01:30:59 · Zach Rash, John Coogan, Mark Gurman
[01:30:59] S8: But, yeah, for anyone that doesn't know, I mean, this is this is what they look like. This is a miniature three d printed version of it. But Yeah. It's a purpose built robot, designed for, moving goods around our cities. [01:31:11] S1: Yeah. And you're expanding to The UK. How did that come together? [01:31:16] S8: Yeah. So we... Here in The US, we work with Uber Eats and and DoorDash. DoorDash had recently acquired Deliveroo, which is based over in London. [01:31:25] S1: Sure. [01:31:25] S8: And so we've been working with DoorDash for, I think, almost four years at this point. We have a really close partnership, and we wanted to bring it across the pond and bring it to London. And so we announced just the other week a kind of multi year bidirectionally exclusive partnership to bring Cocos all throughout The UK and bring it to some real scale. [01:31:45] S1: How do you how do you think about... Or how is... How are these sidewalk robots different from the delivery network that you might see on DoorDash with people in cars? Like, is there a specific robot assigned to an individual restaurant because they're just going back and forth, or is it more like a fleet, a network, it's driving around, and it might stop by any random, restaurant, and the restaurant just has to sort of see the robot and say, oh, I should put this there. How does that work? [01:32:12] S8: Yeah. They're... So they're shared. So they'll... We'll get offers from the different platforms. We also work directly with some retailers. Mhmm. [01:32:18] S7: So we get
01:32:18 · Mark Gurman, Zach Rash, John Coogan
[01:32:18] S7: So we get [01:32:19] S8: a bunch of offers coming to our system, and then the robot will, you know, go somewhere to pick up an order, deliver it to the customer, and then go on to pick up the next one. So just in LA, I mean, we have probably a few thousand merchants across grocery, food, retail Yeah. That use the robots. [01:32:34] S1: Has there been any movement on getting the restaurants to help do some of the charging and some of the infrastructure sort of, like, use their back rooms for space, or do you have to do sort of what Lime and and, Bird did where you have trucks driving around servicing things? Like, what's been the solution there? [01:32:55] S8: You had some mix. I mean, ideally, you don't want to be going around swapping batteries because it means you're, you know, doubling... At least doubling your battery CapEx. It takes forever to drive across a place like Los Angeles to manually swap the batteries, and there's not that many deliveries happening at four in the morning. [01:33:09] S7: Yeah. [01:33:09] S8: So ideally, they go back to merchants. The merchants plug them in, and then they take a few hours of downtime. Right? So you run them twenty hours a day, and then maybe you have four hours of downtime to do cleaning and charging. We do work with some of our merchants to do this. We work with a bunch of kinda ghost kitchen companies that also are a great fit at actually charging these and helping us do some of the maintenance. And then some of our big retail partners, they have a bunch of space. They might have twenty four seven staffing
01:33:09 · Zach Rash, John Coogan, Jordi Hays
[01:33:09] S8: So ideally, they go back to merchants. The merchants plug them in, and then they take a few hours of downtime. Right? So you run them twenty hours a day, and then maybe you have four hours of downtime to do cleaning and charging. We do work with some of our merchants to do this. We work with a bunch of kinda ghost kitchen companies that also are a great fit at actually charging these and helping us do some of the maintenance. And then some of our big retail partners, they have a bunch of space. They might have twenty four seven staffing [01:33:34] S1: Mhmm. [01:33:34] S8: In which case, we can kinda work with them to drive some incremental revenue back to their store in exchange for doing some of the fleet management. So you'll never have a 100% in any one of these categories, but the idea is to kinda work with our partners to reduce the overall cost of operating the fleet. [01:33:48] S0: What's the bottleneck to going from a thousand active robots? Sounds like this new in The UK will will increase that. But how how do you go to a million? Like, what what are the kind of things that need to happen?
01:34:00 · Zach Rash
[01:34:00] S8: Yeah. So, you know, we're gonna bring... Coco's in about 10 cities today across The US and and Europe, and next year, we should be in between 40 to 50 cities across even more continents. And, you know, the first step is regulatory. You know, you need to create new policy for this. This is the new class of of thing. It's not a vehicle. It's not a person. It's it's a robot. And so you need to define laws around how the robots operate in our cities. So you work with different either national transportation agencies or at the state level or at the city level, depending on where we're going, introduce policy, and then ramping up manufacturing of our of our new robots. So this is Coco two. It's the next gen robot that you'll start seeing rolling around the cities in the coming weeks. So we're planning on making about 20,000 of those for 2027. Mhmm. So you'll see a pretty large step up in the fleet size across a lot of new a lot of new territories, and then the path to a million feels pretty inevitable from there. Mean, I you look like, you know, a platform like DoorDash probably has 10,000,000 drivers or something like that working working on the platform in The US, and the volumes on these platforms is still growing 20 to 30% a year. So you just compound this out. I mean, look at the at the end of the next decade, right, at the end of this... In the next ten years, you're gonna have you're gonna have to have, you know, 50% of The US working population, you know, driving for these platforms. So you you need to find a way of increasing supply without getting human drivers because you're just gonna run out of people.
01:35:31 · John Coogan, Zach Rash, Jordi Hays
[01:35:31] S1: Awesome. A lot of people in tech are worried about x risk these days. Hypothetically, where's the vulnerable point if I need to disable this with, like, a five five six round? Where would I shoot it? Can't tell [01:35:45] S8: you that, though, but I [01:35:46] S1: Oh, wow. He's on the side of the robots. This guy's traitor to humanity. [01:35:51] S8: Crazy. It... What's funny is that [01:35:54] S0: You need, like, a dead man switch. Okay. Let me test it out, bro. [01:35:57] S1: Send the delivery over. Send it over. You wanna test [01:36:00] S8: it out. Japan. Sick. In Japan, you actually have to have a big red kill switch button on the robot. [01:36:05] S1: Oh, okay. [01:36:06] S8: Which I feel like would uniquely only work in Japan. [01:36:09] S0: Yeah. Because otherwise, in the back emergency. Of [01:36:12] S1: Of course. But separately, jokes aside, I I have no doubt that, that you are feeling the benefits of general model progress and AI progress. Like, just take an image, identify if there's someone there, if you can go straight on this road. All that's probably happening. But what do evals, what do benchmarks actually look like? What are you measuring, and what are you seeing in those? Like, how do you know that the latest iteration of your software, of your AI stack is actually improving the experience? It feels like we sort of feel the failure mode. Like, the hallucination is like the viral video on Instagram of the Coco robot getting stuck on the curb. But how are you measuring that? What's progress been like?
01:36:12 · John Coogan, Zach Rash, Amiram Shachar
[01:36:12] S1: Of course. But separately, jokes aside, I I have no doubt that, that you are feeling the benefits of general model progress and AI progress. Like, just take an image, identify if there's someone there, if you can go straight on this road. All that's probably happening. But what do evals, what do benchmarks actually look like? What are you measuring, and what are you seeing in those? Like, how do you know that the latest iteration of your software, of your AI stack is actually improving the experience? It feels like we sort of feel the failure mode. Like, the hallucination is like the viral video on Instagram of the Coco robot getting stuck on the curb. But how are you measuring that? What's progress been like? [01:36:56] S8: Yeah. Well, it's it's interesting because I think in robotics, [01:36:59] S9: I mean, you see a [01:36:59] S8: lot of demos of folding clothes. And then in the self driving car world, you have a lot of road driving. Right? Yeah. In a very rule based environment. Yeah. The interesting thing about city sidewalks is they're the most complex, Yeah. Dynamic not rule based environment in the world, and every city is different. [01:37:15] S1: Yeah.