Live: From half-marathon race to real-world applications

Añadido: 2026-05-05

[00:00:39]Hello and welcome to this special edition of Hot Take coming to you as a live video podcast. This is the first time we're doing this format and we are doing it live but you can always tune in on all platforms where you follow us or get your podcast. I'm your host today.

[00:00:56]Now I'm sure you all heard the buzz.

[00:00:58]China's human and robots now runs faster than humans. Much faster. Over the past weekend at the 2026 Beijing Etown Half Marathon, top performing robots completed the course in around 50 minutes. A major leap from last year's best time of about 2 hours and 40 minutes. So, what are these races actually testing? And what are the key technological leaps behind them? and how close are robots to operating autonomously in the real world. Today, I'm joined by my two of my colleagues who were at the front lines of the event. Welcome to the show. Now, let's get started with a quick hello and a self introduction and also a key line from the key takeaways of the event, please.

[00:01:46]>> Okay, who should go first?

[00:01:48]>> Maybe.

[00:01:49]>> Oh, ladies first.

[00:01:50]>> Thank you for um it's an honor for me already. um for enjoying this um event, this talk show. And so my name is John Shan Huay. I'm um anchor at CGTN French channel and uh I'm also very interested in the AI and robot development in China especially around the world and also the cooperations between China and French speaking world. So this time I got chance to be there um on site on the marathon and I bring also some of my observations. One word was I was really impressed by the speed of the Chinese robot this time that we see on the marathon, the semi marathon that we encountered and observed with great attention.

[00:02:32]>> Wonderful. And we're very lucky to have you to bring us more global perspective.

[00:02:36]Welcome. And now Chenzy.

[00:02:38]>> Yes. Well, it was a crazy day right at Sunday. Anyway, I'm Yang Shani. I'm a CGTM reporter and I've always loved tech. I've always loved AI, robotics, but uh it was nothing like seeing that in person. Right. I was at the starting line, but you were at the finishing line. You were in the middle, right?

[00:02:58]>> We completed the race.

[00:02:59]>> Yes, we completed. We can talk more about that later in the show. Uh but seeing those robots running alongside with all the human marathoners like most of you probably didn't know that uh the humans about 12,000 uh 1,200 of them were running alongside the robots and it was a weird and exciting moment when you witnessed the humans cheering on the robots on the other lane. That was a a very electrifying uh atmosphere that we will see more of in the future. So I'm really looking forward to talking more about it.

[00:03:33]>> Yeah. Yeah. So before we dive deep into the discussion, how about we go back to the exciting moment of human running alongside robots. So let's have a recap with the video from from you from Chenzi for humanoid robots in Beijing. You can hear the human marathoners on that lane cheering on the robots on this lane.

[00:04:03]It's the second edition of the event and nearly 40% of the contestants are no longer remote controlled. They are running on their own algorithms. Some of the robots are built for speed, of course. Others are designed to move like us with a gate that looks almost human.

[00:04:20]And then you've got the endurance types that are optimized to go for longer without needing a battery swap. These robots are competing for prices across different race categories. In the future, they'll be expected to serve different specialized functions in the real world. It's even more impressive when you consider that in the same race just last year, many robots couldn't navigate on their own and some failed spectacularly.

[00:04:46]And now though, it makes you wonder, give it a few more years, we might just see them running into burning buildings.

[00:04:54]Human saving robot heroes.

[00:05:00]>> Now, my colleague Jachi is at the finishing line. Let's go there.

[00:05:05]That's right, Chung. Excitement is building here at a finish line. It's about 8:00 a.m. now on Sunday morning, but you can see it's packed with spectators all eager to witness which robot will be the fastest to complete a 21 km race.

[00:05:29]One of the leading robots now making to the finish. The live screen behind me shows a remarkable leap in performance.

[00:05:35]The best performer this year complete the race in under 15 minutes without the 1.2 coefficient calculation compared to last year's 2 hours and 14 minutes. That would make a record-breaking through.

[00:05:45]And China's humanoid robotics are really advancing fast and becoming faster and more stable and autonomous.

[00:05:54]Honor's lightning robot takes the title.

[00:05:56]Beyond the victory, a bigger question emerges. What is the real value of pushing robots to run faster?

[00:06:05]We can transfer key technologies developed in competition like liquid cooling into industrial applications.

[00:06:12]>> It's similar to how racing helped drive innovation in the early automotive industry.

[00:06:18]>> International teams also underscore the growing role of global collaboration through events like this. I really enjoyed my uh collaboration with uh one of the university of China's and I'm looking forward to uh making some further and uh I hope we can uh create more opportunities like these in uh the future.

[00:06:37]>> Industry insiders say the path to commercialization may be closer than many expect.

[00:06:46]The timeline for mass production, deployment, and delivery is continuing to shorten. Once we move into real world applications, development accelerates, creating a flywheel effect. Wider use drives cost down, and lower costs drive wider adoption.

[00:07:00]>> Time and time again, events like this robot marathon show is more than just a race. It's a real world testing ground for algorithms, hardware, and computing power.

[00:07:11]Yangi, CGTN, Beijing. Wow. So yeah, we had many reporters uh witnessing at the event and just like we said, I was at the finish line, you were at the >> starting line, >> starting line and you were somewhere in the middle and there were many turns, right? So uh I know you two are were covering this event for the first time.

[00:07:31]>> So tell us um yeah, how did you feel what you what you saw?

[00:07:36]>> Um really really happy to be there. It was very early because we arrived there I think you too, right? I arrived there at 6 something um to get prepared and then also to see how people local people are very interested on this event. I saw especially moms bringing their children here. Um and then everybody are cheering there and especially when I was in the crowd I can see I can also hear the curiosity that they have to different robots because you see that the you have different kind of robots of different colors of different heights alo of different um there are many tall ones there also very small ones who coming around and then um I think it's really it's a good good time it's a showtime of the Chinese technology and it's also showtime for educating the younger generation how to get used to and to love this new technology and especially when we were having this marathon. There are also other activities that were organized alongside of the road and then you can you get like robot dogs and human robots who are together to giving a show together and so that's something that's really about event about um some very beautiful presentations of how to coordinate different robots together and >> there was also a robotic police that I I saw opposed from from other uh WeChat accounts directing traffic right see how like robots are getting into our lives.

[00:09:00]>> Exactly. And you have also time that these robots can come down onto the lane and to do some interactions with kids there. So I think it's really interesting to see especially I think the people who live there in Etown they are impressed by um the frequency that in their life already because we know that today we have the um autonomic cars and taxis who are available in um Eron in this Etown place and um so this is what impressed me the most and I mean exactly also we've talked about the speed and there's many things that I got really impressed this time especially for the technology that was used and also improved especially for this long duration thermal technology this management and also the real-time autonomous navigation and this energy optimized control how to get a robot to run up to 50 minutes up to uh 90 minutes so that's a lot of work behind it and we can also see how China has upgraded its technology in robotic industry during this past year only one year and you see the progress That's really impressed me a lot.

[00:10:08]>> That's right. And Ji, what about you? I I I saw your reaction in the >> in the in the video. I loved it. Like like that this is the frequency of speed how you turn your head, right?

[00:10:18]>> A genuine reaction, by the way. And you can tell the speed. I I rotated my head.

[00:10:24]This is how fast those robots were running. Uh so I was there pretty early in the morning uh by about six o'clock.

[00:10:31]And I was there throughout the entire process of building up anticipation until the starting gun. So right uh the first impression I got was just how much uh international attention has been given to this event because I was a depressed area and people around me were speaking all kinds of different language like Korean, Japanese, English, uh French and uh obviously it's a topic that everyone's been interested about like if you check YouTube you see you see major news outlets and and media organizations reporting on this event.

[00:11:05]Uh so it is a a a pinnacle of China's AI and robotics and manufacturing and general tech achievement being showcased here in this one event. Right. And at the starting gun, I was genuinely surprised uh of of how fast the robots were. Our camera couldn't really keep up. So, I ended up using a lot of the official footage from from the organizer because our camera was panning too fast, losing focus. I genuinely did not anticipate that because they were too fast. And probably many of you have seen uh some of the footage from last year. A lot of robots had some really epic and sometimes comedic fails.

[00:11:50]>> Yeah.

[00:11:50]>> At the at the at the starting line. Some couldn't even run. Some just fell down after like four or five steps. and some couldn't even hold their own, right? And not to mention that nearly all of them were not running autonomously, but this year about 38% were running on their own algorithm. We can talk more about that later. Uh so I was surprised that I didn't really see a lot of fails this year. There was, as far as I could remember, one fail at the at the starting line. uh one robot fail, but there was a caveat to that. When that robot fell down, the one that was running right behind it automatically stopped >> and then took a detour. Oh, >> okay.

[00:12:36]>> Right. So, that was that was the part that that that was surprising to me.

[00:12:40]Right. The robot behind it uh it was an an autonomous one.

[00:12:44]>> So, it detected an obstacle in in the front. Something that stopped in the front.

[00:12:49]>> Didn't he clap a little bit? Didn't he clap and say, >> "But that was the other one."

[00:12:54]>> Oh, that was the other one.

[00:12:54]>> The one that clapped was controlled by an engineer. So, it was an engineer that decided to clap. But, uh, the one I witnessed at the starting line uh was an autonomous one. It automatically stopped. But >> admittedly, it took a while to really calculate the situation and take the detour, but it did. But it made a stop.

[00:13:13]>> Yeah. But it did make a stop and then and then took a detour. So, that that was that was intelligence.

[00:13:19]>> Yes. And actually the the first robot that crossed the finish line was from Honor, Chinese filmmaker. It also fell right like couple hundred meters in front of the finish line.

[00:13:30]>> It Yeah. crumbled. It It just hit directly to the barrier on the side. M >> uh but then yeah it obviously needed a little bit of assistance from human but the the the workers the the the stuff just just uh pull it up from the ground and it just come back to itself and rushed through the finish line that happened at the finish line. That was something impressed to me as well. And yes this is my second time covering this event. This is the second edition. Yes.

[00:14:00]And last year I was at both the starting point, the starting line and the finish line because I got time to to go from the beginning point to the finish point you and this year the the organizer told me that you can only choose one. I was like why why can't I go both? Like I wanted to do this and last year we we did another live stream throughout the race. We did a this this called like one by one, right? We were had different reporters at different point but this year we couldn't do it and we just simply couldn't do it.

[00:14:30]>> Same here. I was intending to ask the organizer because I was in the middle and I was saying after I saw some because I was in the middle so the only thing that I I can see or just passing by all those robots and I asked them is this possible that I you move me at the end or um 30 minutes after this that you move me to the to the finishing line.

[00:14:50]They said that's impossible because the time that you're arriving in the finishing line, the whole game will be missed the show. They miss the show.

[00:14:55]>> Yeah. Because there was no shuttle bus this year from the beginning line to to the finish line and I couldn't run a half marathon to >> faster than the robots.

[00:15:05]>> The finishing line. Maybe I could do it last year because most of the robots were goofing around and didn't running didn't run pretty fast. But this year you couldn't do it.

[00:15:13]>> Yeah. One of our reporter did last year.

[00:15:15]Uh Jun, he well he didn't run. He he he rode a bicycle. I think >> that was funny. So yeah, let me share with you some of the numbers from last year to this year. As you mentioned, the failure rate and and the the teams the participating teams. So in 2025, the race was at its early stage. There were only about 20 teams participated at the starting point, right? And um only six managed to finish the race.

[00:15:42]>> Six.

[00:15:43]>> Six of them only. And so that brings the failure rate to about 70%. And this year >> the completion rate is 45%.

[00:15:54]More than 100 uh 100 teams participated with more than 300 robots in the competition and a total of 47 teams completed the race.

[00:16:05]>> And what's more interesting was that you all mentioned that they were doing this race some of them were doing this raids fully autonomously. That means um only 29 relied on remote control and as I mentioned the first one that crossed the finish line was under remote uh controlled mode but there is this performance leap.

[00:16:28]>> Last year the winning time last year the winning time was 2 hours 40 minutes and 42 second. This year the top robot finished in just 50 minutes 26 seconds and that was that was not from the first robot uh going through the finish line because you said that there is a ratio accounted to the time >> and >> so they were both from honor >> actually all first three all first five of them >> were from honor and oh just a little bit about the the the record 50 minutes for your reference for the context that's current um human half marathon world record was recently set in March 2026 and that was 57 minutes and 20 seconds.

[00:17:16]So robots >> imagine being that guy.

[00:17:19]>> Imagine being that guy who's proud of breaking the half marathon record for like half a year >> and only one month later.

[00:17:26]>> Yeah, but it's not human. It's machine.

[00:17:29]You can say it's machine but think it looks like me. Yeah.

[00:17:34]That's right. That's right. And as you mentioned like you know witnessing from the starting point >> this year humans were literally running alongside robots or actually they cannot reach the robots. Last year >> all the human renders finished their race and the robots were slowly coming to the finish line. So that's a huge leap from last year. M one takeaway is that although the finish rate is both under 50% for the first two uh years of the match but this year is I think much more significant uh given the fact that about 40% were running autonomously. It is immensely hard to rely solely on algorithm to run instead of human guidance as far as I know because uh the AI has to do a lot of instant calculations because when you're running uh a robot even though it could rely on wheels it would make it much more easier for the robot to finish the race but it has to run like a human on two feet on it pedals. So when you imagine a human running, there will be fractions of a second where you're totally airborne like both of your feet were lifting >> meaning you were in the air.

[00:18:48]>> Yeah. Right. You were lifted off the ground. So you have to calculate how you land on the ground and it is very different from uh running in a actual real world situation compared to running in the lab. Right? you you have crevices, you have pebbles, you have slight inclines and declines in the road and AI has to adjust for all of that and adjust for all kinds of emergencies and not to mention you have overheating uh uh to to deal with. Uh so it is immensely immensely more difficult this year. So it's it is an achievement in and of itself.

[00:19:22]>> That's right. That's right. And it's actually an integration of a lot of technologies together and put them all into one physical robots, right? So, so what about you, Jan? What are the other technological breakthroughs that impressed you the most? The most that press impressed me is that what we can see that it was the honors robot that uh became one of the champions and I was like wow we can theoretically this is a cell phone uh maker and how can this directly one year after that you can see this technology applicated onto um the robots the human robots and this is what impressed me the most. So >> so let's yeah let's talk about that.

[00:20:03]Yeah, the winner winning teams brand.

[00:20:07]It's a phone brand. A Chinese filmmaker who just announced manufacturing robots less than a year ago, right?

[00:20:13]>> And they took away all the champions.

[00:20:16]>> Yeah. Killed the game. Yes.

[00:20:17]>> Yeah. So, were you surprised? Because I was quite shocked.

[00:20:21]>> Personally, I Yeah, me too. I was quite shocked. And in a certain way, I was like, it's hard to imagine how a Chinese smartphone company can make a huge success in only less than one year from the time they announced that we'll start investing and do more researchers in this domain. Um, and then I an analysis a little bit because there could be some similarities and also some technologies that can be applied directly onto robots. for example, some of those little manufacturing processes and also some components that could be shared on the same time, especially for the energy use. Um, how to minimize the energy use and how to make the uh robots to do um the accurate movement on some meanwhile that is advancing and that meanwhile that is running um in a certain direction. So I think that could be interesting and in a certain way it also seems that Chinese technology has an advance compared to some other countries or some other economies is that there is a real use place and there is a news scenario here in China. You have the supply chain already especially with the built and with the manufacturing of these cell phones already in China. So you have some technologies that can be applied directly onto a machine and also onto a robot and another case is that there is also a market here the market for cell phone and also there's the concrete scenarios that can be used to testing the robots to testing um the h and also this is also interesting is that we have a media uh influence and the media coverage on this robot that naturally attract the interest and also the attention of the whole international world. So this is interesting because from a cell phone company and you can use their technology and directly this moves to the whole boom and the whole curiosity of the whole society on the robot development. So this is I find it interesting to talk about this supply chain in China and also this whole environment media environment that was built on and centered on the development of uh robot. That's a very interesting point and that makes much more sense.

[00:22:37]>> If this happened like four or five years ago, I would have been surprised.

[00:22:41]>> But now when you see that Xiaomi also another Chinese filmmaker has made one of the top electric cars in the world within just a few years, but it been development for many years, but it came out recently in recent years. If Xiaomi could do that, I'm not totally surprised that Honor could make a killer robot. Uh >> okay. But I mean it's easy to think I mean for easy for me to think that for a robot to adapt technologies from the EV industries because it's pretty much like an auto autonomous driving EV but on two pads >> but for a filmmaker how come? Well, uh there was there was a take that I saw uh that says well a car a really electric car is essentially a smartphone on wheels >> and you can imagine a robot >> all these industries are connected >> as a cell phone with hands and feet >> right uh so basically this the core idea is that you ha you will have to have a really good algorithm and chip inside it and that is the bread and butter of the cell phone industry.

[00:23:53]Uh so if Honor wants to do a robot, it has a lot of innate advantage as a phone brand. Uh as as you said, it has a huge customer base to begin with. So it has a really comprehensive manufacturing chain that can really produce at precision and at scale, >> right? So you can instantly turn out really high quality robots. uh something that you couldn't do in many manufacturing uh uh hubs around the world uh but you could do it in Shenzhen for example where honor is based you can have you can have a 3D printed uh metal component within 48 hours if you if you designed it right but it's something very difficult to get in other parts of the other parts of the world and um and the advantage of being a phone brand is that they have years of experience both for Xiaomi for in making a car and honor in making a robot is that they have years of experience developing uh the most high-end system where it can work with a AI empowered chip in a tiny form factor >> and then they have spent years of research hiring the best talents in solving how to dissipate heat in the system >> howant how to make a system reliable right all of these expertise although it's not exactly the same between a phone and a robot but it can be transferred to some extent so that I believe is part of the reason why honor has been taking the top place here >> and some of the people are talking about this using of the sensors is that for example for phones uh phones to today with the light on and also the all the cameras that are using the sensors to think and that they can have the perception of the temperature in the room and have the the light in the room and to give the best look and the best um show of the users using them. So that's kind of the similar logic with the sensors and also with the efficient power management how to make robot running the the kilometers and kilometers away and also the lightweight structural materials and as what you just mentioned about the how to make those sophisticated parts into a very tiny space. So I think these are we do have many similarities in these um the these different agencies and >> yeah I think we need to do another episode on how China's industries are connected each other and you know we should do that >> it is interesting because when I was there I saw also also this kind of manifestations um the shows that uh the robots that display together you get the dog ones the dog shaped ones you got also the human ones and I was talking with one of the um techno technological uh expert um in the enterprise who told me that they today they're trying to develop in a way a new platform a brand new platform to make different robots from different makers I mean from uh different brands and from different nationalities to communicate together.

[00:26:59]So which is that they will need to design and invent a new layer in their communication um platform. So this could be um interesting because we have just talked about um the how to make robots communicating together and have their own conversation coded conversations between them to make shows together and that he explained to me that for example the expert himself was in in the past um majoring in foreign communication and so I think in a certain way you can see some similarities here and today you know for engineering uh 30 years ago could be a robot engineer today and there are some similarities.

[00:27:41]>> Yeah. And and I think you mentioned about this communication um system between different makers of robots is getting crucially important. So that means we are building this open platform not a closed ecosystems that the the the the western countries used to lead in in technologies. Right.

[00:28:00]>> Right. So I think essentially uh so I also talked to well I didn't talk to it but I I I I I saw the uh the winning team was talking to the press uh after the race and he mentioned three points that makes their win and first you all mentioned is battery life and well the most important thing they mentioned is the cooling system because last year a lot of robots were >> overheated >> overheated and they because during the test the pre-trial test before the game, before the competition, they were doing it at night and they were all performing well. And then in the morning, everything changed. Even though the race started very early in the morning when you let the robots run and you know the motors, the joints, they they they need to operate for a long time. They all went over here.

[00:28:51]>> Yeah, it was interesting watching the robots running into like a Formula One pit stop where every kind of mechanic was running up, swatching, swapping batteries and spraying this cooling cooling agents. It was very interesting things.

[00:29:03]>> Yeah. And last year was pretty much like that. Everyone had to spray >> but they share they adopted either fans or the liquid cooling system and there was something quite interesting with honors um cooling system or thermal management that they were using this liquid cooling system and they had it at the core at the core there was a high-speed levitation pump. So for for for liquid you need to circulate the blood right just like human. So for human renders you have your skin you have your uh veins to exc to to to to get rid of your heat.

[00:29:39]>> Yes. But for robots the cooling system is at core. And this one this pump is a magnetic pump. So that means this pump doesn't have friction. So it doesn't produce excessive heat >> and all the energy it uses is to circulate the liquid for cooling down the joints. That's a stimulation with heart how the heart pumps.

[00:30:00]>> Yes. And that same technology is also used can be used in um artificial hearts and also for um computing center as well as commercial space. So that technology has a wide range of applications and coming down to that is the joints right all the joints they said were uh or motors or uh it's called is called >> actuators. Yes, that yes they were all uh self-developed or domestically developed here and and frankly this year the reason we can see that many robots on the track was not a surprise because now almost everyone can can easily well I wouldn't say easily but not not with the difficulty put put together a robot >> but the the difference or the difficulty is about how you integrate different technologies right like I mentioned the including navigation and sensing and like real time computing >> a lot about that. So yeah um and lastly it was about the battery they share a lot of them adopted this called hot plugging techn technique meaning the robots does not need to stop to swap batteries. So the robot can still maintain this momentum or motion >> um while changing batteries and one of them like each battery can last for 10 kilometers. So which means they only need to change one or two times in the middle.

[00:31:29]>> I think the holy grail in the in the next few years would be finishing the race without a battery swap.

[00:31:34]>> Yeah. Yeah. I think that's something we can look forward to in the next uh edition.

[00:31:38]>> If it is the case, that would be very very interesting and astonishing.

[00:31:42]>> Yeah. Will make them immensely more useful. And you mentioned joints. I think that was one of the uh the technical marvels that I've that have uh shocked me is that you can imagine how much pressure has been put on the joints of these robots. They've been running non-stop at a pretty fast pace throughout the 21 kilometers and their joints have been moving non-stop. If it's if a part is a component is moving non-stop, it generates immense amount of heat. So heat dissipation is a main issue. And then and then if you look at how the honor lightning robot runs, it runs in a squatting position, right? The center of gravity is quite low. So it allows it to be more stable while maintaining high speed. If a human runs like that, if I run like that, my knees would be destroyed in kilometer one.

[00:32:32]>> Yes.

[00:32:32]>> So uh and when you imagine as we as we talked about when you're running, you have airborne time. So if when your front foot touched the ground uh your foot and your joint would take about two three times uh uh the the body weight uh of of the entire Yeah. Yeah. for the impact. So it it's a test of the cooling system. It's a test of the integrity of the of the system. How how much force it can take and just how long it can maintain that performance at that point.

[00:33:05]Oh, so that explains why there's a new prize that was invented this time that who runs the best in the best way and best looking runner is that who runs really like human because you when we were we were um on the lane and to observing the whole different robots past you got some robots who's running like grandmas like this in this ways and then you have also some some ones running like babies uh but sometimes you see oh okay I'm not sure if it's still keeping balanced and the best I think They have a new prize this time for the best runner for the best uh robot um with the best performance and the most beautifully looking uh in their ways of running >> one that looks runs like human right and you see some staggering >> you see some like staggering like like you're really tired I don't know if they would feel tired but >> I don't think so >> I don't know what that is >> they don't feel tired right yeah because you know they they train their robots on uh with human data So it's naturally that they if they can let their robots that look that runs as much uh as closely as a human runner they can certainly run wild run well on on real roads. So I think that comes down to our next question is why do we let humanoid and robots running yeah >> these half marathons >> right? I mean um this year we we the the robots are running alongside humans. So they all need to make a lot of sharp turns and like I said a lot of them were completing the race autonomously and there were ups and downs according to the organizers that you know they have to go in up slopes uh with small or huge inclination and yeah to to complete the race.

[00:34:59]>> So yeah what do you what do you guys think what what's the meaning of having this race? I mean, a lot of people on on the internet would say that, okay, what's the point of having robots that are built like humans run a half marathon? Like, first of all, why do they have to look like human? Yes. If you want a robot to finish a half marathon, just put wheels on them. You have just have an electric car. They can run tens of marathons in a row, right?

[00:35:24]Yeah. But why do does it have to be designed like human? So there is a a real world uh case for building robots uh to exa look exactly like human and that's because uh throughout thousands of years of evolution the human world is really designed to be operated >> by a biped >> that's about 170 meters tall that have two limbs and 10 fingers have digits. So if you think of something as simple as a doornob >> or this desk, >> yes, it's built at exactly the right height where a humanoid form can operate best in. So if you want robots to for example work in existing factories, you don't have to reinvent everything. You if you want a robot to work seamlessly 24 hours a day at a factory, you want it to operate in in the exact same way as a human. They can operate all the screwdrivers that are designed to be held with one hand and you can hold a hammer in the other hand.

[00:36:24]>> Okay.

[00:36:25]>> If a robot is designed in any other way, we will have to reinvent the entire process and reinvent new tools. That in that's in fact not economical. So, so having a humanoid robot is actually the most economic way >> for us to automate our future and it's the >> a less scary way. You want to talk about that? I was thinking if I unish asked me to interact one day with a very competent uh robot but it's like uh I don't know it could be a um a dog looking or a cat looking I think so that I would need to um put myself a little bit lower and talk to this person or talk to this this thing. Um, so when you have a human, somebody who's shaped like human, who looks like human, it seems that it's much more natural to interact with this >> person, I mean this machine. And also I was thinking about you because you were talking about how come and what's the point of having this human robots especially when running a marathon and sometimes people would say that okay so this time you see that more um attentions media attentions have went to um the robots and I think it's two ways of thinking um it's actually I I calculated a little bit on the internet what it it means for the 21 kilometers of continuous cooper operation for a robot is about 25,000 to 30 thousands of steps.

[00:37:51]>> So which means that from the time that you have a robot can complete all these distance by himself which means that each steps requires the robots to sense the ground conditions to adjust the balance to manage impact forces and also to coordinate dozens of ACU accuras sim simultaneously. So that's uh you need to do it for 50 minutes without failure. So this is what we have achieved today. So it's also what represents the technology that is represented with the speeds and with the distance that the robots covered. So we can uh easily thinking or imagine that one day we can use these kind of technologies in rescuing in uh some training military training in uh especially I think rescuing in disaster situations >> disaster disaster have like robots that can feel the temperature the ground temperature that can feel if there's will be the the environment >> and the robots can be equipped with more like add-ons to wet and to get us more um information from the emergency situations.

[00:38:57]>> Exactly. And that I think is very quick and you can see that how to make a a machine smarter >> smarter so that we can better use it in our daily life in the future.

[00:39:08]>> Yeah.

[00:39:08]>> Yeah. I think running fast is just the first step and then I think a more important challenge is to manage the smaller stuff >> to operate uh operate a kitchen for example. they still couldn't do that 100% competently, but they could run really fast. Uh, and that dexterity I think is kind of the next milestone here. And then I think there is value in itself for a robot to designed to be designed to run like human or look like human because as you said it makes it way less scarier when you interact with it. For example, if I want to interact with a robot, >> I I call it, >> it turns around and look at me, I would feel more I would feel more secure and respected because otherwise I wouldn't know if it's listening to me or not. So this kind of humanlike reaction although it is manufactured, it would make the user feel more secure and and more accepted to towards these kind of technology. And that that's why I think that's why uh the organizers have been putting for example robot dogs at cheerle as cheerleading leading squads and a robot uh traffic conductors to make robots just seem like way less scarier than it actually looks. And then a lot of robots this year have been designed to look cute.

[00:40:29]>> Yeah.

[00:40:29]>> Like the other robots has been designed to look cool. like it's really looks like something out of a sci-fi movie that looks really cool to a lot of boys out there. It it a lot more robots are designed to be less industrial autotype looking. M >> so it is part of a a way to push this idea that okay robots will be coming into real life in in the next few years and we should get used to it and then I think that's part of the social experiment of letting human run alongside robots and that's a a that's a so a psychological barrier that needs to be crossed because I've seen uh some commenters on the internet who have participated in this marathon as a human contest contestants. The majority of them are really thrilled about the novelty of of robots running past them.

[00:41:22]They they feel really excited about it because it's it's something new, right?

[00:41:26]>> But but the elite runners, uh some of them say that it's kind of more demoralizing like when you when you're coughing your lung out halfway there and just passed you >> and the robot passed you like it make makes you look like you're standing still.

[00:41:40]>> Oh yeah. Yeah.

[00:41:41]>> It's kind of demoralizing >> whether they were under pressured or not. Yes. So a lot of uh there will be more more and more cases where human would feel kind of threatened >> by the capability of robots. So one of the big objectives of the tech people out there is really to make this new technology seem less threatening and less scary and they think that's part of part of the thing that they're doing this year. Well, those are great points you two are making and I do have a little bit like different perspective is that so I do agree that for robots to get into our lives they need to look as close to us as possible but in the case of uh let's say factories I think efficiency still comes at first right so if you look at well um if if you look that look at the Boston Dynamics robot atlas every joints including the its head >> can span 360. Yes. and the the the arms and including the the hands and that's for that's designed to let them operate at the fastest speed in factory >> even though they are still maybe designed at a human height because you know I I think in their demo video they were working along the car manufacturing um production line.

[00:43:06]>> Yeah. And for that they can you know bring a part of the car you know from behind to front very smoothly.

[00:43:15]>> And yeah I think there's a reason for us to design robots like that in factories but there's also the need for us to design robots that look like us in daily lives.

[00:43:26]>> First one day to enter in different families to have some uh like but also attention u >> there's they don't necessarily have emotion. you can train them to have emotions. Uh but it's I think it's >> apparently much easier for elder people to get used to a life with accompanied by >> this robust from the time they're human shaped and humanliked. I think it's better or it's easier for them to accept them um to be in their life um daily life to uh doing some chores making washing dishes. Um so I think that could be and and talking about those runners I find it interesting most of the runners perhaps for them the game the aim for running a marathon is to surpass themselves is to um to feel the the the the joy of the sport. uh whereas today when they find in in a certain way surpassed by the robot they could be a little bit depressed but I think most of them are there to test the limit of humanity to test the limit of the human uh performance and today we're talking about the machino or the the the machine performance and that you see this two parallel ways but when we put these two parallel universe together um that could bring perhaps some new things or new combinations in the future. I think I think Elon Musk said a few years back that uh when AI comes and when robotic comes one of the biggest tasks for human being is to redefine the meaning in life right what do you do now like robots can do everything better than you >> uh even though robots can already move faster than you you have cars in a century back but a car is just a box with two box with two wheels it's not moving using human functions but a robot is actually like actually running using the human method of moving. So it there is a a a a philosophical point in itself that uh it can surpass you in human functions. So what is it like what is it like that all of human achievements can be beaten by a robot. So one of the biggest goals for humans in the future is to like redefine what we want to achieve with ourselves, right?

[00:45:41]>> Yeah. I mean, isn't that giving us more time to think about that philosophical question when we have humans doing all the daily deja vu for us, >> right? And also, I think another point for h having this race having robots running a half marathon is that like uh like I said, we're putting everything to an extreme. So, we are ex stretching their battery lives to an extreme. We are stretching their um the motion, testing their motions and testing all the different parts the joints of these robots because at the end of day they are going to work alongside us. If they are going to enter to into our our daily lives the reliability of these robots >> comes at first, right?

[00:46:28]>> And that's essential.

[00:46:30]>> So well let's talk about robots joining our lives. Um what do you expect? Do you think, you know, this year is quite impressive the results?

[00:46:40]>> You think they're ready to join our lives and or or are you ready? Are are you maybe thinking about purchasing a robot?

[00:46:47]>> Well, I wouldn't necessarily need a robot in my life that just runs really fast. But uh but I I think this is uh this half marathon is a big tech showcase. It's a trade show essentially.

[00:47:00]Uh so for robot to show that they can make something really really high-end.

[00:47:05]So, so that it could garner a lot of attention and then maybe the consumers in the future would be more acceptance accepted toward their consumer grade models that will not run as fast but could do a lot of other functions or may may last longer and need needing less battery swap. It's the same idea where where Xiaomi makes a a SU7 Ultra that can break record at the Nurburg Ring, but they're not not really mainly selling that product. They're selling their consumer driving that speed.

[00:47:36]>> Yeah. Yeah. That's the idea I think.

[00:47:39]>> Mhm. Interesting. I got chance to also when I was with the covering the whole marathon, I was um I also got chance to visit it. There's a place called the robot world, the big world to for you to discover and also to have chances to interact with these robots. So, I interact with some of them. I find it really interesting. You know, you have the the small ones. The small ones for the robots that from the chess playing till um marking a goal um in basketball till uh having dog robots um and today all that the technologies that that experienced in that center are able and available to consumers already.

[00:48:18]>> Uh that varies from 2,50 yuan till uh 150,000 yuan. So that's today that we have these various and what impressed me the most is that I saw a Einstein there.

[00:48:31]Um and then I even say hello to >> an artificial >> an artificial right an artificial one but an artificial Einstein with all the knowledges that he have in head. So I asked him questions I asked him how many language do you speak? He told me about German. He told me about English. And I said I asked him can you talk me in German so that I can practice my German with you? He said sure I can do that.

[00:48:52]And I asked him some questions um about the philosophical world and about the universe and about the >> general relativity >> the general >> he squ he did explain to me very well and so I'm thinking that these kind of um technology could be used in the future somebody who looks like Einstein who speaks the language that Einstein speaks >> it's a be best way to to learn somebody in this intellectual and in this um interactive um part in activities with him and there are also many other ways like for example if I want to uh learning playing chess so why don't I have an AI teacher in front of me um already and then there's also the um the machine dogs that impressed me the most because um I asked him many questions like uh was that what day is today what's so better today and can you help me to carry um the thing from point A to point B and they said yes and then um there's also a way that I asked him um like can you see me? Can you take me a photo? Um can you tell me uh how I can go from from from this place to that place. So all these examples that I practice with this dog, I think he did it like beautifully and he gave me all these informations and he can even tell me where I am and he can follow me all around where I walk. Um so that's something that's quite interesting. I'm I'm really h I would be happy to have one dog with me carry that I can carry with me to the supermarket so he can carry me um old stuff that I will buy in these markets and he can also follow me to take me pictures to do videos um as cameraman and he can also um do some researches like on time researches about um what's happening around the information that I would need. So these are ways I think that they can really integrate um the AI into our daily lives. So I mean this is already available on Chinese market and I have also some French journalists friends >> okay >> who have already asked me uh okay next time you've come to France can you bring me this this this this this Chinese um this Chinese um AI stuff are there >> including a robot >> including a robot that I can train it by myself I can train it for my use in France so yes >> you should reply that oh at a higher you you have a price mark you know >> right and uh I think that's really interesting And what we can see is also this time we saw some international teams in the marathon and they trained Chinese-made >> robot with their algorithm with their technology. So you can see there's a beautiful combination between different nalities of artificial intelligence. How can you put them together?

[00:51:33]>> Right.

[00:51:35]>> Great point.

[00:51:36]>> Well, first of all, you know what robot I would really like? I'd like a robot that can be my gym spotter. play basketball with you >> probably like if I'm lifting balls for you >> if I'm lifting a dumbbell >> and I I I I couldn't do it anymore. My robot would be like, "Okay, I got you.

[00:51:53]>> Then what are you practicing?"

[00:51:54]>> And then it has all my health status, all my health stats. So when it runs alongside me on the treadmill and I want I can't no longer go on, the robot would say, robot would say, "I know for a fact that you can >> just push it to all your health stats.

[00:52:11]Look, I'm running as at as at the same speed as you. So, keep going.

[00:52:15]>> Oh, >> that's the best coach that you want.

[00:52:17]>> Yeah, that's the best coach. You cannot lie to that coach because you have he has all your health stats and has all of your movements calculated >> and on time can cheer you up when you can't do it anymore.

[00:52:27]>> Yes. Yes. And have to look like a human, right? Exactly.

[00:52:31]>> That could be interesting. Yeah. If if there is one, would we be wanting to buy one?

[00:52:36]>> Oh, yeah. I'd cons I'd consider if the price goes down. It's really early in the tech curve. So, it's they're really really expensive. I think Honor can make one probably.

[00:52:47]>> Yeah. And then and then they did say that they the next step for them is to put these robots into their retail stores.

[00:52:53]>> Once they have these kind of robots, we'll see that they they could be like the EV cars today is that okay, you have a robot, I can rent it to you. So, we have a EV car, I can rent it to you. It could be the same. So today we have the shared cars and shared bicycles and shared one day we could having shared robots, right? For example, I would need the robot to do this for me for five days and then I'll pass it to somebody else.

[00:53:18]>> Yeah. We can we can we can have a robot at a much lower cost.

[00:53:22]>> Wow. So we're talking about business ideas now. We're too far away.

[00:53:26]>> Well, coming back to robots in life, I have a I don't know. I have a bold idea.

[00:53:31]I don't know if you think it's too dark, but Chi, you and I both covered brain computer interface.

[00:53:37]>> So, imagine if you can, this is actually from a a a comic. If you can upload your brain to a digital world and have a physical robot representing you or looks just like you and go go to work for you.

[00:53:54]>> But then how can I tell who is who? like how can I which is chen number one chen number two >> or chi number one number two I would have problem in >> ethical problems I'm thinking >> yeah it doesn't solve it doesn't really 100% solve the problem for me though like if I could upload my consciousness for 100% onto a robot I'm still here I mean if the robot looks exactly like me and can >> and that goes back to the question of what's the meaning of us >> yeah but I'm still here I will still experience getting old and being getting old and getting sick and and dying in the future.

[00:54:32]>> He can live as me. So what what's me?

[00:54:34]What what my concern?

[00:54:35]>> Imagine if you were let's say Steve Jobs and you were having cancer and dying and you were able to upload your consciousness.

[00:54:43]>> In that case it would work like for a healthy person. I would I would know why you want to do it because if I want if I upload my consciousness to a to to a machine where he can live as exact the exact same life as I life as I do, >> I'm still here. So I I still experience all the pain and inconveniences of a human flesh and blood body. So why would I want to do that? But that's a philosophical question is that the thing is that even if you have a robot that looks exactly like you who thinks exactly like you the conscious that he have is you in the past till today. Oh yeah.

[00:55:16]>> Is this is but the thing changings >> right also in two different beings in the future because what you encountered could be different the environment around you could be different as unless you can like simultaneously get uh renewed with the um machine or it means that there could be you but your version number two there but what you are experiencing in the future could be different. So I think like in 10 years >> that could be still in two 10 different two different persons >> and then the question comes to who holds the responsibility, who is the legal, you know, civil person in in this society. All right? So, I think similar to AI, as we witness the fast development of robots, the same like regulation and safety concerns should also addressed in this industry as well.

[00:56:04]>> Exactly. But yeah, um I'm also confident that because I already see a lot of robotics retail stores existing in shopping malls in Beijing like in I I saw that um in during this spring festival in Muing and also yeah I think this year in China we're seeing this boom in robotics industry is largely thanks to the show that they did at our spring festival gala right and Um yeah, I'm confident that over the next year we'll see more robots in retail stores.

[00:56:40]>> Yeah. But um so apart from the good sides, right? Uh I think Shahi also mentioned that can these robots go to the the dangerous places.

[00:56:52]>> So this year before the half marathon, there was actually a warrior challenge for robots for both human robots and the robotic dogs or quadripetss. Do you guys want to check it out?

[00:57:04]>> Yeah, let's try.

[00:57:05]>> All right, let's uh take a look at this video.

[00:57:09]>> So, that robot just got back online.

[00:57:11]What you are seeing here are not some variety show or a Spartan challenge.

[00:57:16]These robots, including the humanoids and quadroets, are taking on a warrior challenge before running a half marathon tomorrow in Beijing.

[00:57:27]Welcome to Hot. I'm Chen. before the Beijing E- Town half marathon for both human and robots. There's an even tougher test.

[00:57:37]So, behind me are 17 different obstacles and robots have to go past them either autonomously or via remote control. They need to go through narrow bridges, climb up slopes and stairs, and even duck under barriers, and even open doors to clear obstacles.

[00:57:57]It's clear that these challenges are set to simulate real life scenarios, especially disaster relief. So, what are the biggest technical challenges for robots to move like this?

[00:58:50]So what this um challenge was about was that there were 17 different obstacles set in this. It's it looks really like a Spartan race but for robots for both humanoids and quattrabets or the the robotic dogs right and and for the quattrabets they also uh >> classified into the willed ones and the pointed ones >> and they can choose to complete from these 17 challenges 17 tasks and those scenarios are very similar to what we see in real life um rescuing situations.

[00:59:26]So yeah, I think from this type of race that the organizers are hosting, it's really a testing ground for all these robots getting ready to into our daily lives and they are also the test is also classified into like the controlled ones and um automated ones. So robots, some of the robots, they can also complete the race fully autonomously by recognizing, you know, the barriers and the obstacles and there were even heavy bags hitting on them. They they know how to remain balanced and that was quite impressive to me as well.

[01:00:03]>> Wow.

[01:00:03]>> So which means that behind it is also a supply chain density there. You got the technology here and perhaps we can >> we can see already the progress that we have already made here in China in one year. the speed has went up the smart the smart uh level and also the flexibility has impress impra progress a lot and for me I think this also interesting is that when we have the breakthrough technology breakthrough we can also wait for a larger big scale of production in the future and perhaps one day we can also having these new dogs this contrabats to come on to different um very extreme situations on plat toes on uh the earthquake situation and to go there rescue people to go there to bring food and water um really for the humans wellbeing.

[01:00:55]>> Yes. When I saw that warrior uh challenge footage I I instantly think of like disaster relief situations for example the quadropaths uh they don't need to be stationed at like a village next to a mountain. They could be stationed in the mountain >> like as as long as there's electricity there like it doesn't need food doesn't need water so it can respond to disaster much faster like a human have to take like a helicopter from a from a helicopter base towards a disaster area but they can be embedded in all kinds of dangerous places as long as there's electricity there whenever disaster happen they could like just jump out >> like the first second they could be like a really really effective first responder.

[01:01:40]>> Yeah. Or they can be here, you know, let's say in a forest, they can walk around, patrol the forest and detect dangers before it happens.

[01:01:49]>> So to to prevent disaster happen before, >> they could totally do that.

[01:01:53]>> Yeah. Yeah.

[01:01:54]>> So that's very impressive for us to see the challenge and the marathon. So at the end before we wrap up for today's show, we briefly talked about that, but if you're listening to this point or watching till this point, you are going to uh have a big way in the future because we're going to talk about business ideas. What do you think for we can do um with these robots? Rashano, you just talked about shared robots.

[01:02:19]>> Shared robots perhaps. I think that that could be a very gen it could be a genuine idea because you do an investment and that you can rent it out for different uses and I think that could be because nobody want to offer a very expensive robot um for they're only solely used for one person or two people needs. Um I think perhaps for them it's more um okay I'll use this because I have a need for a certain period of time so that I will need a robot that come to my home or Tom come to my workplace to help me to solve this problem. So in this case I think that could could be quite interesting to have perhaps in the future an agency or enterprises that doing the uh renting of the robots. Yes.

[01:03:04]and to your knees because today I want to experience this kind of robots a human robot and to tomorrow I want to experience oh this is better looking robot I can change to one and then I can experience a dog robot so why not I think that could be interesting and um a good um genuine good idea >> yeah and I think um we see that a lot of the influencers are already purchasing these robots but just for entertainment and also a lot of u the conferences are having robots as the uh front desk or uh the receptionist but like soon enough they will be in our lives and uh like you said I mean uh you can experience different functions of these models at a lower cost and I just want to add a little bit to that is that uh last year China already produced delivered what around 40,000 unit of humanoid robots >> and so yeah just imagine if that number goes up, the manufacturing goes up, the price would could easily be cut down.

[01:04:09]>> Yeah. You know what I'm thinking?

[01:04:11]>> Yeah.

[01:04:12]>> Robot fashion. Like maybe that's something you could talk to with your French colleague.

[01:04:18]>> Yes. Because that is a rising industry in the future. I believe.

[01:04:22]>> No, you took away my idea. Yeah, >> if you think about the early days of computers, they all look pretty industrial and non aesthetic and and really rugged >> until companies like Apple come along and make it look really good and everybody wanted one.

[01:04:39]>> So, same with robot. It's has all these wires now looks really metallic and really ugly. So one of the biggest uh businesses in the future I believe is to have robot fashion and to make it appealing to customers to all the consumers in the world and it's has some similarity with designing hum human fashion but like how do you design like a hat for a robot that can rotate their head 30 30 to 60 degrees?

[01:05:10]>> Yeah. How do you make a coat that looks really chic >> for a robot with four hands?

[01:05:15]>> So that there's a huge cooperation with French fashion industry.

[01:05:20]>> Yes.

[01:05:21]>> Yeah. We should we should discuss that with our French colleague, >> right? That could be interesting to see a good-looking and very very fashion uh motto, right? high-tech robot robot, but also it could be it could be a very good-looking girl with a very good looking good good-looking lady that come in front of you or it could be a very good-looking mature um assert that comes that's kind of attracting and um I think why not one day using um this kind of robot to representing a air company or representing a um a a new brand and why not? Yes, you're right. You're throwing a lot of free ideas to them. All right, since you took away my idea, I think that we'll wrap up of this episode of Hot Take. Okay, let's um keep the imagine wild and see you next time.