like for example, I felt like DeepSeek felt like a Sputnik moment in that sense.
Like there used to be like the space race between America and Russia, Soviet Union, and then the Russians launched Sputnik, the satellite. Everybody in America panicked. And then after that, there's a [00:06:00] big push to put a lot of money and then put a man on the moon.
Shiyan Koh: But I don't think the money is an issue, right? And I think there's an open question about how big a deal it is, right? Because the one thing is that the DeepSeek team actually published. It's an open source model, like you can go download it today on Hugging Face, right?
Perplexity has already included it as an option in their search engine. And the techniques that they describe, I think everyone will use from now on to reduce the cost of pre-training. Is it so surprising that a bunch of smart Chinese engineers and scientists could come up with ways to work around the constraints of lower powered NVIDIA chips.
That's not actually that crazy an idea, right? You have a billion people. Amongst those billion people of whom many are very highly trained in STEM, would they not solve around that? And actually the fact that they didn't have access to H100s actually force them to innovate around CUDA in order to get the kind of processing that they needed to do the calculations. And the fact that Americans have no problem getting the [00:07:00] highest rated chips means that they weren't forced to do that, right? They didn't have to go and find that extra few turns of performance from somewhere.
I mean, I don't know if it's a Sputnik moment because also those techniques will be copied by everyone, right? It's open source. And you can't just imagine that everyone else is dumb.
Jeremy Au: Yeah.
Shiyan Koh: No, No, but it's a little bit like, did we talk about this last time? About the Americans getting on Xiaohongshu and then seeing Chinese people and being like, Oh, wow, like your city looks pretty modern.
Did you think we lived in a cave? What do you think's going on over here? I don't think the valley has a monopoly on innovation or brains or ambition. And of course, things are going to happen all over the world. And it's the constraints perhaps led to this.
Jeremy Au: No, I love what you said, which is, from an American, if you look at it as an arms race in AI, which is the framing in the media today, they felt like having those export controls AI chips would prevent them from catching up, but like you said, actually the evolutionary Darwinian approach is that if you created that constraint, people will design around that [00:08:00] constraint.
Shiyan Koh: Yeah.
Jeremy Au: Which is that, if America had let China use the most power hungry chips, then probably, we will probably see the latest iteration be a more powerful model, but probably be much more power intensive.
Shiyan Koh: So yeah. But it's if you, who is it? It's like the Indian Space Agency, right? The Indian Space Agency actually is very well known for being very capital efficient, right? And building rockets that are cheaper and all this, it's like, well, why? It's like, well, they just don't have as much money, so they need to, right?
So I do think that there is some aspect of that, but it's like a mad thing. But you were telling me about legislation banning DeepSeek today. And at first, I thought you were just talking about the federal one, where it's oh, we're gonna ban it on federal devices. You're like, yeah, okay, fine. Sure, right? Anyone who uses any sort of corporate device half the things you're not allowed to use anyway, right? Great. Somebody has some sort of posturing bill to go, ban it on federal devices. Why not? But, I think there's actually a second piece of legislation which I hadn't read about yet.
Jeremy Au: Yeah, let me go get the numbers for you.
Shiyan Koh: Cause [00:09:00] that one sounded much more aggressive.
Jeremy Au: Yeah. I think that Senator Josh Hawley proposed that if you use DeepSeek as an individual, you can have a penalty of up to 20 years in prison. And if a corporation uses DeepSeek, then the fine will be up to $100 million for a corporation.
And I think the big problem, as you can tell, is that we don't know whether it'll become legislation or not, but definitely it's within the Republican Party, it's a good shot at passing. It could potentially become bipartisan. And the potential nature of that does have a chilling effect, that if I was an American company, I would be like, do I really want to build my technical roadmap on top of DeepSeek, right?
Because if I use it, the technical infrastructure for it, I might have to re-engineer in one year's time when it does get passed.
Shiyan Koh: Yeah, sure. But I feel like talking to people in their portfolio like most people are actually building in a way that allows them to swap out the underlying models pretty easily.
And they do that anyway, right? They start with the frontier models, whether from Claude, from Anthropic [00:10:00] or OpenAI. And then as they learn more about how the application performs, they start to swap out with cheaper open source models, right?
And they divide up the workload so that they're not spending so much on the frontier models and they slowly swap smaller models or cheaper models through. I don't know, that feels just, I agree. On the margin, probably, it makes you be like, do I really need this hassle in my life? But it's just like more posturing, just so irritating.
Jeremy Au: Yeah. If you are a startup, you would be pushed to make it even more likely to be able to swappable, right? Because you just need that optionality. Which is, probably one of the intended effects of even having the legislation come up is this. It's not a side effect, it is the one of the main early effects of it. So I think the direction of travel is clear, although I think I also like what you said was true, which is that even though it can be seen as Chinese, but it's also open source, and that open source, it was hosted on American open source platform and community.
Shiyan Koh: When Perplexity rolled [00:11:00] it out, the marketing message was, this is the DeepSeek model hosted on American servers with censorship controls removed.
Jeremy Au: Yeah.
Shiyan Koh: So what are we even talking about?
Jeremy Au: Yeah, because it's open source, right? It's like really also a victory of the American netizen, internet principles of open source, right? Linux and, other folks.
Shiyan Koh: Meta's
AI leader, Yann LeCun was like, this isn't about China versus India, versus the US. It's actually about open source versus closed source. So he, of course, because, Meta has been pushing open source with their Lambda models and things like that. He's like, yeah look, we can see everything and everyone's going to take this stuff and they're all going to improve and innovate with their models that way too.
Jeremy Au: I love what you say, which is the idea of open source versus closed source, right? And I think the virtue of closed source is that somebody gets to control it.
Therefore, the country can control it, right? It exists. If you have open source, then no country can control it by nature, which is the whole ethos of open source as well. It's just that, open source used to be, I think it still is obviously American led, American [00:12:00] values, libertarian American, like push, right?
No? It's like me using Microsoft and then sometimes I was using like OpenOffice.
Shiyan Koh: I'm trying to think whether it is an American idea. Maybe I feel like we have to do some research on what is the origins, I feel like.
Jeremy Au: Internet originated in America the early folks had a very yeah.
Shiyan Koh: But it was a DoD. The internet was a DoD project, right?
Jeremy Au: Upper net, but within that space it was an open playground. The open internet net neutrality a lot of that stuff is were we really debating this in Southeast Asia?
No, definitely not. Has China been a big advocate of open source? No, historically, no.
EU, I think it's definitely following America's footsteps on this stuff. I think if I define America as a country, American as a governance and regulatory agency, America in terms of communities and, like scientific community, and then also there's American values, right? Which is the culture and ethos, right? Four very distinct layers, I would say, where we say something is American.
Shiyan Koh: And America is a big country, right? There's a fair amount of heterogeneity. All right. I'll [00:13:00] consider it. I'll allow it.
Jeremy Au: But, I think what's interesting is that, yeah, like you said, DeepSeek is, open source. And so it's on open source servers. Everybody in America has access to it. A lot of startups are using it, DeepSeek canal. And then the economic market went to a downward turn because they digested the news, right? Which was like, one is we need potentially less GPUs than we thought.
That's one.
Shiyan Koh: Two is we need less energy than we thought because it's more energy efficient. And both of those things is NVIDIA and obviously American energy companies.
Yeah, NVIDIA and the Magnificent Seven in general are like price to perfection, right? They're trading at like historically high multiples and a bunch of that was around like AI. And in general though, when the cost of something goes down, there's a name for this. Jevons paradox or whatever. Yeah, that's right. It's actually it grows demand for the thing rather than saying, so like, you could debate about whether NVIDIA is fairly valued or not. But in terms of like the demand for the chips, I don't think that actually goes away. So it's like item one.
Shiyan Koh: I actually think if the cost of intelligence goes down, [00:14:00] it just means that you can apply to way more things. Yeah. Economically. And the demand for that thing goes up.
Jeremy Au: Yeah.
Shiyan Koh: And then I think the second thing is I've read a few, more technical papers about the cost, right?
Because there's a big debate about can we believe this cost estimate? Is it true? And, they say this is the pre-training number. It doesn't include, like, all the R&D that was done before, or whatever. And so if you think about pre-training as a percent of the total overall cost, it is not the 100 percent cost.
So there's still like the cost of inference and all these other things. And that still is gonna require GPUs. And that next phase of development I think is gonna become more important as models advance. So I just think on a demand side, from the demand side, I'm not actually that worried about demand for GPUs.
I do think it begs the question of, okay, because I think all the tech companies reported in the last 10 days or so, did they need to be spending that much on CapEx? Could they use more effective ways to do it? Probably, right? If they probably start to move some of these techniques through to the stuff that they're [00:15:00] building, maybe they'll save money that way.
But they're all kind of in an arms race with each other, right? So it's also just that we've just pushed the frontier of what you can do here. We're going to implement that, but that doesn't mean that there isn't other stuff for us to keep trying to push the boundaries on and spending on.
Jeremy Au: Yeah. And I think this is actually a good parallel to the US broadband push, right? I think there was also a similar kind of like push where everybody was bidding broadband and they were pushing really aggressively, huge overinvestment, but it was an arms race.
Shiyan Koh: Yeah.
Jeremy Au: And the winners were the ones who did overinvest.
And didn't die.
So the tricky part is if you overinvested the most and you didn't die,
Shiyan Koh: you can last long enough for the demand to catch up.
Jeremy Au: Those who overinvested and died got killed, and those who underinvested and survived end up dying and eaten anyway. So it was like the sweet spot Goldilocks of like, how do you overinvest but also raise enough capital so that we don't die? And then we eat everybody else and consolidate that.
Shiyan Koh: And maybe overinvest is not the right word but you're basically investing ahead of demand.
Jeremy Au: Yeah. So I don't think, overinvest implies a negative thing, but you are investing way ahead.
Shiyan Koh: It's a time [00:16:00] problem.
Jeremy Au: It's like railroads, right? Like once upon a time, all the railroad barons were like, let's over invest in railroads, right? They were investing ahead of demand and then they built all that track.
And if you look at today, everybody's like, America, there could be more railways, but it just happens to over invest at that point of time versus what they had. It was interesting that dynamic for infrastructure. And I think it's key, which is like what you said about it.
It's AI is a commodity, right? And the more and cheaper we get it, the more people are going to use it. So maybe you're economical to use like, we talked about AI teddy bears, right? In the past, the compute wouldn't have made it because you'd be like, oh, the subscription cost of this AI teddy bear will be $10 per month. It's too expensive. So just AI teddy bear. But now you're like, oh, with DeepSeek and the way it is done, you can imagine a scenario where you're like, it's preloaded with enough whatever and it's a one time cost. It's $10 one time cost, right? Into the teddy bear.
Shiyan Koh: Yeah, I think you'll just end up applying it to like lots of stuff, right?
You have a talking tree, you can have a talking chair. I'm just saying you have a talking hat. I think when you talk about cheap, that's what we're getting to it.
Yeah. I have a very silly example. When you go to Sushiro, [00:17:00] the conveyor belt sushi, and every plate is chipped.
Shiyan Koh: But do you remember when RFID chips first came out? They were very expensive. And then they're like how would you ever use this thing? It's so expensive.
Jeremy Au: Yeah. I remember all those crazy founders are talking about RFID chips at the future. And then everyone's like, why would we use RFID?
Shiyan Koh: But now it's they're in every plate at Sushiro. And at the end of your meal, the person just comes in scan your stack of plates and gives you your bill.
Jeremy Au: Crazy, right? So they were right.
Shiyan Koh: But it's a trivial example.
Jeremy Au: No, but the thing is like the founders that were doing RFID, I remember now from those days, they were right.
It's just that, of course, the question is what kind of, that was too early, but too is what's the business model, right? Can you imagine their business model was like every sushi plate will have RFID. Then everyone would just laugh at them and say like, why are you talking about? It's too cheap, such a low value, blah, blah, blah.
But if you think about it, somebody made money selling RFID sushi plates.
Shiyan Koh: Yeah, it's just but it's all labor savings, right? That's what it is. But I think every book in the National Library is RFID tagged.
Jeremy Au: Most trees [00:18:00] in Singapore are also tagged as well.
Shiyan Koh: Yeah, because you know when you return the library books, right? It's just boop, and it just knows what you have. I don't know why I suddenly thought about Sushiro. It's the weekend, I have to deal with my kids, but
Jeremy Au: No, we're just talking about the fact that, the cost of compute is dropping, and it's getting to the point now that it's going to be a big thing cause and, it reminds me all of a sudden, of that, Beauty and the Beastbeast where the teacups, every teacup has a voice and personality.
They all dance at the candlestick. The clock, you're like, wait a moment. We can totally do that now.
I'm just saying is this cheap enough where you're like,
Shiyan Koh: Just because you can do something doesn't mean you should.
Jeremy Au: Wouldn't you want to like, have a tea set and coffee set and all them all talking to each other and having a very dramatic korean drama style debate with one another about who's, getting the coffee and who's pouring the tea. If you think about it, we now, because of DeepSeek and the AI boom, you're like, wait a moment, it probably cost you like, maybe like a hundred bucks to do it, right?
Because this is like a small speaker for each one of them. And then you put a wifi signal and then you run it through DeepSeek and you'd be like, Oh, I recreated Beauty and [00:19:00] Beast,
Shiyan Koh: My house is noisy enough. I don't need my teacups and my pot to talk to me or to each other.
Jeremy Au: You're going to be disrupted by the future. You're like bifocals. And you're like looking at some print newspaper.
Shiyan Koh: Hey, don't make fun of me, okay. I'm probably about at that level, I need to get bifocals. I was like, oh man. Presbyopia is setting in, like, why can't I read this small text anymore?
Jeremy Au: Now imagine at your breakfast table, you have your seven teacups and every place talking to each other and while they're talking to each other, they also read the news to you. See, there you go. And you know what? This is gonna become a product, actually. Like a Beauty and the Beast talking tea set.
Shiyan Koh: I'm sure disney's already working on it.
Jeremy Au: And it's doing, you know what? It's totally doable, right? Anyway, such a novelty gift, can you imagine? But yeah, it's possible now. It's just that, the AI model will be proudly made in America. That's the difference for the Disney one, right? You know what I mean?
It'll keep the censorship and the fantasy modules installed, sorry.
Shiyan Koh: You are really into this teapot thing.
Jeremy Au: I know, I'm just like, [00:20:00] okay, I think somebody's going to make it, so why not?
Shiyan Koh: For sure. And I think somebody is working on the AI stuffed animal stuff. Yeah, there's a lot of them just working on stuffed animals. It's totally a thing.
Jeremy Au: So now we're just brainstorming so that when it does happen, we'll be like, Oh, we already thought about that. We're like, just too lazy to do it ourselves and make bank on it. We just speculated for fun, right? I guess another question that I was thinking about is Okay, so what does that mean, right? Because we saw the acceleration of geopolitics, I think, in January, since the January 6th inauguration. So I'm just saying like the piece that I'm thinking about is I think one piece is that I think the anti China or the decoupling will accelerate even faster than I had thought even one month ago, right?
And I think the writing of the law is quite clear for every Chinese national founder out there.
Shiyan Koh: Yeah, I don't know whether it's going to be faster than I thought. I think the writing on the wall was pretty clear, right? Because, it started in the first Trump administration and then continued. Biden's policies continued there. I do think that that is something that Chinese founders have to continue [00:21:00] to think about.
Because I guess the other side to take on this Trump stuff is like, all of this stuff he's doing by executive order rather than through legislation. He doesn't want to make things more permanent or pass legislation. He doesn't want to deal with Congress. He just wants to be like, do this. And he's like a mercurial person. I think there's a little bit of let's wait and see what happens.
Like it hasn't actually written into law yet.
Yeah. I don't know.
Jeremy Au: I just feel like, whatever impact it has, four years is a long time. It's forever in startup years. There's nothing you can do. Actually, one interesting thing that I've talked to some Chinese founders about this is that for the smaller Chinese founders, they're actually okay with some of these changes because the large Chinese players like Temu and Shein, like all those companies were doing well under the old regime or if you want to call it by the old set of policies.
And so to some extent, Trump's changes rewards, for example, Chinese founders are willing to move to Mexico or smaller outfits are willing to be more nimble because a lot of the pain, for [00:22:00] example, on closing the minimis bracket for e commerce goods impacts Temu input Shein, for smaller e commerce, somebody else's problems at the top is at least a way to change the game for my current dynamics.
Shiyan Koh: Yeah. Oh, that's an interesting point of view, which is it's all aimed at incumbents. And so if you're a small startup, you benefit because your big Goliath is occupied.
Jeremy Au: I think another thing that happened that I didn't expect as much as personally was that I think Crypto made a big comeback.
Shiyan Koh: Heading into the next bull market. You ready?
Jeremy Au: I think we saw that Trump actually went towards pro crypto perhaps in about Q3 last year. Because individuals, but people felt like it was primarily we're pushing back against Biden
and Kamala for being anti crypto rather than Trump being pro crypto.
And I think in January, we saw Trump launched the Trump coin. And then, he promised to be the most pro crypto administration.
Shiyan Koh: There's also this like strategic Bitcoin reserve proposal. Did you see that one?
Jeremy Au: Yeah. If the government is [00:23:00] buying, then obviously, it's like an oil market, the prices will rise, it's, there's a lot of support, supposedly right for the lift.
Shiyan Koh: Yeah. I mean, I think it's part of the general like deregulation move, right? Which is we want to be the center of financial markets, right? So we're going to like, and there's people in the administration who are pro crypto. He has a crypto czar, right? David Sachs. And, but even before, right?
You can already see some of this stuff opening up with the ETFs being approved and things like that. I think, yeah, like I think the mood is good. But I think partially also the crypto market has matured a little bit, right? With the many more sort of the stablecoin use cases and things like that.
But yeah, the Trump coin actually seems quite ridiculous to me. Can you imagine any other political leader, like on the eve of like, basically coming into power?
Jeremy Au: Yeah. Lawrence Wong coin. There you go.
Shiyan Koh: But he's the beneficiary of it, right? Anybody who buys that thing, and the company that did it, is based out of Singapore.
Jeremy Au: Oh, really?
Shiyan Koh: And but I think that is actually quite you don't want to talk about conflicts of interest, like launching a coin for [00:24:00] your own. And Melania has a coin. The whole thing is just insane.
Jeremy Au: Yeah. And I think it's interesting because I think it does have an impact on Southeast Asia crypto from my perspective.
So I think that when the Biden administration was more anti crypto from the perspective of the crypto community, and obviously China banned crypto as well, then Singapore was seen as a neutral place as well as Dubai.
Shiyan Koh: Yeah.
Jeremy Au: And I think what's interesting is that I think now that Trump is pro crypto, I think there's going to be a move.
From my perspective of the more American affiliated crypto companies, I think back to America, perhaps, because, they might domicile, for example, in Texas or Nevada rather than Delaware I think there's a certain stack where you see them and they still have some optionality, obviously, between, different geographies, just in case the next administration is a democratic administration that's anti crypto again, but I just think that there's going to be that move of crypto talent there.
And I think, the Chinese talent will still continue to be in Singapore and Dubai because I don't think they feel too welcome in the US But you can also imagine regional crypto folks [00:25:00] might just move to the US as well rather than move to Singapore. So I think Singapore's crypto hub that really existed as a regulatory alternative, alongside Dubai.
I think both of these alternatives have decreased in terms of. Value to a crypto business.
Shiyan Koh: Yeah, but it takes time, right? I think part of it was that there wasn't a lot like it was like regulatory confusion, right? It was like, is this legal? Is this not legal? What's going on? So it'll take time to actually put in, I think, like clear regulatory guidance.
Yeah. But yeah, you're right. Like probably on margin, right?
Jeremy Au: Yeah, definitely. I think what's going to be interesting as well in the coming year is I think that a lot of the Chinese companies are going to start really aggressively pushing Southeast Asia, EU as growth markets rather than the US I think a lot of the Chinese folks were like, they think that Trump is, obviously not friendly to China, but they've had a wait and see approach. And then Trump really came out strong and raised tariffs by 50%. So from their perspective. So I think everyone's just okay, got it.
The message is we better make it work in EU or [00:26:00] Southeast Asia.
Shiyan Koh: Yeah, it's I go back to the US end of next month and so I'll want to see what the temperature is.
Jeremy Au: Wrapping up things here from your perspective is what's the one thing that you hope happens in Southeast Asia Tech in the next month?
Obviously we're looking at the whole region, and I think the big piece of news as well for the Southeast Asia tech ecosystem is that the Grab and Gojek slash GoTo merger acquisitions talks apparently have been leaked to be said they're ongoing. We have a target for the deal to be done this year.
Shiyan Koh: Yeah.
Jeremy Au: What do you think?
Shiyan Koh: This isn't a new rumor, right? There've been talks about this a number of times in the past. And so I think there is some commercial logic to it, right? Is to consolidate these players remove a fair amount of the cost structure because they compete in the same markets, but they are strong in different markets, right?
And so to be able to eliminate one layer of competition, I guess my question is is it anti competitive?
Jeremy Au: Yeah. And I think that's going to be the question I'm sure that the regulator is going to have between the two companies, Grab is a stronger position, enterprise [00:27:00] value in terms of its core profitability, whereas Gojek is still losing money.
So frankly, this is more of an acquisition than a true merger of equals. It's acquisition of Grab of Gojek. But so the Singapore regulator is going to be okay from my perspective and from my experience doing M&A. And I think the Indonesian regulator is going to be the question mark about that. But I think that's where I think some of the optics gonna be very key.
Like saying there's a merger rather than acquisition. I think the optics around maybe some of the commitments to preserve jobs in Indonesia would be an important part to allay the Indonesia regulator. Maybe as we're talking about service levels or even price levels it's gonna be quite important, right?
Because, if you look at where Grab and Gojek are competing, they're primarily competing in the Indonesia market. And both of them therefore are keeping prices relatively low. But if Grab and Gojek kind of merge as a service, then as a primary player, the prices can rise. And then I think from a regulator perspective, that would be quite problematic from their perspective, right?
So there may be, I [00:28:00] can imagine some level of commitment to be like, we're not going to change prices maybe for the next, five years.
Shiyan Koh: I mean, it's not like they're the only people, right? There's still taxis, Bluebird is a big player. There's a bunch of other yeah, they're just the big like ride hailing.
Jeremy Au: Yeah.
Shiyan Koh: And I think there's a couple of Vietnamese players that enter that market as well, right? They're trying to go regional
Jeremy Au: Yeah, so you can imagine a scenario where you know, and I've no access to any of this private information, but I can imagine that Gojek will also be fielding opportunities talking to Bluebird and these Vietnamese companies just be like, does that level of integration make sense as alternative because you never want to be an M&A composition of only one party, but of course, who's going to have the money to pay, and streamline? But like you said, the economic part of it is you have two major ride hailing companies that are highly duplicative in terms of their marketing and their infrastructure stack.
They can save a lot of money just by consolidating, so
Shiyan Koh: But if they have to make concessions on the Indonesian jobs and all that sort of stuff, then, does the math still pencil out?
Jeremy Au: I think it pencils out from the Grab perspective, right? Because, it's just a 10 year horizon, [00:29:00] and then you continue being a super app, you make money on other things as well. Because I think the transportation side is the most sensitive bit but if you raise prices on food delivery, or some other adjacent parcel delivery, I think people are not going to be as, politically sensitive about that point.
And of course I think driver salaries is going to be a big part of it as well. So I think it's going to be interesting because these rumors have been happening for a long time, right? The truth is SoftBank has been pushing both of them for a long time to do SoftBank had investments in Uber, Gojek, in Grab. And they were the ones who told Grab and Uber has said, Hey, Uber in Singapore because both of you are going to a price war to show a lot of value as a VC in both of them. So Uber should exit so that Grab can go up.
Obviously, what Uber's management thought about this is a different thing. But from SoftBank's perspective, it's less competition makes it easier to rise. At that time, they said no to SoftBank because the two CEOs could not agree. Who wants to be CEO?
But now I think it's totally different,
Shiyan Koh: Cause the founders are out.
Jeremy Au: For Gojek is out. So right now it's a Patrick from [00:30:00] North Star who's committed to be leading Gojek to 2029, but of course, when you say it's only there till 2029, then obviously the question is like what happens in 2030?
And so I think some of this timeline is less emotional. Is this going to be more like there's synergy, there's cost savings, there's revenue opportunity. This is the value. How do we slice the pie? And I think the issue that we have is that Gojek doesn't have as many alternatives because Greg can keep going.
Shiyan Koh: Yeah.
Jeremy Au: But Gojek is going to find it more difficult. So I think to me, I think that's where the struggle is from the Gojek perspective is how do I get the best transaction price out of this relatively like single buyer auction, which is not.
Shiyan Koh: Yeah. And they have the same color. So it should be easy. Brand assets perspective.
Jeremy Au: Your cost saving, is this going to put white tape for your back? We're not even going to change the color. That's how cheap you're going to be. Elon Musk style is like, we're not going to buy new swag with this. We're just going to use a crayon and just draw on the back.
Yeah, no, I think it's going to be a big piece. So what would the impact be? I think Grab is the bigger player of the high enterprise value is acquirer. So [00:31:00] the question is, will they overpay? I think that's a sub market perspective. So we don't know that, is less competition in the Indonesia side.
So we should expect that over the medium to long term, prices will rise for indonesian consumers or you can also make an argument that driver salaries will stay stagnant in the Indonesia market. Gojek already exited from Vietnam. The sort of Vietnamese consumer has already been yeah, no more impact from this piece So it's really I think the indonesia consumer does be impacted maybe
Shiyan Koh: Singapore, they don't have much here.
Jeremy Au: Singapore is not much. It's very small piece as well. So it's primarily,
I see some Singaporean people, they check between Gojek and Grab and Comfort Delgro. But I think now that Gojek's out, this is going to be back to Comfort Delgro and Grab. And Grab also pulls taxis from Comfort Delgro.
So they're both like same fleet, different apps with different margin structures. Anyway, I don't, so I think you can expect maybe some small rise maybe in Singapore prices.
Shiyan Koh: They've already gone up, right? Yeah, they've already gone up a lot. I don't take Grab that much anymore.
Jeremy Au: Yeah. I definitely been a few times where I was like I'd rather [00:32:00] drive and listen to something or sometimes I'm just like, you know what? I'd rather take the bus.
Shiyan Koh: Especially peak period. Like yesterday I was, going from one meeting to dinner and it was both downtown, right?
I was like, you know what? I'm just going to take the MRT and walk because it's going to be super expensive and very crowded.
Jeremy Au: Yeah. You're going to save a 40 minute trip to 20 minute trip. And then your price difference is going to be effectively 40 bucks versus a dollar on the train.
Shiyan Koh: And I'm going to get in 2000 steps or whatever.
End of the day. Get a bit of walking in. Not so bad.
Jeremy Au: Peak of hours. Like it's basically $2 per minute. Saved from my perspective, which is a crazy, that's $120 per hour salary equivalent. So you better be doing something massively more productive.
Shiyan Koh: But it is, it's like, it was like gotten really expensive for a peak period.
So I was like, all right.
Jeremy Au: But when I was in SF in January as well, Uber is also more expensive now as well. So I think this is a function of labor, right? It's if a lot of cheap labor, way cheaper. Yeah, so exactly.
Shiyan Koh: But Waymo is not that much cheaper. It's a little bit cheaper. No, But you get novelty [00:33:00] value.
Jeremy Au: Waymo strategy is price skimming, right? Similar to like the lab or cultured diamond market, right? If you have natural diamonds that are very expensive, and you have a lab diamond, would you price it as very cheap? No. You're going to price it as 5 to 10 percent cheaper than a normal diamond.
And then people start switching because there's enough for price difference to get a better stone. And then the next step is after that you drop it by another 10%. And so the price drop happens, a price war happens. People are trying to skim that curve. So it takes 10 years to really see that full price drop.
So that, during this time period, the lab diamonds can extract as much value as they can. Does it make sense? Skimming off the natural diamonds. So the natural diamonds nests never move. Only lab diamonds are moving, but at least you have the price premium. So I think same thing for Waymo is they have no need to set the prices massively cheaper.
As long as it's slightly 10 percent cheaper than Uber, then people switch. That's what we saw in Grab and Gojek. It's just if you're 10 percent cheaper, people just switch anyway. So you don't need to go like a massively cheaper. But yes, eventually, I think [00:34:00] self driving cars would be a way to get it really cheap.
Shiyan Koh: Don't want it because once again, then it leads to overconsumption.
Jeremy Au: Yeah. And also if you make it too cheap, then governments will be very angry because a lot of these drivers are low income folks. So you're disrupting the taxi fleet, right? Which is still a relatively protected class in many countries.
So you know, and then, I think that's what happened in Chinatown and SF, right? Like they burned a way more robo taxi. Instead of if I had hit it, it was a bunch of rods.
Shiyan Koh: But if you think that's the future, then maybe you shouldn't give out new taxi licenses.
Jeremy Au: As a government?
Shiyan Koh: Yeah. Cause it's like older people, basically. So it's like employment of last resort. And so maybe it's okay, like you just let that population nationally a trip out, but don't enable new people. And then you slowly implement the self driving cars.
Jeremy Au: Yeah. And I see that argument internally. I'm sure people are arguing about that internally.
Of course, from my perspective is, we also don't know how long the robo taxi curve is going to take.
Shiyan Koh: They work there and they're licensing it, right? Waymo's licensing it to [00:35:00] Uber in Austin. You can start doing it today, basically. And I would make the argument, I don't know though, but there's like young people who are Grab drivers.
And I guess there's a question which is, is that the best most productive job they could be having.
Jeremy Au: This is where Singapore government for skills future to be future proof your career. I always love the phrase "future proof". It's such a Singaporean phrase like, like making yourself invulnerable against the future.
I always found it such an interesting phrase that only exists in
Shiyan Koh: Singapore. Is it only Singaporeans use it?
Jeremy Au: Ever heard a SF person say, I'm going to future proof my career? No. No, right? Because they're like, my job is to be tech. My job is to disrupt people. Why would I want to future proof myself, right?
Isn't such a Singaporean phrase?
Shiyan Koh: I guess so.
Jeremy Au: I never heard anybody say that they want to future proof themselves. Say, if I was in America, I say I want to bulletproof myself. It makes total sense because bullets in America, but nobody in Singapore says bulletproof.
It'd be crazy. I want a bulletproof car. I was like, what, why? It's just crazy, right? Anyway, I just find it interesting that in America, I don't think you're going to hear anybody say they're going to future proof themselves, [00:36:00] but anyway.
Shiyan Koh: Thank you for my fun fact of the day, Jeremy.
Jeremy Au: On that note, let's wrap things up and I'll see you next month.
Sounds good.