AI transcript
0:00:01 This is an iHeart podcast.
0:00:06 Run a business and not thinking about podcasting?
0:00:07 Think again.
0:00:11 More Americans listen to podcasts than ad-supported streaming music from Spotify and Pandora.
0:00:15 And as the number one podcaster, iHeart’s twice as large as the next two combined.
0:00:17 Learn how podcasting can help your business.
0:00:19 Call 844-844-iHeart.
0:00:22 Why are TSA rules so confusing?
0:00:24 You got a hoodie on, take it off!
0:00:25 I’m Manny.
0:00:25 I’m Noah.
0:00:26 This is Devin.
0:00:31 And we’re best friends and journalists with a new podcast called No Such Thing,
0:00:33 where we get to the bottom of questions like that.
0:00:34 Why are you screaming at me?
0:00:36 I can’t expect what to do.
0:00:39 Now, if the rule was the same, go off on me.
0:00:39 I deserve it.
0:00:40 You know, lock him up.
0:00:46 Listen to No Such Thing on the iHeart Radio app, Apple Podcasts, or wherever you get your podcasts.
0:00:48 No such thing.
0:00:56 Pushkin.
0:01:07 What do you understand about Maps and the world, because you have been, you know, studying it and working on it for so long?
0:01:15 I think the reason I personally love Maps is just because they’re a super, super complex problem and very, very hard thing to solve.
0:01:17 That’s why I’m so passionate about it.
0:01:18 This is Philip Kondal.
0:01:19 This is Philip Kondal.
0:01:22 He’s the chief product officer at a company called Grab.
0:01:30 Grab is huge in Southeast Asia, and its main businesses are delivery and mobility, kind of like a combination between Instacart and Uber.
0:01:33 And as a result, Maps are at the core of its business.
0:01:47 But the fascinating thing for me is, like, once you get a digital representation of the real world, you can basically, the vision that I always had for long, long years, and we haven’t solved that yet, is imagine when you went from Yahoo to Google, right?
0:01:53 Yahoo was like the static index of the web that you could, like, save for something that just leads you to a website.
0:01:57 Search for sports, and it gets you to a sports site, and then you need to navigate your way.
0:02:03 And then Google, you can say, you search for a very specific thing, and it brings you exactly to that page, right?
0:02:03 Yeah.
0:02:07 And that’s what we haven’t solved with Maps yet, right?
0:02:11 You think we’re still in the Yahoo era of Maps, is what you’re telling me?
0:02:13 Maybe slightly beyond the Yahoo era.
0:02:19 But let’s say, like, when you say, where do I find the, in Southeast Asia, durian is, like, a super popular fruit, right?
0:02:24 I say, where do I find the freshest durian for this price right now?
0:02:25 Right now.
0:02:26 In the real world.
0:02:27 Right now.
0:02:28 In the real world.
0:02:28 Right?
0:02:31 Like, which store stocks a durian right now?
0:02:34 Show me where it is, and then guide me to that store.
0:02:36 There’s nobody who has solved that problem.
0:02:43 And that’s the fascinating part for me, that I want what Google has done for the web, I want that for the real world.
0:02:45 I mean, that’s a wild problem.
0:02:50 When you formulate it that way, you would have to know everything all the time, right?
0:02:54 That’s, there’s an information problem there that is quite hard.
0:02:56 I mean, the hard problems are the fun ones, no?
0:02:56 Yeah.
0:03:06 I’m Jacob Goldstein, and this is What’s Your Problem?
0:03:10 The show where I talk to people who are trying to make technological progress.
0:03:14 Philip built and sold a mapping startup before he wound up at Grab in 2019.
0:03:19 Today, he’s based in Singapore, which is one of several countries where Grab operates.
0:03:24 Others include Indonesia, Malaysia, the Philippines, Thailand, and Vietnam.
0:03:29 And in addition to doing delivery and mobility, they also do payments.
0:03:30 It’s what they call a super app.
0:03:42 And I wanted to talk to Philip about maps in particular because, I mean, I guess I just kind of thought maps were solved or assumed that without really thinking about it.
0:03:46 And then when I started learning about Grab, I realized I was very wrong.
0:03:51 So there’s that dream of real-time maps that Philip mentioned a minute ago.
0:03:56 But there’s also something that’s really interesting and kind of more immediately relevant.
0:03:57 And that is this.
0:04:05 The online maps we use in the U.S. and other developed countries just don’t work very well in a lot of other parts of the world.
0:04:10 And when Grab launched several years ago in Malaysia, that turned out to be a big problem.
0:04:12 I mean, from day one, you needed maps.
0:04:14 So you can’t run the company without maps.
0:04:17 And Grab started using a third-party service.
0:04:22 But then what we realized, a lot of these services are built for like a developed market like the U.S.
0:04:26 and very built around the mental model of like cars.
0:04:29 And Southeast Asia is really different.
0:04:32 We’re operating primarily on motorbikes.
0:04:36 Even two-thirds of our transport trips are on the back of a motorbike.
0:04:41 And then if you know Southeast Asian cities, then you have these narrow alleys and sideways and so on.
0:04:45 And traditional maps just don’t cover them because…
0:04:46 Oh, interesting.
0:04:51 Basically, how maps are made traditionally is with these like big mapping vans.
0:04:55 I’m sure you’ve seen driving through cities, $150,000, $200,000.
0:04:57 Basically look like a Waymo, right?
0:04:58 That’s kind of like how they look like.
0:05:04 And they don’t cover the roads that we need to deliver our services and really reach our customer
0:05:07 because they expect to be picked up in their home.
0:05:12 They expect that the food get delivered to their front door and not just to the nearest street
0:05:16 that might be 200 meters away in terms of like a big car-drivable street.
0:05:23 So lots of life, lots of people live, lots of businesses are on streets that a van couldn’t even drive down if it wanted to.
0:05:24 Yeah, no chance.
0:05:32 I mean, when I do the immersions, like on the back of a motorbike, there’s absolutely no chance that with a van you can get there.
0:05:36 It was a scooter-centric world and Southeast Asia update so quickly.
0:05:39 Like, I mean, there’s like entire new neighborhoods springing up.
0:05:45 And it was traditional maps are built in a way that these big vans collect once every one or two years.
0:05:48 And we need to refresh the maps in like days.
0:05:50 So you have this problem.
0:05:52 Like, what’s the first step?
0:05:53 So you realize the maps aren’t working.
0:05:54 You’re a mobility company.
0:05:56 That’s not going to work.
0:06:00 How do you, it seems impossible to think, oh, we’ll just map everything.
0:06:01 How do you go, how do you start doing that?
0:06:03 You’re exactly right.
0:06:04 It’s a crazy problem, right?
0:06:09 I mean, the numbers that were out there when Google started mapping the U.S.,
0:06:12 they apparently spent like anywhere between half a billion to $1 billion.
0:06:16 And we clearly didn’t have that amount of money to map it.
0:06:19 So, and it seemed crazy, right?
0:06:25 Like people, when we started doing this, I mean, I got like so much feedback that people thought we were completely insane
0:06:26 because it’s going to cost us.
0:06:28 And Southeast Asia, right?
0:06:33 I mean, we are home to like close to 700 million people, which is like a lot larger than the U.S.
0:06:36 So you can imagine like how much it would normally cost.
0:06:37 Yes, 2X.
0:06:37 Exactly.
0:06:41 And crazy dense cities, traffic, tiny alleys.
0:06:44 So how do you even start to undertake this project?
0:06:45 What do you do?
0:06:47 Yeah.
0:06:49 No, so this was really the fun part.
0:06:51 That’s why it’s such a fun challenge to solve.
0:06:56 But basically what we started is we started taking it from first principles, right?
0:06:58 Like why was it so expensive to map?
0:07:00 And then we tried to solve those problems.
0:07:00 I love it.
0:07:06 And the problems why it’s so expensive normally to produce a map are a few things.
0:07:09 The one is that I just said, the mapping vehicles are extremely expensive.
0:07:18 Then the people sitting in the mapping vehicle are extremely expensive because you send them around driving to the country, sleeping in hotels.
0:07:23 So the cost like to send these vans with people around costs a fortune.
0:07:28 And like operating that is just super, super costly.
0:07:37 And that’s what we had a unique advantage because obviously we had our drivers already crisscrossing the city in like insane amounts.
0:07:43 Like, I mean, our drivers, any city is like crossed by our drivers like a hundred times or more in a day.
0:07:46 So there’s like no roads that our drivers don’t see.
0:07:50 So we thought about two things that we needed to drive the cost radically down.
0:07:53 One is we build our own collection hardware.
0:08:01 So instead of building these like massive rigs, we build small GoPro-like cameras with an AI chip in there that we can give to our drivers.
0:08:08 And instead of costing $100,000, $200,000 for a full mapping van, these cameras are on the order of like a bunch of $100.
0:08:10 So orders of magnitude is cheaper.
0:08:14 And then the second thing is we said the drivers drive around the city anyway.
0:08:16 Can they just have the camera?
0:08:18 And then we just need to pay them a little.
0:08:27 For them, it’s a great extra income, but they get their main income from doing food deliveries or doing mobility trips, and they get an extra income from that.
0:08:37 And that’s because we could deploy, like we have in Southeast Asia, probably a hundred times at least more cameras than anybody else deployed because they’re so cheap.
0:08:39 And then we don’t need to send drivers traveling everywhere.
0:08:41 So let’s break it down a little bit.
0:08:46 Like building your own hardware, like the driver part is kind of obvious, right?
0:08:48 Like, oh, we’ve got these guys already out there.
0:08:49 Could we get them to do the mapping?
0:08:53 It’s not at all obvious to me that you would need to build your own hardware.
0:08:54 So tell me about that.
0:08:57 Like, first of all, why do you need to build your own hardware?
0:08:59 Why not just, dumb question, buy a camera?
0:09:01 No, great question, actually.
0:09:02 I mean, that’s what we tried.
0:09:02 Okay.
0:09:07 So I think when we went and looked into this, because we didn’t want to build hardware, so basically we tried two things.
0:09:14 So there were these GoPros, a few hundred bucks cameras, and then there were those more professional mapping-grade cameras, 20,000 bucks.
0:09:15 Oh, wow.
0:09:18 So the problem with GoPro was a bunch of things.
0:09:20 They’re made usually as action cameras, right?
0:09:21 Like you can go skiing and so on with them.
0:09:22 Yeah.
0:09:24 But they don’t have great GPS in them.
0:09:25 Oh, interesting.
0:09:27 And basically because you don’t need it, right?
0:09:31 Like if you go, like, somewhere mountain biking, it’s typically open sky.
0:09:34 It’s not like a dense urban environment with big skyscrapers that reflect GPS.
0:09:43 So they’re decent for what people usually use GoPros, but they’re not the meter-level precision we need for high-density urban center mapping.
0:09:45 So that was one problem.
0:09:53 And the second problem with GoPros was they just have no tooling that they can operate 24-7, drivers upload automatically data.
0:10:06 So they’re not made, like usually made for like a one-two-hour recording and not this heavy-duty recording for the back of our motorbikes for 12 hours in blistering heat in Southeast Asia, tough rain and so on.
0:10:12 So that’s basically why both of these cameras that existed didn’t fit the needs that are very harsh environments.
0:10:16 So would the mapping camera have worked, but you didn’t want to spend 20 grand per camera?
0:10:19 That was presumably the simple problem with the mapping camera?
0:10:21 Yeah, 20 grand was the problem.
0:10:23 I mean, the other problems were they’re also quite heavy and bulky.
0:10:29 So those cameras are close to 10 kilograms, which is not super easy to operate.
0:10:32 So they were too bulky and too costly.
0:10:36 So what, do you just get on a plane to Shenzhen and tell them what you need?
0:10:38 Like, how do you build your own camera?
0:10:41 So this was really serendipity.
0:10:48 We actually had a small team in Shenzhen already building our battery swap lockers for our scooters.
0:10:59 And back then I talked to our CTO at that time and he was kind enough to say, like, hey, Philip, we really need to find something next for this team to do so you can have them all.
0:11:00 They can build your cameras.
0:11:02 Oh, amazing.
0:11:03 We’re very lucky.
0:11:05 So tell me about the camera they came up with.
0:11:09 Yeah, so the first camera that we built, we called it a CARTA cam.
0:11:13 So like after like CARTA in like some language means like map.
0:11:19 So basically those cameras where the first version was mounted on the helmet of drivers predominantly.
0:11:23 So it was super light, 250 grams, mounted on the helmet of drivers.
0:11:28 And it would basically be a camera with an AI chip and a really highly accurate GPS.
0:11:30 And like that’s all we needed.
0:11:38 And so they would just mount them on their helmets and go about their day and just at the end of the day, take it home, take it on their Wi-Fi, upload the data.
0:11:41 And yeah, we would get data from tons of cameras.
0:11:44 And then you made a CARTA cam 2, right?
0:11:46 Why’d you tell me about CARTA cam 2?
0:11:48 Yeah, we made a bunch of iterations of our hardware.
0:11:54 So CARTA cam 2 is our latest generation, which is basically a 360 camera.
0:12:02 So it has basically four camera lenses in all directions, has also like more advanced sensors, like LiDAR sensors in there.
0:12:12 So we basically made the setup a lot more professional that we can capture maps in even higher level of accuracy to get things like lane level navigation for our drivers.
0:12:13 Our advanced safety features.
0:12:20 So the latest camera is basically a great iteration that has taken it just a step further to capture more details on the map.
0:12:21 And where does it go?
0:12:24 Where does the camera go if somebody’s driving a scooter?
0:12:27 Yeah, so now we’ve built a mount on the back of a motorbike.
0:12:32 So we have built actually the cameras that you can either mount them at the back of your motorbike, then it doesn’t disturb you.
0:12:35 And it’s in a pole so that it’s high, so that it can see.
0:12:38 Because the camera obviously needs to have good view, even if there’s cars around.
0:12:41 So we have built a special mount on the back of a motorbike.
0:12:45 We have mounts also for cars, so some drivers also mount them on cars.
0:12:48 And then we have a backpack that you can carry if you want to map indoor.
0:12:54 The other thing that’s really important in Southeast Asia is malls are gigantic.
0:12:57 You have like these malls, which it takes you 15 minutes to walk around.
0:13:04 And so people will get a delivery to whatever, the noodle shop in the middle of the mall, and you got to put that on the map?
0:13:09 Well, the other way around, people get a delivery from the noodle shop in the mall.
0:13:10 Of course, of course, of course.
0:13:13 And our driver needs to find their way there.
0:13:18 And if they get into the wrong entrance of the mall, it might cost them 10, 15 minutes extra to walk back and forth.
0:13:26 So we need to precisely not only tell them go into the mall, but we also precisely need to tell them where to park their motorbike.
0:13:30 So what we emphasize when we build our maps, that we build them like end-to-end.
0:13:32 And we say, here’s where you park your motorbike.
0:13:33 Here’s where you walk.
0:13:35 Here’s where you pick up the noodles.
0:13:37 And then you can do the same in reverse.
0:13:40 How many of these cameras do you have out there, like, today?
0:13:43 How many people are driving around mapping with these cameras today-ish?
0:13:47 By end of this year, we have about 20,000 cameras in the field.
0:13:54 And to give you a sense, like, professional mapping companies in Southeast Asia, to my best of my knowledge, have about tens of cars.
0:13:57 But not tens of thousands, tens of cars.
0:14:05 So does that mean they have basically ceded it to you, like, that you have won the mapping Southeast Asia fight?
0:14:07 I think it’s still so, so early.
0:14:08 Oh, fair. Interesting.
0:14:10 I mean, there’s still so many things we want to do.
0:14:11 Interesting. I love that.
0:14:16 So let’s talk a little more about what you’re doing now, and then let’s talk about what you want to do.
0:14:19 So you have an incredible amount of data coming in now, right?
0:14:28 I mean, if you have 20,000 cameras driving around every day, like, that is a wild amount of input.
0:14:30 What are you doing with all that data?
0:14:32 I mean, there’s basic mapping, and I’m sure you’ve got that.
0:14:33 But what’s the, like, next level?
0:14:35 Presumably you have AI, you said you have LiDAR.
0:14:38 Like, tell me all the things you know because you have all this data.
0:14:39 You’re right.
0:14:45 I think the basic vision that we have is get an accurate representation of the real world in real time.
0:14:49 That’s the goal of mapping.
0:14:50 In real time is crazy.
0:14:51 In real time is crazy.
0:14:53 Yeah, that’s wild.
0:14:54 But fine, that’s the goal.
0:14:57 Just, like, right now, what do you know?
0:15:04 Yeah, so a bunch of things is, I mean, first of all, everything that you need for navigation, roads, how are roads drivable?
0:15:06 Is the road safe to drive?
0:15:09 Does it have a lot of potholes and things like that?
0:15:10 And then signage, obviously.
0:15:11 Is it a one-way road?
0:15:12 Is it not a one-way road?
0:15:15 Can you warn me about a particular pothole?
0:15:18 Can you be, like, be careful there’s a pothole on the left coming up?
0:15:19 Try the right lane?
0:15:21 And we’re doing actually very cool things.
0:15:29 We’re doing right now, we’ve already launched safety navigation that tells you, A, is the road safe to drive based on potholes?
0:15:30 And does it have street lighting?
0:15:31 Is it safe?
0:15:35 Because if the road at night is well lit, it’s a lot safer to drive than it’s not.
0:15:39 So that’s kind of like a practical use case that we already can do with that.
0:15:39 Interesting.
0:15:49 As I’m sure you know, like, Waze in this country tells you where there’s, like, a speed camera or, like, a speed trap.
0:15:50 Do you do that?
0:15:51 Yeah, absolutely.
0:15:59 So we have an AI voice reporting that drivers can share anything, say, like, oh, the right lane of this road is closed.
0:16:03 And then it gets processed and basically used for all the other drivers.
0:16:07 So we’ve launched that and it’s been really successful as well.
0:16:12 So when you say it’s AI, does that mean the driver just says it and it gets integrated into the maps?
0:16:14 Or what does that mean?
0:16:16 Yeah, that’s exactly right.
0:16:19 There’s a feature we actually work closely together with our friends at OpenAI.
0:16:24 So the driver in natural language reports an issue and then it gets processed.
0:16:26 And if you’re not clear, we ask you a follow-up question.
0:16:28 If you say, hey, the right lane is closed.
0:16:30 And we say, like, oh, do you mean this road?
0:16:32 And then it says, driver, yeah, that’s exactly the road I’m talking about.
0:16:36 And then we process it and basically warn all the other drivers accordingly.
0:16:38 So you mentioned your friends at OpenAI.
0:16:41 It seems like good friends to have, good place to have friends.
0:16:43 I know you have a deal with them.
0:16:45 Like, what is the nature of your deal with OpenAI?
0:16:48 And more generally, what’s the work you’re doing with them?
0:16:55 So mapping is one of the key things we do together with them and use their models to improve our maps.
0:17:00 The voice stuff that I shared earlier is like one of our highlight features.
0:17:04 And in general, we are just embedding like AI in everything we’re doing.
0:17:09 And we’re having like over a thousand different AI models that we’re working on.
0:17:13 So it’s basically deeply, deeply embedded in many, many of the things that we’re doing.
0:17:15 Tell me more.
0:17:17 Like, what are some more examples of how you’re using AI?
0:17:27 So one of the latest things that I really like, that’s really cool, is the other thing that has really huge impact on our marketplace is weather.
0:17:30 In Southeast Asia, the rain is insane.
0:17:32 It is like pouring down.
0:17:34 The roads are flooding.
0:17:40 So, and for our services, for mobility, for deliveries, that has a tremendous impact on that.
0:17:46 So knowing when it rains early is super, super critical so we can adjust the marketplace.
0:17:50 We’ve launched AI-based rain detection that does a bunch of things.
0:17:55 So we deploy sensors, other sensors in cars.
0:17:59 So we have basically a device that’s called a Kata dongle that we deploy in the cars that we own.
0:18:02 And it plugs into a port in the car that’s called an OBD2 port.
0:18:04 And that reads when the windshield wiper is going.
0:18:05 Oh, genius.
0:18:11 Such a simple way, such a simple way to know if it’s raining.
0:18:13 Does the car have the windshield wipers on?
0:18:14 It’s so second order.
0:18:15 I love it, though.
0:18:16 Yeah, and how fast it is.
0:18:21 Oh, how fast the car is going because the slower it’s going, the heavier the rain.
0:18:22 Is that the inference?
0:18:26 How fast the windshield wiper is going because you can have it like this or this, right?
0:18:31 So, okay, so, and what do you do with that data?
0:18:32 That’s the input.
0:18:33 What’s the output?
0:18:39 The output is basically we know the moment it rains, we know demand will go up and supply
0:18:40 of drivers will go down.
0:18:43 So what we do, we try to activate more drivers.
0:18:47 So we would send out to all the drivers, hey, it’s starting to rain.
0:18:48 There’s a fantastic earning opportunity.
0:18:53 If you’re not working, now’s a great chance to get on the road and make extra money.
0:19:00 And then we try to actively get the supply of drivers up so that we can keep our reliability
0:19:01 at the levels we need.
0:19:03 I know people get all worked up about surge pricing.
0:19:05 Do you use surge pricing in that setting?
0:19:07 It sounds like a classic use of surge pricing.
0:19:08 Everybody wants a driver.
0:19:09 Nobody wants to drive.
0:19:10 What you need is a higher price.
0:19:11 Do you do that?
0:19:12 That’s exactly right.
0:19:15 Or like you need to motivate the drivers to come on the road.
0:19:15 Yeah.
0:19:16 No, there’s a market.
0:19:18 I’m pro surge pricing.
0:19:19 That’s a great example.
0:19:21 What’s another example of the way you’re using AI?
0:19:26 So we use AI to translate in many scenarios.
0:19:32 For example, when I go to Jakarta, I obviously don’t speak any Bahasa, but I can message our
0:19:36 drivers and we real-time translate any message and send in the chat to the driver.
0:19:39 And he can reply back to me in Bahasa.
0:19:40 And I see it in English on my side.
0:19:42 He sees it in Bahasa on their side.
0:19:45 But the more important one for me was like food menu translation.
0:19:50 So we invested quite a bit because whenever I go to Thailand and then like, I mean, I can’t
0:19:51 read any Thai script, obviously.
0:19:56 And I, in the past, I look at like some things on the menu and I have no idea.
0:19:59 Is it like, for me, always the word is ultra spicy.
0:20:00 Can I eat it or not?
0:20:02 And I didn’t even know what the item is.
0:20:06 I can look at the picture and kind of guess, but sometimes merchants don’t have pictures.
0:20:11 So we use AI to translate all these menus in all kinds of languages.
0:20:14 So that, that has been a super impactful one as well.
0:20:20 We’ll be back in just a minute.
0:20:30 Run a business and not thinking about podcasting?
0:20:31 Think again.
0:20:36 More Americans listen to podcasts than ad-supported streaming music from Spotify and Pandora.
0:20:40 And as the number one podcaster, iHeart’s twice as large as the next two combined.
0:20:43 So whatever your customers listen to, they’ll hear your message.
0:20:47 Plus, only iHeart can extend your message to audiences across broadcast radio.
0:20:49 Think podcasting can help your business?
0:20:50 Think iHeart.
0:20:52 Streaming, radio, and podcasting.
0:20:55 Call 844-844-IHEART to get started.
0:20:58 That’s 844-844-IHEART.
0:21:02 Imagine that you’re on an airplane and all of a sudden you hear this.
0:21:10 Attention passengers, the pilot is having an emergency and we need someone, anyone, to land this plane.
0:21:11 Think you could do it?
0:21:18 It turns out that nearly 50% of men think that they could land the plane with the help of air traffic control.
0:21:21 And they’re saying like, okay, pull this until this.
0:21:22 Do this, pull that, turn this.
0:21:23 It’s just…
0:21:23 I can do it with my eyes closed.
0:21:24 I’m Manny.
0:21:25 I’m Noah.
0:21:26 This is Devin.
0:21:31 And on our new show, No Such Thing, we get to the bottom of questions like these.
0:21:34 Join us as we talk to the leading expert on overconfidence.
0:21:41 Those who lack expertise, lack the expertise they need to recognize that they lack expertise.
0:21:45 And then as we try the whole thing out for real.
0:21:46 Wait, what?
0:21:47 Oh, that’s the runway.
0:21:49 I’m looking at this thing, see?
0:21:56 Listen to No Such Thing on the iHeartRadio app, Apple Podcasts, or wherever you get your podcasts.
0:22:02 So I’m curious about GrabMaps Enterprise, right?
0:22:08 Like, you’re selling something related to your maps to big companies, right?
0:22:11 To Microsoft, to Amazon, to governments.
0:22:11 Like, what?
0:22:12 Tell me about that business.
0:22:15 That was quite an interesting one, right?
0:22:19 Like, we would have never imagined early on when we built our maps that this business would be created.
0:22:25 But we got people, like, once we publicized our maps, and people have seen that they work generally quite well,
0:22:28 we got people approaching us saying, hey, can we use them as well?
0:22:33 And that was, like, the genesis of the enterprise business.
0:22:40 And then we started working very closely with AWS, where any developer on their platform can use our maps now.
0:22:44 So we have over 100 different developers already using it.
0:22:52 We partnered with Microsoft, as you said, and a bunch of other large tech companies that in Southeast Asia started using our maps,
0:22:59 because they’ve seen, I mean, all these things that I shared earlier, that it just really serves their needs in Southeast Asia a lot better than anything else out there.
0:23:03 What’s an example of something someone has built, you know, on top of it?
0:23:15 A really cool startup that I’ve seen that used our maps, what they do, they go from merchant to merchant and collect, like, the old oil that they use for cooking, and then basically recycle it.
0:23:19 So they send somebody around going to all these merchants collecting that.
0:23:26 And in the past, and a lot of these merchants are, like, small neighborhood, mom-and-pop shops, you know, all these, like, little side roads and alleys.
0:23:31 And they had a hard time for the person going around collecting all of the stuff, finding that.
0:23:39 And that’s one of these examples where they use grab maps to make the navigation of the person going around and collecting all of that a lot easier.
0:23:50 And there’s many of these kind of stories where people use it for things that are very similar in this kind of finding the last mile has been really, really hard before.
0:23:54 So I’m curious what you’re working on now.
0:23:58 Like, yeah, what are some of the things you’re trying to figure out that you haven’t figured out yet?
0:24:03 I think the one thing that we’re really passionate about is solving more of the indoor problem.
0:24:05 So that’s one thing that we’re really mapping more.
0:24:09 So we have a decent amount of malls already mapped, but there’s still so much more to do.
0:24:12 So hopefully we can, like, find that.
0:24:21 Whenever you go into these malls, we can exactly help you to find whatever you’re looking for, a specific store or a general kind of, like, shop or so.
0:24:24 So that’s one thing that I really want us to crack.
0:24:34 So is the general shop idea that, like, if I’m just not working for Grab, I’m just in the general public and I’m like, I want to buy a pair of shorts.
0:24:37 I just type that into Grab Maps and Grab Maps tells me where to go.
0:24:38 Is that what you’re thinking of there?
0:24:41 Yes, but we always think about it with a hyper-local twist.
0:24:52 And what that means, as an example, let’s say, in Indonesia, a large part of the population is Muslim, which means they generally eat halal.
0:24:56 And this is, like, all the mapping platforms, they don’t support that.
0:25:02 You can, of course, for every search query, you can say restaurant halal, this halal, this halal, but you cannot make it part of your user profile.
0:25:06 But it’s not something that you switch every day.
0:25:10 Either your preference is halal or it’s not.
0:25:18 And none of the mapping platforms support putting in your profile, I only want halal restaurants because I don’t eat anything else.
0:25:26 And those are the kind of things that we see in Southeast Asia we need to really solve because, again, nobody else would solve those kind of problems.
0:25:34 So capturing this data accurately and knowing all these details, those are the kind of things that we really put a lot of emphasis on.
0:25:38 Are there technical problems you’re working on?
0:25:48 Like, are there things where you haven’t figured out the right tech or where you need to build something or where AI models aren’t quite where they need to be, but you’re trying to push them?
0:25:50 I think what we’re trying to do is what I said earlier.
0:25:53 Like, you want to have a real-time, accurate model of the world.
0:25:55 Yeah, so let’s talk about that.
0:26:00 Like, real-time is a crazy phrase in that context, right?
0:26:04 Like, when you say real-time model of the world, what are you thinking of?
0:26:09 Yeah, I mean, a simple use case would be know where there’s parking right now.
0:26:10 Yeah.
0:26:14 Not like where there’s parking in general, but let’s say a roadside parking.
0:26:24 If you had another grab car driving past and say, hey, 15 seconds ago there was a freak parking spot on the right, that’s extremely useful.
0:26:25 That is extremely useful.
0:26:31 I know people in Brooklyn and San Francisco who would love to have that functionality.
0:26:38 Yeah, I think for us, really, the goal is always what we know is like shaving off seconds of every delivery, right?
0:26:41 Like, that’s what really, really makes a Delta.
0:26:49 I think that I really loved Steve Jobs’ old mental model when he convinced people to make the Mac boot, like, 10 seconds faster.
0:26:57 And he said, like, well, if you take this, I don’t remember the exact math for him, but he said, like, if you make the Mac boot 10 seconds faster, there’s 50 million people using it.
0:27:04 Every year you save, like, I don’t know, it’s like 10 lifetimes of people’s lifetime waiting for the Mac to boot, something like that.
0:27:15 And for us, right, like, across, I mean, across, I calculate and I share this always with the team, across a billion deliveries, 2.5 seconds saved across every delivery is roughly one lifetime that you can save.
0:27:27 So any second we can shave off by getting cars to park faster, by getting motorbikes to park faster, park at the right space at the right time, that’s kind of the problems that we’re really passionate on solving.
0:27:30 How long does it take you to do a billion deliveries?
0:27:32 Is it a billion per what?
0:27:38 We don’t publish exact data, but it’s the order of magnitude of, like, a year or less.
0:27:39 Oh, wow.
0:27:39 Okay.
0:27:42 That’s a big number.
0:27:45 So how do you get real-time parking data?
0:27:49 I mean, so is the constraint now just getting more cameras to more drivers?
0:27:51 Like, what’s the rate-limiting step?
0:27:53 Yeah, great question.
0:27:55 I think more cameras is one constraint.
0:28:04 The other constraint is doing all the smart processing on the edge, because obviously you cannot upload all these data, because that would be extremely costly.
0:28:09 And mobile networks in Southeast Asia aren’t quite that powerful that we could upload millions of video streams to the cloud.
0:28:15 So we are working a lot on what is called, like, edge AI, that we can run all these models.
0:28:19 That are powerful, but not in the cloud, but on the edge, or at least on a mix of both.
0:28:23 So in this case, is the edge the actual camera?
0:28:30 I mean, what is, like, is the dream the camera itself is doing the work and just uploading something very simple to the network?
0:28:32 That’s in many places already happening.
0:28:35 We already have quite powerful AI chips in the cameras.
0:28:42 But, I mean, of course, like, no mobile phone, no edge camera right now is as powerful as, let’s say, chat GPT with GPT-4-0.
0:28:43 Well, sure.
0:28:52 I mean, you have this sort of spectrum where you have a, whatever, a hundred million dollar data center on one end and a hundred dollar camera on the other and some things in between, right?
0:28:55 So what can you do on the camera now?
0:28:57 We do things like privacy.
0:28:59 So we blur people’s faces.
0:29:02 We blur license plates because we never want to upload this information.
0:29:05 We do weather detection, what I said earlier, with rain.
0:29:09 So we detect on the cameras if it’s raining, things like that.
0:29:12 We detect traffic signs and see if they’ve changed.
0:29:16 So we actually already run a large part of processing on the edge.
0:29:22 And then only if something has changed, then we upload some data and do a validation on the server.
0:29:26 But that already allows us to reduce what we upload by more than 90 percent.
0:29:29 Just tell the server if something is different, basically.
0:29:31 Yeah, exactly.
0:29:37 So I know, you know, we’ve talked mostly about maps, but obviously Grab is doing a lot of different things.
0:29:40 Are there other parts of the business that we should talk about?
0:29:42 The business is so fascinating.
0:29:45 So if we have time, we should talk about all of our businesses.
0:29:49 Yeah, just tell me, what’s one other sort of frontier, one other thing you’re working on?
0:29:58 I think the other thing which we’ve really invested deeply, which is also quite closely connected to our maps, is when we look at across the delivery journey.
0:30:04 The other part where a lot of time is spent is in the merchant, cooking and preparing the food.
0:30:17 And that’s an area where we spend a lot of time optimizing together with the merchants to make sure that the food is prepared in the right time, that the driver arrives in the right time.
0:30:19 So we’ve done lots and lots of cool things.
0:30:27 So, for example, we’ve built a data science model that accurately predicts at every time of the day how long the merchant needs to prepare an order.
0:30:33 So we can detect the busyness of the merchant and say, if you know, oh, the merchant has gotten a lot of orders.
0:30:40 Normally they take to prepare, they prepare to like a banh mi in like three minutes, but when they’re very busy, they take seven minutes.
0:30:45 And the merchants don’t want that our drivers crowd their store and wait in troves.
0:30:53 So we typically allocate the driver only when we know, okay, the food is ready in seven minutes and we don’t allocate the driver immediately.
0:31:00 When we know, oh, there’s a driver two minutes away, we just allocate them like five minutes later so that he’s in the shop just in time.
0:31:12 And that also cuts the delivery journey, makes it a lot more pleasant for the merchant, not have like a lot of people crowd their store and allows us to offer more affordable price to the consumer because they don’t need to pay for all these minutes of the driver waiting.
0:31:17 So just all these optimizations, all these different margins where you can optimize.
0:31:26 So if we think about the whatever, five years, 10 years out, when you think about this sort of medium to long term, like what’s your dream?
0:31:28 Like how’s it work?
0:31:28 What’s going on?
0:31:31 I think like for me, the world is changing so fast.
0:31:35 Five to 10 years prediction is really, really hard with all the AI advancements.
0:31:36 How far you want to go?
0:31:37 Two years?
0:31:38 Four years?
0:31:38 You tell me.
0:31:40 Like what’s your, just give me a dream for the future.
0:31:44 I mean, the obvious one that I’m very passionate about is robotics.
0:31:47 They make so much sense in our marketplace.
0:31:56 So if you can add robotics to our marketplace, that will be a huge, huge change and something we’re actively working on to make happen.
0:32:00 So that’s, I think, probably the thing I’m most passionate about to change.
0:32:02 Tell me what you’re actively working on with robotics.
0:32:06 We haven’t shared much, which we do on the delivery side.
0:32:09 But for example, I mean, things like autonomous cars.
0:32:16 We’ve signed an agreement with a bunch of like autonomous car providers to come with us to Singapore and so on.
0:32:18 So we’ll do a bunch of things in that space.
0:32:23 But basically, you can imagine that it will be in many, many parts of our delivery chain.
0:32:26 You need to build an autonomous scooter based on what you’ve told me, right?
0:32:31 I feel like the analogy to the map story is the autonomous scooter that can go down all the alleyways.
0:32:32 Great.
0:32:35 If you ever want to have a job in our product team, feel free to join us.
0:32:38 Build autonomous scooters with us.
0:32:38 Because you’re exactly right.
0:32:41 We need to build products that work in our region.
0:32:43 So that’s exactly the right mindset that we’re trying.
0:32:48 Like what we say always, we want to build hyper-local products that work in Southeast Asia.
0:32:53 We’ll be back in a minute with the lightning round.
0:33:04 Run a business and not thinking about podcasting?
0:33:05 Think again.
0:33:10 More Americans listen to podcasts than ad-supported streaming music from Spotify and Pandora.
0:33:14 And as the number one podcaster, iHeart’s twice as large as the next two combined.
0:33:18 So whatever your customers listen to, they’ll hear your message.
0:33:22 Plus, only iHeart can extend your message to audiences across broadcast radio.
0:33:24 Think podcasting can help your business?
0:33:25 Think iHeart.
0:33:27 Streaming, radio, and podcasting.
0:33:30 Let us show you at iHeartAdvertising.com.
0:33:32 That’s iHeartAdvertising.com.
0:33:36 Imagine that you’re on an airplane, and all of a sudden you hear this.
0:33:44 Attention passengers, the pilot is having an emergency, and we need someone, anyone, to land this plane.
0:33:45 Think you could do it?
0:33:52 It turns out that nearly 50% of men think that they could land the plane with the help of air traffic control.
0:33:56 And they’re saying like, okay, pull this, do this, pull that, turn this.
0:33:57 It’s just…
0:33:57 I can do it with my eyes closed.
0:33:58 I’m Manny.
0:33:59 I’m Noah.
0:34:00 This is Devin.
0:34:05 And on our new show, No Such Thing, we get to the bottom of questions like these.
0:34:08 Join us as we talk to the leading expert on overconfidence.
0:34:15 Those who lack expertise, lack the expertise they need to recognize that they lack expertise.
0:34:19 And then, as we try the whole thing out for real.
0:34:20 Wait, what?
0:34:21 Oh, that’s the runway.
0:34:23 I’m looking at this thing, see?
0:34:30 Listen to No Such Thing on the iHeartRadio app, Apple Podcasts, or wherever you get your podcasts.
0:34:34 Okay, let’s finish with the lightning round.
0:34:44 So, I read on your bio, you write that you mostly travel between Singapore, San Francisco, Berlin, and Cluj.
0:34:48 Tell me about Cluj, the other three I’m familiar with.
0:34:50 Cluj, I don’t know anything about.
0:34:51 How’s that wind up on that list?
0:34:56 It’s a city in Romania, in the heart of Transylvania, actually.
0:35:01 And we have a small engineering center there in our maps team.
0:35:07 So, I’ve spent a lot of time there because for my own startup, that’s how I wound up in Cluj originally.
0:35:13 For my startup, Scobler, which was a maps and navigation startup, I built an engineering team in Cluj.
0:35:20 So, I’ve been going to Cluj and working with engineers there for the last 15 years or so.
0:35:21 What’s Cluj like?
0:35:22 It’s fun.
0:35:27 It’s a student town, so high energy, young population, very smart.
0:35:30 The computer science department, very good.
0:35:33 They used to clone the IBM mainframes for the Soviet Union.
0:35:35 So, a bunch of hardcore engineers.
0:35:38 You also write in that bio, you wrote,
0:35:42 In Berlin, I experienced some of the best parties ever.
0:35:44 Tell me about a Berlin party.
0:35:46 That was prior life.
0:35:50 That was before I moved to Southeast Asia.
0:35:55 But Berlin was a fun, fun, fun journey for, I lived there for five, six years, maybe.
0:35:58 You didn’t tell me anything about any party.
0:36:03 So, and then you also say you lived and worked in Singapore and Berlin and San Francisco.
0:36:10 And I’m curious, like, how is sort of, whatever, professional culture, work life, different in those places?
0:36:12 What are the sort of striking differences?
0:36:15 I mean, Germany is extremely direct, right?
0:36:16 Like, that’s where I’m originally from.
0:36:17 I’m originally German.
0:36:27 And Germans are known to be super, super direct, which in Southeast Asia doesn’t work super well if they’re not accustomed to it.
0:36:40 So, I think for me, the thing that I needed to adjust most is to become a lot more indirect, to become a lot more, like, have those conversations in, like, smaller groups, not in front of everybody.
0:36:46 So, I definitely had to adjust my style quite a lot when moving around.
0:36:47 But I found that always incredibly fun.
0:36:49 I always loved learning about new cultures.
0:36:54 So, after some adjustments, I really, really like it here.
0:36:55 That is super interesting.
0:37:08 Was there, as you were saying that, I was wondering, like, was there a particular moment when you realized, oh, I’m not behaving correctly in this cultural context?
0:37:09 Yeah.
0:37:11 I mean, this predates Grab.
0:37:13 So, this was when I was first doing business in China.
0:37:23 So, I sold my startup to a Silicon Valley company, but I also got a 200-people team in China report to me, which was the first time that I ever managed a team in China.
0:37:31 And I was extremely surprised why they wouldn’t tell me about all the mistakes, all the things that I did wrong.
0:37:33 And I said, like, that’s very unusual.
0:37:37 Normally, I get a lot of, like, pushback from engineers.
0:37:41 And then I started, like, asking people, why don’t they do this?
0:37:46 And then people said, like, well, it’s kind of rude in a public forum to say the boss is wrong.
0:37:55 And then I realized, like, oh, that’s why I just need to, like, change how I’m asking questions, how I’m managing teams.
0:37:57 And that’s when I first managed teams in China.
0:37:59 It took me quite a while to figure out, honestly.
0:38:08 And how does San Francisco fit in relation to both working in Berlin and working in, you know, China, Singapore?
0:38:11 Where’s the U.S. on that continuum?
0:38:18 The U.S. for me, I think the thing that I really loved in San Francisco was just the craziness of ambition.
0:38:26 When I spent, like, April back in SF for a month to, like, Vibecode and hack on a bunch of hobby projects.
0:38:30 And, like, you go to every random coffee shop and there’s somebody who said they work on a billion dollar idea.
0:38:32 And they’re just starting.
0:38:33 They have nothing yet.
0:38:36 They’re just like, okay, I’m going to make it big.
0:38:46 So, I think that that ambition and that willingness to openly declare it, even if people know it’s super unlikely, there’s not thousands of people who start billion dollar companies.
0:38:50 But in the Bay Area, there’s thousands of people who say they will.
0:38:57 And this conviction and this, like, optimism, I think that was, for me, one of the most striking things in the U.S.
0:39:00 and that I still love the energy whenever I go there, basically.
0:39:10 Philip Kondal is the Chief Product Officer at Graph.
0:39:14 Please email us at problematpushkin.fm.
0:39:17 We are always looking for new guests for the show.
0:39:21 Today’s show was produced by Trina Menino and Gabriel Hunter-Chang.
0:39:26 It was edited by Alexander Gerriton and engineered by Sarah Brugger.
0:39:30 I’m Jacob Goldstein, and we’ll be back next week with another episode of What’s Your Problem?
0:39:43 Why are TSA rules so confusing?
0:39:56 Why are you screaming at me?
0:39:57 I can’t expect what to do.
0:40:00 Now, if the rule was the same, go off on me.
0:40:01 I deserve it.
0:40:02 You know, lock him up.
0:40:08 Listen to No Such Thing on the iHeartRadio app, Apple Podcasts, or wherever you get your podcasts.
0:40:09 No such thing.
0:40:12 This is an iHeart Podcast.
0:40:15 you
0:00:06 Run a business and not thinking about podcasting?
0:00:07 Think again.
0:00:11 More Americans listen to podcasts than ad-supported streaming music from Spotify and Pandora.
0:00:15 And as the number one podcaster, iHeart’s twice as large as the next two combined.
0:00:17 Learn how podcasting can help your business.
0:00:19 Call 844-844-iHeart.
0:00:22 Why are TSA rules so confusing?
0:00:24 You got a hoodie on, take it off!
0:00:25 I’m Manny.
0:00:25 I’m Noah.
0:00:26 This is Devin.
0:00:31 And we’re best friends and journalists with a new podcast called No Such Thing,
0:00:33 where we get to the bottom of questions like that.
0:00:34 Why are you screaming at me?
0:00:36 I can’t expect what to do.
0:00:39 Now, if the rule was the same, go off on me.
0:00:39 I deserve it.
0:00:40 You know, lock him up.
0:00:46 Listen to No Such Thing on the iHeart Radio app, Apple Podcasts, or wherever you get your podcasts.
0:00:48 No such thing.
0:00:56 Pushkin.
0:01:07 What do you understand about Maps and the world, because you have been, you know, studying it and working on it for so long?
0:01:15 I think the reason I personally love Maps is just because they’re a super, super complex problem and very, very hard thing to solve.
0:01:17 That’s why I’m so passionate about it.
0:01:18 This is Philip Kondal.
0:01:19 This is Philip Kondal.
0:01:22 He’s the chief product officer at a company called Grab.
0:01:30 Grab is huge in Southeast Asia, and its main businesses are delivery and mobility, kind of like a combination between Instacart and Uber.
0:01:33 And as a result, Maps are at the core of its business.
0:01:47 But the fascinating thing for me is, like, once you get a digital representation of the real world, you can basically, the vision that I always had for long, long years, and we haven’t solved that yet, is imagine when you went from Yahoo to Google, right?
0:01:53 Yahoo was like the static index of the web that you could, like, save for something that just leads you to a website.
0:01:57 Search for sports, and it gets you to a sports site, and then you need to navigate your way.
0:02:03 And then Google, you can say, you search for a very specific thing, and it brings you exactly to that page, right?
0:02:03 Yeah.
0:02:07 And that’s what we haven’t solved with Maps yet, right?
0:02:11 You think we’re still in the Yahoo era of Maps, is what you’re telling me?
0:02:13 Maybe slightly beyond the Yahoo era.
0:02:19 But let’s say, like, when you say, where do I find the, in Southeast Asia, durian is, like, a super popular fruit, right?
0:02:24 I say, where do I find the freshest durian for this price right now?
0:02:25 Right now.
0:02:26 In the real world.
0:02:27 Right now.
0:02:28 In the real world.
0:02:28 Right?
0:02:31 Like, which store stocks a durian right now?
0:02:34 Show me where it is, and then guide me to that store.
0:02:36 There’s nobody who has solved that problem.
0:02:43 And that’s the fascinating part for me, that I want what Google has done for the web, I want that for the real world.
0:02:45 I mean, that’s a wild problem.
0:02:50 When you formulate it that way, you would have to know everything all the time, right?
0:02:54 That’s, there’s an information problem there that is quite hard.
0:02:56 I mean, the hard problems are the fun ones, no?
0:02:56 Yeah.
0:03:06 I’m Jacob Goldstein, and this is What’s Your Problem?
0:03:10 The show where I talk to people who are trying to make technological progress.
0:03:14 Philip built and sold a mapping startup before he wound up at Grab in 2019.
0:03:19 Today, he’s based in Singapore, which is one of several countries where Grab operates.
0:03:24 Others include Indonesia, Malaysia, the Philippines, Thailand, and Vietnam.
0:03:29 And in addition to doing delivery and mobility, they also do payments.
0:03:30 It’s what they call a super app.
0:03:42 And I wanted to talk to Philip about maps in particular because, I mean, I guess I just kind of thought maps were solved or assumed that without really thinking about it.
0:03:46 And then when I started learning about Grab, I realized I was very wrong.
0:03:51 So there’s that dream of real-time maps that Philip mentioned a minute ago.
0:03:56 But there’s also something that’s really interesting and kind of more immediately relevant.
0:03:57 And that is this.
0:04:05 The online maps we use in the U.S. and other developed countries just don’t work very well in a lot of other parts of the world.
0:04:10 And when Grab launched several years ago in Malaysia, that turned out to be a big problem.
0:04:12 I mean, from day one, you needed maps.
0:04:14 So you can’t run the company without maps.
0:04:17 And Grab started using a third-party service.
0:04:22 But then what we realized, a lot of these services are built for like a developed market like the U.S.
0:04:26 and very built around the mental model of like cars.
0:04:29 And Southeast Asia is really different.
0:04:32 We’re operating primarily on motorbikes.
0:04:36 Even two-thirds of our transport trips are on the back of a motorbike.
0:04:41 And then if you know Southeast Asian cities, then you have these narrow alleys and sideways and so on.
0:04:45 And traditional maps just don’t cover them because…
0:04:46 Oh, interesting.
0:04:51 Basically, how maps are made traditionally is with these like big mapping vans.
0:04:55 I’m sure you’ve seen driving through cities, $150,000, $200,000.
0:04:57 Basically look like a Waymo, right?
0:04:58 That’s kind of like how they look like.
0:05:04 And they don’t cover the roads that we need to deliver our services and really reach our customer
0:05:07 because they expect to be picked up in their home.
0:05:12 They expect that the food get delivered to their front door and not just to the nearest street
0:05:16 that might be 200 meters away in terms of like a big car-drivable street.
0:05:23 So lots of life, lots of people live, lots of businesses are on streets that a van couldn’t even drive down if it wanted to.
0:05:24 Yeah, no chance.
0:05:32 I mean, when I do the immersions, like on the back of a motorbike, there’s absolutely no chance that with a van you can get there.
0:05:36 It was a scooter-centric world and Southeast Asia update so quickly.
0:05:39 Like, I mean, there’s like entire new neighborhoods springing up.
0:05:45 And it was traditional maps are built in a way that these big vans collect once every one or two years.
0:05:48 And we need to refresh the maps in like days.
0:05:50 So you have this problem.
0:05:52 Like, what’s the first step?
0:05:53 So you realize the maps aren’t working.
0:05:54 You’re a mobility company.
0:05:56 That’s not going to work.
0:06:00 How do you, it seems impossible to think, oh, we’ll just map everything.
0:06:01 How do you go, how do you start doing that?
0:06:03 You’re exactly right.
0:06:04 It’s a crazy problem, right?
0:06:09 I mean, the numbers that were out there when Google started mapping the U.S.,
0:06:12 they apparently spent like anywhere between half a billion to $1 billion.
0:06:16 And we clearly didn’t have that amount of money to map it.
0:06:19 So, and it seemed crazy, right?
0:06:25 Like people, when we started doing this, I mean, I got like so much feedback that people thought we were completely insane
0:06:26 because it’s going to cost us.
0:06:28 And Southeast Asia, right?
0:06:33 I mean, we are home to like close to 700 million people, which is like a lot larger than the U.S.
0:06:36 So you can imagine like how much it would normally cost.
0:06:37 Yes, 2X.
0:06:37 Exactly.
0:06:41 And crazy dense cities, traffic, tiny alleys.
0:06:44 So how do you even start to undertake this project?
0:06:45 What do you do?
0:06:47 Yeah.
0:06:49 No, so this was really the fun part.
0:06:51 That’s why it’s such a fun challenge to solve.
0:06:56 But basically what we started is we started taking it from first principles, right?
0:06:58 Like why was it so expensive to map?
0:07:00 And then we tried to solve those problems.
0:07:00 I love it.
0:07:06 And the problems why it’s so expensive normally to produce a map are a few things.
0:07:09 The one is that I just said, the mapping vehicles are extremely expensive.
0:07:18 Then the people sitting in the mapping vehicle are extremely expensive because you send them around driving to the country, sleeping in hotels.
0:07:23 So the cost like to send these vans with people around costs a fortune.
0:07:28 And like operating that is just super, super costly.
0:07:37 And that’s what we had a unique advantage because obviously we had our drivers already crisscrossing the city in like insane amounts.
0:07:43 Like, I mean, our drivers, any city is like crossed by our drivers like a hundred times or more in a day.
0:07:46 So there’s like no roads that our drivers don’t see.
0:07:50 So we thought about two things that we needed to drive the cost radically down.
0:07:53 One is we build our own collection hardware.
0:08:01 So instead of building these like massive rigs, we build small GoPro-like cameras with an AI chip in there that we can give to our drivers.
0:08:08 And instead of costing $100,000, $200,000 for a full mapping van, these cameras are on the order of like a bunch of $100.
0:08:10 So orders of magnitude is cheaper.
0:08:14 And then the second thing is we said the drivers drive around the city anyway.
0:08:16 Can they just have the camera?
0:08:18 And then we just need to pay them a little.
0:08:27 For them, it’s a great extra income, but they get their main income from doing food deliveries or doing mobility trips, and they get an extra income from that.
0:08:37 And that’s because we could deploy, like we have in Southeast Asia, probably a hundred times at least more cameras than anybody else deployed because they’re so cheap.
0:08:39 And then we don’t need to send drivers traveling everywhere.
0:08:41 So let’s break it down a little bit.
0:08:46 Like building your own hardware, like the driver part is kind of obvious, right?
0:08:48 Like, oh, we’ve got these guys already out there.
0:08:49 Could we get them to do the mapping?
0:08:53 It’s not at all obvious to me that you would need to build your own hardware.
0:08:54 So tell me about that.
0:08:57 Like, first of all, why do you need to build your own hardware?
0:08:59 Why not just, dumb question, buy a camera?
0:09:01 No, great question, actually.
0:09:02 I mean, that’s what we tried.
0:09:02 Okay.
0:09:07 So I think when we went and looked into this, because we didn’t want to build hardware, so basically we tried two things.
0:09:14 So there were these GoPros, a few hundred bucks cameras, and then there were those more professional mapping-grade cameras, 20,000 bucks.
0:09:15 Oh, wow.
0:09:18 So the problem with GoPro was a bunch of things.
0:09:20 They’re made usually as action cameras, right?
0:09:21 Like you can go skiing and so on with them.
0:09:22 Yeah.
0:09:24 But they don’t have great GPS in them.
0:09:25 Oh, interesting.
0:09:27 And basically because you don’t need it, right?
0:09:31 Like if you go, like, somewhere mountain biking, it’s typically open sky.
0:09:34 It’s not like a dense urban environment with big skyscrapers that reflect GPS.
0:09:43 So they’re decent for what people usually use GoPros, but they’re not the meter-level precision we need for high-density urban center mapping.
0:09:45 So that was one problem.
0:09:53 And the second problem with GoPros was they just have no tooling that they can operate 24-7, drivers upload automatically data.
0:10:06 So they’re not made, like usually made for like a one-two-hour recording and not this heavy-duty recording for the back of our motorbikes for 12 hours in blistering heat in Southeast Asia, tough rain and so on.
0:10:12 So that’s basically why both of these cameras that existed didn’t fit the needs that are very harsh environments.
0:10:16 So would the mapping camera have worked, but you didn’t want to spend 20 grand per camera?
0:10:19 That was presumably the simple problem with the mapping camera?
0:10:21 Yeah, 20 grand was the problem.
0:10:23 I mean, the other problems were they’re also quite heavy and bulky.
0:10:29 So those cameras are close to 10 kilograms, which is not super easy to operate.
0:10:32 So they were too bulky and too costly.
0:10:36 So what, do you just get on a plane to Shenzhen and tell them what you need?
0:10:38 Like, how do you build your own camera?
0:10:41 So this was really serendipity.
0:10:48 We actually had a small team in Shenzhen already building our battery swap lockers for our scooters.
0:10:59 And back then I talked to our CTO at that time and he was kind enough to say, like, hey, Philip, we really need to find something next for this team to do so you can have them all.
0:11:00 They can build your cameras.
0:11:02 Oh, amazing.
0:11:03 We’re very lucky.
0:11:05 So tell me about the camera they came up with.
0:11:09 Yeah, so the first camera that we built, we called it a CARTA cam.
0:11:13 So like after like CARTA in like some language means like map.
0:11:19 So basically those cameras where the first version was mounted on the helmet of drivers predominantly.
0:11:23 So it was super light, 250 grams, mounted on the helmet of drivers.
0:11:28 And it would basically be a camera with an AI chip and a really highly accurate GPS.
0:11:30 And like that’s all we needed.
0:11:38 And so they would just mount them on their helmets and go about their day and just at the end of the day, take it home, take it on their Wi-Fi, upload the data.
0:11:41 And yeah, we would get data from tons of cameras.
0:11:44 And then you made a CARTA cam 2, right?
0:11:46 Why’d you tell me about CARTA cam 2?
0:11:48 Yeah, we made a bunch of iterations of our hardware.
0:11:54 So CARTA cam 2 is our latest generation, which is basically a 360 camera.
0:12:02 So it has basically four camera lenses in all directions, has also like more advanced sensors, like LiDAR sensors in there.
0:12:12 So we basically made the setup a lot more professional that we can capture maps in even higher level of accuracy to get things like lane level navigation for our drivers.
0:12:13 Our advanced safety features.
0:12:20 So the latest camera is basically a great iteration that has taken it just a step further to capture more details on the map.
0:12:21 And where does it go?
0:12:24 Where does the camera go if somebody’s driving a scooter?
0:12:27 Yeah, so now we’ve built a mount on the back of a motorbike.
0:12:32 So we have built actually the cameras that you can either mount them at the back of your motorbike, then it doesn’t disturb you.
0:12:35 And it’s in a pole so that it’s high, so that it can see.
0:12:38 Because the camera obviously needs to have good view, even if there’s cars around.
0:12:41 So we have built a special mount on the back of a motorbike.
0:12:45 We have mounts also for cars, so some drivers also mount them on cars.
0:12:48 And then we have a backpack that you can carry if you want to map indoor.
0:12:54 The other thing that’s really important in Southeast Asia is malls are gigantic.
0:12:57 You have like these malls, which it takes you 15 minutes to walk around.
0:13:04 And so people will get a delivery to whatever, the noodle shop in the middle of the mall, and you got to put that on the map?
0:13:09 Well, the other way around, people get a delivery from the noodle shop in the mall.
0:13:10 Of course, of course, of course.
0:13:13 And our driver needs to find their way there.
0:13:18 And if they get into the wrong entrance of the mall, it might cost them 10, 15 minutes extra to walk back and forth.
0:13:26 So we need to precisely not only tell them go into the mall, but we also precisely need to tell them where to park their motorbike.
0:13:30 So what we emphasize when we build our maps, that we build them like end-to-end.
0:13:32 And we say, here’s where you park your motorbike.
0:13:33 Here’s where you walk.
0:13:35 Here’s where you pick up the noodles.
0:13:37 And then you can do the same in reverse.
0:13:40 How many of these cameras do you have out there, like, today?
0:13:43 How many people are driving around mapping with these cameras today-ish?
0:13:47 By end of this year, we have about 20,000 cameras in the field.
0:13:54 And to give you a sense, like, professional mapping companies in Southeast Asia, to my best of my knowledge, have about tens of cars.
0:13:57 But not tens of thousands, tens of cars.
0:14:05 So does that mean they have basically ceded it to you, like, that you have won the mapping Southeast Asia fight?
0:14:07 I think it’s still so, so early.
0:14:08 Oh, fair. Interesting.
0:14:10 I mean, there’s still so many things we want to do.
0:14:11 Interesting. I love that.
0:14:16 So let’s talk a little more about what you’re doing now, and then let’s talk about what you want to do.
0:14:19 So you have an incredible amount of data coming in now, right?
0:14:28 I mean, if you have 20,000 cameras driving around every day, like, that is a wild amount of input.
0:14:30 What are you doing with all that data?
0:14:32 I mean, there’s basic mapping, and I’m sure you’ve got that.
0:14:33 But what’s the, like, next level?
0:14:35 Presumably you have AI, you said you have LiDAR.
0:14:38 Like, tell me all the things you know because you have all this data.
0:14:39 You’re right.
0:14:45 I think the basic vision that we have is get an accurate representation of the real world in real time.
0:14:49 That’s the goal of mapping.
0:14:50 In real time is crazy.
0:14:51 In real time is crazy.
0:14:53 Yeah, that’s wild.
0:14:54 But fine, that’s the goal.
0:14:57 Just, like, right now, what do you know?
0:15:04 Yeah, so a bunch of things is, I mean, first of all, everything that you need for navigation, roads, how are roads drivable?
0:15:06 Is the road safe to drive?
0:15:09 Does it have a lot of potholes and things like that?
0:15:10 And then signage, obviously.
0:15:11 Is it a one-way road?
0:15:12 Is it not a one-way road?
0:15:15 Can you warn me about a particular pothole?
0:15:18 Can you be, like, be careful there’s a pothole on the left coming up?
0:15:19 Try the right lane?
0:15:21 And we’re doing actually very cool things.
0:15:29 We’re doing right now, we’ve already launched safety navigation that tells you, A, is the road safe to drive based on potholes?
0:15:30 And does it have street lighting?
0:15:31 Is it safe?
0:15:35 Because if the road at night is well lit, it’s a lot safer to drive than it’s not.
0:15:39 So that’s kind of like a practical use case that we already can do with that.
0:15:39 Interesting.
0:15:49 As I’m sure you know, like, Waze in this country tells you where there’s, like, a speed camera or, like, a speed trap.
0:15:50 Do you do that?
0:15:51 Yeah, absolutely.
0:15:59 So we have an AI voice reporting that drivers can share anything, say, like, oh, the right lane of this road is closed.
0:16:03 And then it gets processed and basically used for all the other drivers.
0:16:07 So we’ve launched that and it’s been really successful as well.
0:16:12 So when you say it’s AI, does that mean the driver just says it and it gets integrated into the maps?
0:16:14 Or what does that mean?
0:16:16 Yeah, that’s exactly right.
0:16:19 There’s a feature we actually work closely together with our friends at OpenAI.
0:16:24 So the driver in natural language reports an issue and then it gets processed.
0:16:26 And if you’re not clear, we ask you a follow-up question.
0:16:28 If you say, hey, the right lane is closed.
0:16:30 And we say, like, oh, do you mean this road?
0:16:32 And then it says, driver, yeah, that’s exactly the road I’m talking about.
0:16:36 And then we process it and basically warn all the other drivers accordingly.
0:16:38 So you mentioned your friends at OpenAI.
0:16:41 It seems like good friends to have, good place to have friends.
0:16:43 I know you have a deal with them.
0:16:45 Like, what is the nature of your deal with OpenAI?
0:16:48 And more generally, what’s the work you’re doing with them?
0:16:55 So mapping is one of the key things we do together with them and use their models to improve our maps.
0:17:00 The voice stuff that I shared earlier is like one of our highlight features.
0:17:04 And in general, we are just embedding like AI in everything we’re doing.
0:17:09 And we’re having like over a thousand different AI models that we’re working on.
0:17:13 So it’s basically deeply, deeply embedded in many, many of the things that we’re doing.
0:17:15 Tell me more.
0:17:17 Like, what are some more examples of how you’re using AI?
0:17:27 So one of the latest things that I really like, that’s really cool, is the other thing that has really huge impact on our marketplace is weather.
0:17:30 In Southeast Asia, the rain is insane.
0:17:32 It is like pouring down.
0:17:34 The roads are flooding.
0:17:40 So, and for our services, for mobility, for deliveries, that has a tremendous impact on that.
0:17:46 So knowing when it rains early is super, super critical so we can adjust the marketplace.
0:17:50 We’ve launched AI-based rain detection that does a bunch of things.
0:17:55 So we deploy sensors, other sensors in cars.
0:17:59 So we have basically a device that’s called a Kata dongle that we deploy in the cars that we own.
0:18:02 And it plugs into a port in the car that’s called an OBD2 port.
0:18:04 And that reads when the windshield wiper is going.
0:18:05 Oh, genius.
0:18:11 Such a simple way, such a simple way to know if it’s raining.
0:18:13 Does the car have the windshield wipers on?
0:18:14 It’s so second order.
0:18:15 I love it, though.
0:18:16 Yeah, and how fast it is.
0:18:21 Oh, how fast the car is going because the slower it’s going, the heavier the rain.
0:18:22 Is that the inference?
0:18:26 How fast the windshield wiper is going because you can have it like this or this, right?
0:18:31 So, okay, so, and what do you do with that data?
0:18:32 That’s the input.
0:18:33 What’s the output?
0:18:39 The output is basically we know the moment it rains, we know demand will go up and supply
0:18:40 of drivers will go down.
0:18:43 So what we do, we try to activate more drivers.
0:18:47 So we would send out to all the drivers, hey, it’s starting to rain.
0:18:48 There’s a fantastic earning opportunity.
0:18:53 If you’re not working, now’s a great chance to get on the road and make extra money.
0:19:00 And then we try to actively get the supply of drivers up so that we can keep our reliability
0:19:01 at the levels we need.
0:19:03 I know people get all worked up about surge pricing.
0:19:05 Do you use surge pricing in that setting?
0:19:07 It sounds like a classic use of surge pricing.
0:19:08 Everybody wants a driver.
0:19:09 Nobody wants to drive.
0:19:10 What you need is a higher price.
0:19:11 Do you do that?
0:19:12 That’s exactly right.
0:19:15 Or like you need to motivate the drivers to come on the road.
0:19:15 Yeah.
0:19:16 No, there’s a market.
0:19:18 I’m pro surge pricing.
0:19:19 That’s a great example.
0:19:21 What’s another example of the way you’re using AI?
0:19:26 So we use AI to translate in many scenarios.
0:19:32 For example, when I go to Jakarta, I obviously don’t speak any Bahasa, but I can message our
0:19:36 drivers and we real-time translate any message and send in the chat to the driver.
0:19:39 And he can reply back to me in Bahasa.
0:19:40 And I see it in English on my side.
0:19:42 He sees it in Bahasa on their side.
0:19:45 But the more important one for me was like food menu translation.
0:19:50 So we invested quite a bit because whenever I go to Thailand and then like, I mean, I can’t
0:19:51 read any Thai script, obviously.
0:19:56 And I, in the past, I look at like some things on the menu and I have no idea.
0:19:59 Is it like, for me, always the word is ultra spicy.
0:20:00 Can I eat it or not?
0:20:02 And I didn’t even know what the item is.
0:20:06 I can look at the picture and kind of guess, but sometimes merchants don’t have pictures.
0:20:11 So we use AI to translate all these menus in all kinds of languages.
0:20:14 So that, that has been a super impactful one as well.
0:20:20 We’ll be back in just a minute.
0:20:30 Run a business and not thinking about podcasting?
0:20:31 Think again.
0:20:36 More Americans listen to podcasts than ad-supported streaming music from Spotify and Pandora.
0:20:40 And as the number one podcaster, iHeart’s twice as large as the next two combined.
0:20:43 So whatever your customers listen to, they’ll hear your message.
0:20:47 Plus, only iHeart can extend your message to audiences across broadcast radio.
0:20:49 Think podcasting can help your business?
0:20:50 Think iHeart.
0:20:52 Streaming, radio, and podcasting.
0:20:55 Call 844-844-IHEART to get started.
0:20:58 That’s 844-844-IHEART.
0:21:02 Imagine that you’re on an airplane and all of a sudden you hear this.
0:21:10 Attention passengers, the pilot is having an emergency and we need someone, anyone, to land this plane.
0:21:11 Think you could do it?
0:21:18 It turns out that nearly 50% of men think that they could land the plane with the help of air traffic control.
0:21:21 And they’re saying like, okay, pull this until this.
0:21:22 Do this, pull that, turn this.
0:21:23 It’s just…
0:21:23 I can do it with my eyes closed.
0:21:24 I’m Manny.
0:21:25 I’m Noah.
0:21:26 This is Devin.
0:21:31 And on our new show, No Such Thing, we get to the bottom of questions like these.
0:21:34 Join us as we talk to the leading expert on overconfidence.
0:21:41 Those who lack expertise, lack the expertise they need to recognize that they lack expertise.
0:21:45 And then as we try the whole thing out for real.
0:21:46 Wait, what?
0:21:47 Oh, that’s the runway.
0:21:49 I’m looking at this thing, see?
0:21:56 Listen to No Such Thing on the iHeartRadio app, Apple Podcasts, or wherever you get your podcasts.
0:22:02 So I’m curious about GrabMaps Enterprise, right?
0:22:08 Like, you’re selling something related to your maps to big companies, right?
0:22:11 To Microsoft, to Amazon, to governments.
0:22:11 Like, what?
0:22:12 Tell me about that business.
0:22:15 That was quite an interesting one, right?
0:22:19 Like, we would have never imagined early on when we built our maps that this business would be created.
0:22:25 But we got people, like, once we publicized our maps, and people have seen that they work generally quite well,
0:22:28 we got people approaching us saying, hey, can we use them as well?
0:22:33 And that was, like, the genesis of the enterprise business.
0:22:40 And then we started working very closely with AWS, where any developer on their platform can use our maps now.
0:22:44 So we have over 100 different developers already using it.
0:22:52 We partnered with Microsoft, as you said, and a bunch of other large tech companies that in Southeast Asia started using our maps,
0:22:59 because they’ve seen, I mean, all these things that I shared earlier, that it just really serves their needs in Southeast Asia a lot better than anything else out there.
0:23:03 What’s an example of something someone has built, you know, on top of it?
0:23:15 A really cool startup that I’ve seen that used our maps, what they do, they go from merchant to merchant and collect, like, the old oil that they use for cooking, and then basically recycle it.
0:23:19 So they send somebody around going to all these merchants collecting that.
0:23:26 And in the past, and a lot of these merchants are, like, small neighborhood, mom-and-pop shops, you know, all these, like, little side roads and alleys.
0:23:31 And they had a hard time for the person going around collecting all of the stuff, finding that.
0:23:39 And that’s one of these examples where they use grab maps to make the navigation of the person going around and collecting all of that a lot easier.
0:23:50 And there’s many of these kind of stories where people use it for things that are very similar in this kind of finding the last mile has been really, really hard before.
0:23:54 So I’m curious what you’re working on now.
0:23:58 Like, yeah, what are some of the things you’re trying to figure out that you haven’t figured out yet?
0:24:03 I think the one thing that we’re really passionate about is solving more of the indoor problem.
0:24:05 So that’s one thing that we’re really mapping more.
0:24:09 So we have a decent amount of malls already mapped, but there’s still so much more to do.
0:24:12 So hopefully we can, like, find that.
0:24:21 Whenever you go into these malls, we can exactly help you to find whatever you’re looking for, a specific store or a general kind of, like, shop or so.
0:24:24 So that’s one thing that I really want us to crack.
0:24:34 So is the general shop idea that, like, if I’m just not working for Grab, I’m just in the general public and I’m like, I want to buy a pair of shorts.
0:24:37 I just type that into Grab Maps and Grab Maps tells me where to go.
0:24:38 Is that what you’re thinking of there?
0:24:41 Yes, but we always think about it with a hyper-local twist.
0:24:52 And what that means, as an example, let’s say, in Indonesia, a large part of the population is Muslim, which means they generally eat halal.
0:24:56 And this is, like, all the mapping platforms, they don’t support that.
0:25:02 You can, of course, for every search query, you can say restaurant halal, this halal, this halal, but you cannot make it part of your user profile.
0:25:06 But it’s not something that you switch every day.
0:25:10 Either your preference is halal or it’s not.
0:25:18 And none of the mapping platforms support putting in your profile, I only want halal restaurants because I don’t eat anything else.
0:25:26 And those are the kind of things that we see in Southeast Asia we need to really solve because, again, nobody else would solve those kind of problems.
0:25:34 So capturing this data accurately and knowing all these details, those are the kind of things that we really put a lot of emphasis on.
0:25:38 Are there technical problems you’re working on?
0:25:48 Like, are there things where you haven’t figured out the right tech or where you need to build something or where AI models aren’t quite where they need to be, but you’re trying to push them?
0:25:50 I think what we’re trying to do is what I said earlier.
0:25:53 Like, you want to have a real-time, accurate model of the world.
0:25:55 Yeah, so let’s talk about that.
0:26:00 Like, real-time is a crazy phrase in that context, right?
0:26:04 Like, when you say real-time model of the world, what are you thinking of?
0:26:09 Yeah, I mean, a simple use case would be know where there’s parking right now.
0:26:10 Yeah.
0:26:14 Not like where there’s parking in general, but let’s say a roadside parking.
0:26:24 If you had another grab car driving past and say, hey, 15 seconds ago there was a freak parking spot on the right, that’s extremely useful.
0:26:25 That is extremely useful.
0:26:31 I know people in Brooklyn and San Francisco who would love to have that functionality.
0:26:38 Yeah, I think for us, really, the goal is always what we know is like shaving off seconds of every delivery, right?
0:26:41 Like, that’s what really, really makes a Delta.
0:26:49 I think that I really loved Steve Jobs’ old mental model when he convinced people to make the Mac boot, like, 10 seconds faster.
0:26:57 And he said, like, well, if you take this, I don’t remember the exact math for him, but he said, like, if you make the Mac boot 10 seconds faster, there’s 50 million people using it.
0:27:04 Every year you save, like, I don’t know, it’s like 10 lifetimes of people’s lifetime waiting for the Mac to boot, something like that.
0:27:15 And for us, right, like, across, I mean, across, I calculate and I share this always with the team, across a billion deliveries, 2.5 seconds saved across every delivery is roughly one lifetime that you can save.
0:27:27 So any second we can shave off by getting cars to park faster, by getting motorbikes to park faster, park at the right space at the right time, that’s kind of the problems that we’re really passionate on solving.
0:27:30 How long does it take you to do a billion deliveries?
0:27:32 Is it a billion per what?
0:27:38 We don’t publish exact data, but it’s the order of magnitude of, like, a year or less.
0:27:39 Oh, wow.
0:27:39 Okay.
0:27:42 That’s a big number.
0:27:45 So how do you get real-time parking data?
0:27:49 I mean, so is the constraint now just getting more cameras to more drivers?
0:27:51 Like, what’s the rate-limiting step?
0:27:53 Yeah, great question.
0:27:55 I think more cameras is one constraint.
0:28:04 The other constraint is doing all the smart processing on the edge, because obviously you cannot upload all these data, because that would be extremely costly.
0:28:09 And mobile networks in Southeast Asia aren’t quite that powerful that we could upload millions of video streams to the cloud.
0:28:15 So we are working a lot on what is called, like, edge AI, that we can run all these models.
0:28:19 That are powerful, but not in the cloud, but on the edge, or at least on a mix of both.
0:28:23 So in this case, is the edge the actual camera?
0:28:30 I mean, what is, like, is the dream the camera itself is doing the work and just uploading something very simple to the network?
0:28:32 That’s in many places already happening.
0:28:35 We already have quite powerful AI chips in the cameras.
0:28:42 But, I mean, of course, like, no mobile phone, no edge camera right now is as powerful as, let’s say, chat GPT with GPT-4-0.
0:28:43 Well, sure.
0:28:52 I mean, you have this sort of spectrum where you have a, whatever, a hundred million dollar data center on one end and a hundred dollar camera on the other and some things in between, right?
0:28:55 So what can you do on the camera now?
0:28:57 We do things like privacy.
0:28:59 So we blur people’s faces.
0:29:02 We blur license plates because we never want to upload this information.
0:29:05 We do weather detection, what I said earlier, with rain.
0:29:09 So we detect on the cameras if it’s raining, things like that.
0:29:12 We detect traffic signs and see if they’ve changed.
0:29:16 So we actually already run a large part of processing on the edge.
0:29:22 And then only if something has changed, then we upload some data and do a validation on the server.
0:29:26 But that already allows us to reduce what we upload by more than 90 percent.
0:29:29 Just tell the server if something is different, basically.
0:29:31 Yeah, exactly.
0:29:37 So I know, you know, we’ve talked mostly about maps, but obviously Grab is doing a lot of different things.
0:29:40 Are there other parts of the business that we should talk about?
0:29:42 The business is so fascinating.
0:29:45 So if we have time, we should talk about all of our businesses.
0:29:49 Yeah, just tell me, what’s one other sort of frontier, one other thing you’re working on?
0:29:58 I think the other thing which we’ve really invested deeply, which is also quite closely connected to our maps, is when we look at across the delivery journey.
0:30:04 The other part where a lot of time is spent is in the merchant, cooking and preparing the food.
0:30:17 And that’s an area where we spend a lot of time optimizing together with the merchants to make sure that the food is prepared in the right time, that the driver arrives in the right time.
0:30:19 So we’ve done lots and lots of cool things.
0:30:27 So, for example, we’ve built a data science model that accurately predicts at every time of the day how long the merchant needs to prepare an order.
0:30:33 So we can detect the busyness of the merchant and say, if you know, oh, the merchant has gotten a lot of orders.
0:30:40 Normally they take to prepare, they prepare to like a banh mi in like three minutes, but when they’re very busy, they take seven minutes.
0:30:45 And the merchants don’t want that our drivers crowd their store and wait in troves.
0:30:53 So we typically allocate the driver only when we know, okay, the food is ready in seven minutes and we don’t allocate the driver immediately.
0:31:00 When we know, oh, there’s a driver two minutes away, we just allocate them like five minutes later so that he’s in the shop just in time.
0:31:12 And that also cuts the delivery journey, makes it a lot more pleasant for the merchant, not have like a lot of people crowd their store and allows us to offer more affordable price to the consumer because they don’t need to pay for all these minutes of the driver waiting.
0:31:17 So just all these optimizations, all these different margins where you can optimize.
0:31:26 So if we think about the whatever, five years, 10 years out, when you think about this sort of medium to long term, like what’s your dream?
0:31:28 Like how’s it work?
0:31:28 What’s going on?
0:31:31 I think like for me, the world is changing so fast.
0:31:35 Five to 10 years prediction is really, really hard with all the AI advancements.
0:31:36 How far you want to go?
0:31:37 Two years?
0:31:38 Four years?
0:31:38 You tell me.
0:31:40 Like what’s your, just give me a dream for the future.
0:31:44 I mean, the obvious one that I’m very passionate about is robotics.
0:31:47 They make so much sense in our marketplace.
0:31:56 So if you can add robotics to our marketplace, that will be a huge, huge change and something we’re actively working on to make happen.
0:32:00 So that’s, I think, probably the thing I’m most passionate about to change.
0:32:02 Tell me what you’re actively working on with robotics.
0:32:06 We haven’t shared much, which we do on the delivery side.
0:32:09 But for example, I mean, things like autonomous cars.
0:32:16 We’ve signed an agreement with a bunch of like autonomous car providers to come with us to Singapore and so on.
0:32:18 So we’ll do a bunch of things in that space.
0:32:23 But basically, you can imagine that it will be in many, many parts of our delivery chain.
0:32:26 You need to build an autonomous scooter based on what you’ve told me, right?
0:32:31 I feel like the analogy to the map story is the autonomous scooter that can go down all the alleyways.
0:32:32 Great.
0:32:35 If you ever want to have a job in our product team, feel free to join us.
0:32:38 Build autonomous scooters with us.
0:32:38 Because you’re exactly right.
0:32:41 We need to build products that work in our region.
0:32:43 So that’s exactly the right mindset that we’re trying.
0:32:48 Like what we say always, we want to build hyper-local products that work in Southeast Asia.
0:32:53 We’ll be back in a minute with the lightning round.
0:33:04 Run a business and not thinking about podcasting?
0:33:05 Think again.
0:33:10 More Americans listen to podcasts than ad-supported streaming music from Spotify and Pandora.
0:33:14 And as the number one podcaster, iHeart’s twice as large as the next two combined.
0:33:18 So whatever your customers listen to, they’ll hear your message.
0:33:22 Plus, only iHeart can extend your message to audiences across broadcast radio.
0:33:24 Think podcasting can help your business?
0:33:25 Think iHeart.
0:33:27 Streaming, radio, and podcasting.
0:33:30 Let us show you at iHeartAdvertising.com.
0:33:32 That’s iHeartAdvertising.com.
0:33:36 Imagine that you’re on an airplane, and all of a sudden you hear this.
0:33:44 Attention passengers, the pilot is having an emergency, and we need someone, anyone, to land this plane.
0:33:45 Think you could do it?
0:33:52 It turns out that nearly 50% of men think that they could land the plane with the help of air traffic control.
0:33:56 And they’re saying like, okay, pull this, do this, pull that, turn this.
0:33:57 It’s just…
0:33:57 I can do it with my eyes closed.
0:33:58 I’m Manny.
0:33:59 I’m Noah.
0:34:00 This is Devin.
0:34:05 And on our new show, No Such Thing, we get to the bottom of questions like these.
0:34:08 Join us as we talk to the leading expert on overconfidence.
0:34:15 Those who lack expertise, lack the expertise they need to recognize that they lack expertise.
0:34:19 And then, as we try the whole thing out for real.
0:34:20 Wait, what?
0:34:21 Oh, that’s the runway.
0:34:23 I’m looking at this thing, see?
0:34:30 Listen to No Such Thing on the iHeartRadio app, Apple Podcasts, or wherever you get your podcasts.
0:34:34 Okay, let’s finish with the lightning round.
0:34:44 So, I read on your bio, you write that you mostly travel between Singapore, San Francisco, Berlin, and Cluj.
0:34:48 Tell me about Cluj, the other three I’m familiar with.
0:34:50 Cluj, I don’t know anything about.
0:34:51 How’s that wind up on that list?
0:34:56 It’s a city in Romania, in the heart of Transylvania, actually.
0:35:01 And we have a small engineering center there in our maps team.
0:35:07 So, I’ve spent a lot of time there because for my own startup, that’s how I wound up in Cluj originally.
0:35:13 For my startup, Scobler, which was a maps and navigation startup, I built an engineering team in Cluj.
0:35:20 So, I’ve been going to Cluj and working with engineers there for the last 15 years or so.
0:35:21 What’s Cluj like?
0:35:22 It’s fun.
0:35:27 It’s a student town, so high energy, young population, very smart.
0:35:30 The computer science department, very good.
0:35:33 They used to clone the IBM mainframes for the Soviet Union.
0:35:35 So, a bunch of hardcore engineers.
0:35:38 You also write in that bio, you wrote,
0:35:42 In Berlin, I experienced some of the best parties ever.
0:35:44 Tell me about a Berlin party.
0:35:46 That was prior life.
0:35:50 That was before I moved to Southeast Asia.
0:35:55 But Berlin was a fun, fun, fun journey for, I lived there for five, six years, maybe.
0:35:58 You didn’t tell me anything about any party.
0:36:03 So, and then you also say you lived and worked in Singapore and Berlin and San Francisco.
0:36:10 And I’m curious, like, how is sort of, whatever, professional culture, work life, different in those places?
0:36:12 What are the sort of striking differences?
0:36:15 I mean, Germany is extremely direct, right?
0:36:16 Like, that’s where I’m originally from.
0:36:17 I’m originally German.
0:36:27 And Germans are known to be super, super direct, which in Southeast Asia doesn’t work super well if they’re not accustomed to it.
0:36:40 So, I think for me, the thing that I needed to adjust most is to become a lot more indirect, to become a lot more, like, have those conversations in, like, smaller groups, not in front of everybody.
0:36:46 So, I definitely had to adjust my style quite a lot when moving around.
0:36:47 But I found that always incredibly fun.
0:36:49 I always loved learning about new cultures.
0:36:54 So, after some adjustments, I really, really like it here.
0:36:55 That is super interesting.
0:37:08 Was there, as you were saying that, I was wondering, like, was there a particular moment when you realized, oh, I’m not behaving correctly in this cultural context?
0:37:09 Yeah.
0:37:11 I mean, this predates Grab.
0:37:13 So, this was when I was first doing business in China.
0:37:23 So, I sold my startup to a Silicon Valley company, but I also got a 200-people team in China report to me, which was the first time that I ever managed a team in China.
0:37:31 And I was extremely surprised why they wouldn’t tell me about all the mistakes, all the things that I did wrong.
0:37:33 And I said, like, that’s very unusual.
0:37:37 Normally, I get a lot of, like, pushback from engineers.
0:37:41 And then I started, like, asking people, why don’t they do this?
0:37:46 And then people said, like, well, it’s kind of rude in a public forum to say the boss is wrong.
0:37:55 And then I realized, like, oh, that’s why I just need to, like, change how I’m asking questions, how I’m managing teams.
0:37:57 And that’s when I first managed teams in China.
0:37:59 It took me quite a while to figure out, honestly.
0:38:08 And how does San Francisco fit in relation to both working in Berlin and working in, you know, China, Singapore?
0:38:11 Where’s the U.S. on that continuum?
0:38:18 The U.S. for me, I think the thing that I really loved in San Francisco was just the craziness of ambition.
0:38:26 When I spent, like, April back in SF for a month to, like, Vibecode and hack on a bunch of hobby projects.
0:38:30 And, like, you go to every random coffee shop and there’s somebody who said they work on a billion dollar idea.
0:38:32 And they’re just starting.
0:38:33 They have nothing yet.
0:38:36 They’re just like, okay, I’m going to make it big.
0:38:46 So, I think that that ambition and that willingness to openly declare it, even if people know it’s super unlikely, there’s not thousands of people who start billion dollar companies.
0:38:50 But in the Bay Area, there’s thousands of people who say they will.
0:38:57 And this conviction and this, like, optimism, I think that was, for me, one of the most striking things in the U.S.
0:39:00 and that I still love the energy whenever I go there, basically.
0:39:10 Philip Kondal is the Chief Product Officer at Graph.
0:39:14 Please email us at problematpushkin.fm.
0:39:17 We are always looking for new guests for the show.
0:39:21 Today’s show was produced by Trina Menino and Gabriel Hunter-Chang.
0:39:26 It was edited by Alexander Gerriton and engineered by Sarah Brugger.
0:39:30 I’m Jacob Goldstein, and we’ll be back next week with another episode of What’s Your Problem?
0:39:43 Why are TSA rules so confusing?
0:39:56 Why are you screaming at me?
0:39:57 I can’t expect what to do.
0:40:00 Now, if the rule was the same, go off on me.
0:40:01 I deserve it.
0:40:02 You know, lock him up.
0:40:08 Listen to No Such Thing on the iHeartRadio app, Apple Podcasts, or wherever you get your podcasts.
0:40:09 No such thing.
0:40:12 This is an iHeart Podcast.
0:40:15 you
Philipp Kandal is the chief product officer of Grab, an app that serves several countries across Southeast Asia. Two of Grab’s main businesses are delivery and mobility – like a combination between Instacart and Uber. And maps are at the core of its business.
On today’s show, Philipp talks about improving online maps for places like Southeast Asia, where streets are often winding, narrow, and harder to access than those in the US and other developed countries.
See omnystudio.com/listener for privacy information.