Live: A walk through AI – present and future – at WAIC 2026

[00:00:40]Well, greetings to our viewers around the world. This is CGTN live streaming at the 2026 World Artificial Intelligence Conference, also known as WAC. I'm CGTN reporter Wubin here in Shanghai. And this year the world artificial intelligence conference has you know has its largest scale with the greatest participation. The exhibition center expands to three locations and I'm currently uh standing inside the expo exhibition hall which is one of the major exhibition centers and today in our live streaming we will try to understand one qu answer one question that what can AI do for us today and a as you might know the AI is such a big uh you know has covered a lot of aspects but one of the most important aspects and most hot topic trendy topic in recent years is embodied intelligence or humanoid robots. And today we will explore the H hall uh the hall three uh to explore a lot of u uh robotic companies to see some of the latest developments. And the first and of all let's check out this robot because this is the f uh the robot uh standing here serving as a volunteer. Although in the past years you probably have seen uh robots volunteers but this is the first time that they have a have this uh vivid humanoid robot. In the past maybe it's just like you have a tablet but this year look check out the uh shape of this uh robot. It looks just like a human being. And now I will try to speak to him or or her to to see what it can down.

[00:02:35]Hello.

[00:02:39]Hello.

[00:02:50]So, can you speak English?

[00:02:54]>> I already support English US. So, you can just talk to me in English US.

[00:02:59]>> So, what are you doing here?

[00:03:04]>> I'm here as Agibot, star performer at W AIC. I'm in charge of booth, reception, interactive shows, and answering questions about our robots. I can even do street dance and stunts on the spot if you want to see. Would you like a quick demo?

[00:03:21]>> Sure. Can you uh show a little bit of dance to me?

[00:03:26]>> This feature is being upgraded. Try something else first.

[00:03:30]>> Can you tell me where's the restaurant?

[00:03:34]>> This feature is being upgraded. Try something else first. So, can you tell me where the Agribot counter is?

[00:03:42]>> This feature is being upgraded. Try something else first.

[00:03:45]>> It seems that uh it needs up some upgrading, but this is the first time that a volunteer robot is used at the exhibition hall. And now we will go to some of the pavilions, some of the counters to explore some of the uh uh latest achievements uh of the robotic companies. Thank you very much. Uh and this year uh some of the background information uh over 1,100 companies uh participating this 4-day event. They bring about 3,000 uh uh uh products and their latest technology breakthroughs. Uh and over 300 of them is being showed for the first time. And here we will go to the first stop of our live streaming today is the uh pavilion of a Shanghai based company called Ajibot. And why we are going here? Because Ajibbot is the uh now is the top producer of the world's humanoid robot. uh la uh just last month the company has rolled off its uh 15,000 humanoid robots and this time it has presented one of its latest model and actually listed one of the 10 most important XFits of the 2026 world artificial intelligence which is this one. So for more uh for more we will talk to the staff of this of Agibbot to tell us more. So it is one of the most important 10 uh exhibits of the 2026 world artificial intelligence. So why is it Fore!

[00:06:09]Fore!chch!

[00:06:14]Uh okay. So I will first uh translate a little bit to our audience. Uh so this this model of uh a robot can work 24 hours. Just mention non-stop work for 24 hours and it has lasers, cameras both in front and back and it can uh see the uh environment uh of the of the environment all around it. And uh just now you also also mentioned it has a flexible hands to help it grab the detector hand detector's hand to help it to grab some of the most do some of the challenging tasks. Uh okay now this robot will try to uh hold this objects in front of us. Let's take a look.

[00:07:13]Johnnychech.

[00:07:41]Okay. So although holding this uh a a can of coke seems very simple for human being but for robots is still a challenging task and doing this requires multiple sensors to work together and also the controllable detector's hands.

[00:07:58]Now So I'm now asking we now see three objects in front of the robot. One is a box and the second one is a fluffy toy and the third one is a can of Coke. So why are we choosing these three objects for for for for the robot to to to hold?

[00:08:58]So these three objects represents for for example this one is more is fluffy.

[00:09:02]This one is uh is is is harder to hold. And uh this requires the robot to think simultaneously and decide how much strength will he use to to hold these objects.

[00:09:22]So it's also a test of the robot's thinking ability. Is that right?

[00:09:38]Wow. So doing this simple task requires the uh uh function of work different functions to working together both the brains the sensors and also the dexterous hands. So now I would like to see just now we have successfully hold the can of coke.

[00:10:05]So now I'm giving this robot a test to ask him to hold this fluffy toy because he has to use a softer, you know, he he he can't use this much of strength to hold this toy.

[00:10:33]So now uh the robot says this present is for me. So thank you very much.

[00:10:39]Now one more task because this box is quite large.

[00:10:51]So now he will try to uh grab this very large box to see if it he will be successful in that.

[00:11:08]ah >> by hold uh so I can tell our audience what I feel because this box is uh he's using much stronger strength to grab this box.

[00:11:32]So this is also doing these simple tasks but actually it's a test for the robot's thinking ability and also its flexibility both in hands and also its its body. So thank you very much for for for sharing this with us. Let's continue our journey here. Uh so uh for robotic uh development in the past years we've seen u the development of ro robots being incre uh developed really really quickly. I remember two years ago in 2024 when I first uh covered the world artificial intelligence conference. At that time uh the robots can only perform can do simple tasks like do street dances and also those dance moves are not that um delicate. But this year we've seen much more improvement. First of all, the performance side uh the robots can do much complex movements and more importantly we've seen robots are entering our everyday life and also entering factories. So here what we are seeing is a uh scenario of robots working in factories. So let's first give our audience uh a minute to see what this robot is doing.

[00:13:18]There's not Heat. Heat.

[00:13:57]All right. So just now what you were seeing is a scenario a simulated scenario of robots walking in factories.

[00:14:04]For more understanding let's talk to the staff here. So uh thank you for joining our live streaming and first of all please introduce to us what this robot is doing what this scenario represents.

[00:14:52]So this is just now the staff here told me that this is actually not a demo scenario. It's actually used in factories and it's quite mature. So this is not for demo. This it's already used.

[00:15:03]>> Yes. Yes. Yes.

[00:15:23]So at first only two robots are working in the production line and now eight robots are working together and in the future they plan to have dozens of robots working in the production line and uh I would like to uh ask my cameraman to to give a closeup to the uh television out there. What's showed on the television is a live streaming that the company uh did last month in a factory in Jangi province. Um and it shows that several robots are working non-stop in the production line. So this is not for demo. This is real ones. The robots are already in factories uh helping enhance productivity. Um so that's that's one important aspect because just I I was telling our audience for the past two years the robots are developing real quickly not only into our everyday life but also in the robotic sector which is uh very important So how much productivity will the robot enhance?

[00:16:59]Foreers.

[00:17:23]So, so free of human hands.

[00:17:29]So just now the staff told me one of the very important aspect. So deploying robotic uh forces in u factories is to is to make human beings free of uh uh to do more creative tasks and those repetitive tasks or maybe in some dangerous environments can be replaced by robots.

[00:18:18]All right. And one more aspect is some dangerous dangerous environment or some environment that is not h suitable for human beings to work in that can be uh replaced by robots.

[00:18:37]Okay, now let's move here. So, so how about this? This side. Yeah.

[00:19:36]So this area shows that the three robots are working together. The first robot will grab those materials and move to the second robot. And then the second robot will move the material to the box.

[00:19:48]And the third robot will lift the box and maybe carry to carry them to to trucks or somewhere else and deliver them the uh operation the cooperation between different robots. So that's also it looks very simple but just imagine that when humans are collaborating uh together or working together you still have a lot of problems sometimes negotiation sometimes you know uh different kinds of problems not let alone a robot so making robots working together actually is a very very complex task.

[00:21:02]So more importantly this agent platform the company is building the it's it will be simpler for robots to actually adapted to factories en environment. In the past maybe it takes a month to recode but now it maybe just need a day for the adapt to the different factories. Excuse me.

[00:21:26]So thank you very much. Thank you for sharing. H uh so let's continue our uh live streaming here.

[00:21:36]Just now what we see in this uh uh pavilion of this company, we've seen some of the latest uh breakthroughs of uh this uh company and also we've seen some industrial applications.

[00:21:54]So this is these are very important aspects of the robotic development. is not only for showing not only for for doing street everyday life and industrial sector to really boost and in the past days I've talked to some experts in the industry and we know that now China has a very strong um capability of building these robots but also China's capability is also benefiting the whole world because uh China has a rather complete industrial chain and can uh produce all these robots with a relatively cheaper price, cheaper cost. And these can be uh can benefit more developing countries because they can we will make uh like robotic products more affordable.

[00:22:50]the just I'm so today is the second day of the 4-day event and the exhibition center is open to the general public.

[00:22:59]Tens of thousands of visitors are here.

[00:23:00]So you see it's quite crowded in the exhibition hall and in this uh pavilion we've noticed something that's very very cute. Let's first give a close closer look of this uh of this toy. I don't know how to describe it, but maybe we'll ask uh the staff later to see what this is. But it looks really, really adorable. And this is the first time I've seen things like this.

[00:23:30]Let's talk to the staff here.

[00:23:40]So can you please introduce a little bit of this what it is and how do we call uh call it >> okay so like firstly say hi to our like desktop companion digital life gr nano and it's like we designed it for companion and also um we are mainly focused on the physical and emotional interaction with people [music] so um the most important thing that We designed it was like it's very easy to carry out like it's really light like 200 less than 200 g and >> so this is a companion robot >> that's a companion droid >> and also very important aspect of robotic development for companion so I've seen you have here you also have one uh carried into a pocket so I love it you just mentioned you can carry it >> yeah we can carry it out he or she has a nano Okay.

[00:24:34]>> Yeah.

[00:24:40]>> Basically, uh we're focused on the the physical interacting like compared to the droid and the human. So, it has a camera can do the face tracking. So, we can confirm it just always focus on you.

[00:24:52]So, it's a very special design and also it can know up to eight people around you. So, it can remember all the interaction that you have made with him.

[00:25:01]Like you can touch him, he will remember it. He can hold him in hand, he will remember it. So basically it can remember all like all the interactions.

[00:25:12]>> Yeah. And so we mainly uh designed that feature.

[00:25:15]>> Yeah. So can we interact with uh with yours or with the uh other ones for a little bit?

[00:25:20]>> Uh we can just randomly pick one of it and just show you how we interact. Sure.

[00:25:25]Sure.

[00:25:34]this emotional bond with it.

[00:25:36]>> Yeah. So basically we can touch it. So when we feel good we can touch it and it also feel comfortable and when we feel bad we can just shake it and we feel sad as long as we can. Yeah. Yeah. It's just you can see fire in eyes. So basically it's like a like an emotional bonding between it and within us. So also the feature that it can know people that we really want to introduce it to the circle of friends of us. So we do really want it become like just not a a companion joy but also one of our friend.

[00:26:10]>> So how so I just rub his head >> touch it. Yeah. Yeah.

[00:26:16]>> So you can see the interaction events on this iPad. So so like you're touching it and then we just feel it.

[00:26:24]Just look at look at his eyes. He's so adorable.

[00:26:28]>> Yeah. Glad you like it. And and also those emotion bubbles are just really close connected to its own emotion. Like just like I shake it, it feels sad and it will feel sad in here.

[00:26:39]>> Yeah.

[00:26:40]>> Yeah.

[00:26:40]>> I've seen the angry fire in the eyes again, but I don't know. So sorry about that.

[00:26:46]>> It's okay. It's fine.

[00:26:49]>> Thank you very much. Thank you for sharing with us. So this is also another very important aspect of robotic development. The emotional the companion robots form a kind of emotional bond with human beings to provide sometimes a company by your side.

[00:27:10]And now let's uh continue our exploration to those viewers who just joined our live streaming. This is uh a live streaming in the world artificial intelligence conference hosted in Shanghai.

[00:27:29]And we are now in the uh exhibition uh expo exhibition center in the Pudonu area which is one of the major exhibition hall. This year over a thousand companies are here bringing their latest development and this the hall three is mainly focused on embodied intelligence focused on humanoid robots.

[00:27:52]So we see uh humanoid robots, robot dogs and different applications of robots.

[00:28:10]Just now we have se uh talked to different companies about some of the applications of robots like uh uh in the past the robots can perform can do street dances and now robots entering factories and also we explored some of the companion robots.

[00:28:36]And now this company is showing mainly the uh wearable bones for to support your to support you when you do some heavy tasks.

[00:28:47]This is also one of the star companies of this conference.

[00:28:52]And this here is another star companies of the conference. The Brinkco is a company working for computer uh brain uh interface technology. And now what you are seeing is quite interesting. This uh staff of the company is using his his his brain signal, his mind to control the dctor's hands to hold waters.

[00:29:22]And we we see that there's no control uh no machines to control the doctor's hands, but only with his brain signals.

[00:29:33]And this is quite meaningful because those technologies can be used for uh those those uh people who are disabled uh who uh and can let them to maybe stand once again.

[00:30:02]This just by looking at it is it's quite magical that just by thinking about thinking about holding the glass of water, it actually happens.

[00:30:21]And just now we've experienced some uh applications of robots. But here's one more question that I'm uh always thinking about because just now we've seen they've all saying that thinking the thinking ability of robots is very important. So uh what needs what what does the need for a the enhancing the thinking ability uh of robots? So now we are here at this counter and I will explore this question with the staff of uh uh no metrics.

[00:30:54]>> Uh so this company works for uh brains work for models for for robots. So uh first of all please introduce yourself.

[00:31:04]>> Yeah. Oh okay. Okay. Thanks. Hello everyone. I'm I'm I'm Kayman. I'm the uh product solution manager of no matrix and we are a company that uh we are trying to uh put uh general purpose robot intelligence into the phys physical world and we are developing uh force centrate large embodied model to control any robot to do any task. Yeah.

[00:31:29]>> So simply understanding you are working on the brains for robots. Yeah. So why does robots need a need a specific brain? Why can robots just use like large language models or other models built from uh you know from other sectors?

[00:31:43]>> Okay, basically like for uh chat GBT this kind of large language model they can only uh work in the digital world.

[00:31:51]So they can they are unable to interact with the real object in the physical world. So that's why we need embodied model to help robot to interact with the environment and to do and perform the task in the real world >> to interact react with the physical world to understand the the laws of the physical world that's very important. So how are we proceeding so far >> uh currently like uh we have uh launched our own model like uh I will introduce you to you guys later and for the first one is like our intelligent pharmacy. So I will show you our the first uh ability.

[00:32:30]Okay.

[00:32:32]>> So this is a like this is intelligence farmers you just mentioned. So what will the robot doing here?

[00:32:39]>> So basically like a the robot can receive the order from all night and it will collecting the medicines. So it's it's able to collect nearly 3,000 SKU of medicines. So there are three yes there are three uh big features for our robot.

[00:32:56]The first one is like zero changing. So which means like uh you don't have to change the environment of the real pharmacy even the chef. So the robot is able to collect the uh medicines from the real uh shelf. Uh second like it's very low cost which means like so see here you can see it's one arm robot.

[00:33:17]>> Yes I think our audience can see that. I have one question. uh that is that why does this robot uh doesn't need a dexterous hands it just have this three uh >> uh I think yeah yeah there are three like uh changing tools here yeah because like for our robot uh our our brain is able to uh instruct our uh robot to collect different uh medicines with different uh tools which means like uh it will be a very low cost for the robot to do this so hands is pretty expensive.

[00:33:52]Yeah.

[00:33:53]>> So, >> and for and for uh tasks like picking up uh medicines, the texture hands may be a little bit too much.

[00:34:01]>> Yes. Yeah.

[00:34:02]>> Just now you mentioned cost is also a very important thing to think about especially when uh like the developing countries other countries around the world if they want to use this cost is also something very important for them to think about.

[00:34:15]>> Of course. Yeah. Well, we're not uh having deployed our robot in uh three brands uh chain pharmacy customers.

[00:34:24]Yeah. So, they really care about the cost and also the how easy to use our robot.

[00:34:30]>> Okay. So, >> yeah. So, just now we've seen that the robot have have very smoothly >> Yeah.

[00:34:38]>> Yeah. pick up those medicines and bring it to the counter.

[00:34:41]>> Yeah. So, are we going to the next stop?

[00:34:44]So in the future maybe uh we will have 24-hour uh like pharmacy without any human to do the night shift. That would be some very good good news to for for the workers there to hear about.

[00:35:05]>> And just now you talked about brains for the models for for for robots. So now what what are we showing here?

[00:35:13]>> Yeah. uh we are showing like we just launched our uh world action models.

[00:35:18]Yeah. So the the biggest feature of this model is that uh we are not using robot data. So which means like you don't have to uh control a robot to collect data.

[00:35:28]You can just use our uh you you can just uh let people to wearing uh devices to collect the data in the real world. Here we are showing like uh the demo is live.

[00:35:39]Our robot our model is is very has strong robotness and generalization. So you can see here it's very easy. You can just change the environment. It will uh percept the environment is changing and adapt to this kind of changing. Yeah.

[00:35:54]The secondly like uh secondly like uh uh we are we are having this kind of large scale data infra. So I will show you like uh our devices. Mhm.

[00:36:06]>> So this is quite similar to what we are experiencing at the Abibot pavilion. So you change the environment, you change the object and now the robot needs to think, needs to adapt to the changes.

[00:36:18]>> Yeah. So here like it will like pull the coins into the team you are v you are voting for for the champion cup. Here it's Argentina.

[00:36:26]>> So the robot is seeing which where did you put the football and then put the coins inside the box.

[00:36:32]>> Yeah. here which angular it looks very simple but actually it needs uh some strong thinking process for robots.

[00:36:38]>> Yeah.

[00:36:38]>> Yeah. Okay.

[00:36:40]>> Right. So I'm now here I've seen uh a uh washing machine. So I'm wondering why does he put two washing machines here.

[00:36:49]>> Okay. So here like our robot is doing the laundry.

[00:36:52]>> Our robot is helping people do laundry.

[00:36:54]That is quite helpful I think.

[00:36:55]>> Yeah. Yeah. And like um for robot is the first ability is like it has very strong uh coordination ability. Here you can see like the robot can work with this kind of mobile robot right here. And also like it's able to do this kind of long horizon task which means like he can do the laundry whole process of the laundry including opening the door and closing the door and also transform transforming the clothes to from the washing machine to the dryer. And also I noticed something very interesting because this robot has to detect you know the depth of this uh this washing machine and to to to and to put the clothes in it because if he uh didn't reach deep enough >> yes >> the clothes will will fall on the floor.

[00:37:40]>> Yeah that's correct. So our robot still uh has a very strong environment uh perception ability. So which means like for this kind of uh drawing machine drum so it's very compact so when when the robot is working it it should avoiding collision of this kind of environment.

[00:38:00]>> Yeah.

[00:38:00]>> Yeah. So this is also very important because we've seen that we've also have experienced robots is helping human beings to uh free human beings of doing those repetitive very uh tiring exhausting work.

[00:38:16]>> Yes actually. So currently like our robot is uh working in some hotels to help people to do this kind of laundry.

[00:38:22]>> This is already used in hotels.

[00:38:24]>> Yeah. Yes. Okay. Yes.

[00:38:28]>> All right. So thank you very much. Thank you very much for sharing with us. And now let's so today live streaming so far we have experienced uh we have tried some uh give robots tasks. We have seen robots are doing uh different kinds of tasks from uh uh being a volunteer and telling you where the locations of where you want to go the locations to some industrial uh application and also to the uh to the to the use of our in our daily life like in a pharmacy like in uh helping people doing laundry. So this is what robot can do today. But what about the future? And also different uh uh interviewees I've talked about have shared that uh now the cost of uh producing robots is also something that's uh important to think about and the China has announced at the world 2026 artificial intelligence conference that we will help more developing countries especially in the global south to um enjoy the uh development uh opportunities of artificial intelligence and with and and China and with China's uh improvement in artificial intelligence I think um we will bring the world more opportunities and uh if you have time do come to Shanghai and do join the uh world artificial intelligence conference and then see those development yourself and I think that brings us to the conclusion of live streaming today and bye for now