China Just Challenged Tesla’s FSD… So I Flew There to Test It

本站添加: 2026-05-06

[00:00:00]There is no urgency. There's no fear.

[00:00:02]There's no concern.

[00:00:03]>> Well, they're used to it.

[00:00:06]>> This I mean, look at this.

[00:00:08]>> The sky is just in the middle of the road.

[00:00:09]>> Suicidal.

[00:00:10]>> It really feels that way. Today, I'll take you to China to test something that shouldn't be possible.

[00:00:15]>> I'm speechless. I mean, edge case after edge case. Seriously, >> this is unbelievable.

[00:00:21]>> Do people not sue each other in this country?

[00:00:24]>> Oh gosh, he's hitting a street sign. Oh, he's going to rip that sign right off.

[00:00:28]He is. Oh. Oh, >> in one of the most chaotic cities on Earth, full self-driving to put Tesla's autonomy, the gold standard, headto-head with a brand new Challenger you've probably never heard of in the States.

[00:00:41]>> This is an extreme.

[00:00:43]>> Look at this intersection.

[00:00:45]>> It's unbelievable.

[00:00:47]>> One has nearly 10 billion miles of training data. The other just rolled out its self-driving stack in 2026. And what happens next might completely change how you think about autonomy. Now, both cars rely purely on cameras and AI, but Xben quietly added something Tesla has removed. I don't know how it's able to take in so much information and react so quickly. So, we're putting them head-to-head on the exact same drive.

[00:01:12]Full self-driving through one of the most unpredictable, densely packed cities on Earth. Strap in. You're going to want to see this.

[00:01:20]>> See how it's not slowing down, but clear as day, it sees it.

[00:01:22]>> Mhm. But it's almost like I'm going to do what I need to do and you need to figure it out >> and I'll only stop if I calculate.

[00:01:30]>> It's driving like a human drives here.

[00:01:32]>> Yeah.

[00:01:32]>> Hey guys, quick one. We've noticed that over a million of you guys are watching every single month, but seven out of 10 aren't subscribed yet. It doesn't cost you a thing, but it helps us get invited to events just like this one. All right, that's it. If you enjoy our content, please go ahead and subscribe and let's get back to it. But before we can even do any of that, I have to get my Chinese driver's license. And guess what?

[00:01:53]Everything is in Mandarin. So, if I don't pass, we might not even get to the good part, guys. I'm so sorry. Just kidding. I passed. Which means we are going straight to the first test. Some ground rules first. The Tesla is running on the latest version of FSD here in China, version 13.2.9.

[00:02:14]And the P7 is running on its latest version of self-driving called NextGen Pilot on VA 2.0. And the cars both have the preset way points they'll navigate to. The lane markings right now are not clear dividing the two lanes on the road. I don't know if you noticed that, but there was an area right there where the lane markings >> were pretty light, but it was able to know where we were supposed to go now.

[00:02:40]>> Oh god, somebody stopped.

[00:02:41]>> Right now, someone has stopped. We're kind of in between the two lanes.

[00:02:46]We'll figure it out. It did. Sure did.

[00:02:48]And honestly, that's how a human driver would react as well in that scenario.

[00:02:53]>> You've got sidewalks here where scooters are >> paralleling you going faster than the car. So, I mean, there's just so much. I mean, look, there's a child in the back of a a scooter there leaning forward.

[00:03:06]>> There's just a lot. M.

[00:03:07]>> Okay. And now bike coming.

[00:03:10]>> Scooters every direction. Bikes every direction.

[00:03:18]>> Okay. That car just cut us off.

[00:03:24]>> Definitely feel safer with the car driving than me.

[00:03:28]>> Yes.

[00:03:29]>> You too. I don't know if I should take that as a compliment or an insult. Well, I mean, knowing that we haven't ever been to these areas ever in our life, just it feels good driving the car make all the decisions.

[00:03:41]>> Turn right, then your destination will be on the right.

[00:03:46]>> So, the next way point.

[00:03:48]>> So, with Tesla, once the car reaches a waypoint, we had to disengage to navigate to the next waypoint, which for a few seconds would give me a chance to drive it on these streets manually.

[00:03:58]>> Do we want to turn right or do we want to just stay on this road?

[00:04:03]Should we just let it do its thing? Um, >> it's going to go into an area where it should.

[00:04:08]>> That's not right.

[00:04:09]>> Okay.

[00:04:10]>> Yeah, it was just >> continue trip.

[00:04:12]>> Continue trip.

[00:04:15]>> Back out if you can. Yeah.

[00:04:17]>> All right.

[00:04:18]>> Oh, no. Woo.

[00:04:19]>> Now you're This is why I wish the car was driving. These little moments like this where there's a lot going on, but woo. I don't know how people drive here without having this.

[00:04:33]Okay. Cuz the scooters, they just do their thing and they expect you to stop.

[00:04:39]>> Unless we went down that road.

[00:04:41]>> I don't think that's an actual road.

[00:04:43]>> Okay. After this, I don't know.

[00:04:45]>> I think I just need to slowly do it and hopefully they stop for me. Do I drive like how they drive where I just say, "Hey, >> you can drive more aggressive.

[00:04:57]This is your chance. This is my chance.

[00:04:58]Okay, >> good job.

[00:05:02]>> So, next time if we see a waypoint that takes us just off the road, we'll try to stay on the road.

[00:05:07]>> All right, >> re-engage as fast as possible.

[00:05:12]>> That's the first time.

[00:05:15]>> Instead of just pulling us off to the side of the road.

[00:05:17]>> And look at this. That might be that might be a disengage. I need Are we supposed to go?

[00:05:21]>> No, we're It's at a red light.

[00:05:22]>> Are we in the middle of the street?

[00:05:23]>> We're just stopped in the middle of the street at a red light. Okay. So, yeah, I'm going to put it back on now. So, second disengagement within couple minutes of each other.

[00:05:35]>> Now, turn left.

[00:05:36]>> I see what you're saying.

[00:05:41]>> All right.

[00:05:42]>> So, it's a yielding green. Look at this.

[00:05:44]>> Yes.

[00:05:46]>> All right.

[00:05:47]>> Oh my god.

[00:05:47]>> This is crazy.

[00:05:49]>> Super scary. I'm going to put my hands on the steering wheel just in case because >> this is a scary intersection right here.

[00:05:56]Oh my gosh.

[00:06:00]>> All right.

[00:06:02]>> It did it.

[00:06:02]>> Very confident.

[00:06:04]>> It was more confident than I was. All right. So, we just reached our destination. I'm going to disengage it and then continue the trip.

[00:06:14]>> We're good to go. Let's put it in FSD before anything scary happens.

[00:06:22]>> Now turn right >> again. Signs that none of which you can actually understand.

[00:06:28]>> Do we go right or left? I don't know.

[00:06:30]>> I see the word park.

[00:06:32]>> I know. I go right because of this.

[00:06:36]Okay. So, we kind of went into the lane.

[00:06:39]>> Yes. It's a very narrow lane.

[00:06:40]>> It's a narrow lane. There's barricades on the right side and we're kind of driving >> in the middle. Barricades on the right and we're driving in between both lanes.

[00:06:50]>> Oh, look at this.

[00:06:52]>> All right, so we are taking a very sharp left turn right here. Scooter's coming at us.

[00:07:00]We're going under a tunnel. It's dark.

[00:07:02]Again, no LAR in here, so it's camera based, which is actually the same with the way Xpang does it. They both use camera only. So, it's a vision based system for both systems.

[00:07:24]It feels like we're going the wrong way, but I know we're not. It's just a narrow road right here. A moment years ago when Tesla first came out with >> somebody somebody considered jaywalking and it's actually trying to decide if this person is going to jaywalk. Look at that. See, a human driver would be like, "Sorry." Well, unless they're super nice, they like, "Sorry, you can wait for me." Where the car kind of read the mind of a person.

[00:07:49]>> Okay. This is also car's coming. It's narrow. We have a vehicle on our right.

[00:07:55]>> Okay.

[00:07:56]>> Yes.

[00:07:57]>> Figured it out.

[00:07:59]>> You can see how much reasoning is involved in in the Tesla.

[00:08:02]>> Person bending over middle of the road.

[00:08:05]>> Middle of the road cleaning the road.

[00:08:07]And >> oh my god, then they're not going to move.

[00:08:10]Wow.

[00:08:11]>> Will the car go or do I need to take a car? Let's let the car do its thing.

[00:08:16]>> I feel like the car wanted to do that, but it's obviously also noticing that people are going.

[00:08:20]>> Yeah, but people behind me are getting >> Okay, we might have to disengage.

[00:08:22]>> And he's still in the middle of the road. Okay, we're easing forward. The car is easing >> cuz he finally moved his hose.

[00:08:28]>> Okay.

[00:08:30]>> Wow.

[00:08:32]The Tesla.

[00:08:33]>> All right. I didn't disengage, but maybe I should have that time.

[00:08:35]>> That was probably a disengagement worthy.

[00:08:38]>> Yes.

[00:08:40]more polite than most Chinese drivers.

[00:08:43]>> So, now let's do the identical route with the Xpang P7 on his VLA 2.0 self-driving.

[00:08:49]>> Okay, this is my first time driving in an Xpang in general. So, I don't know all the different commands. It's a little bit nerve-wracking when you're unfamiliar with the car, also unfamiliar with the lay of the land in general. Also, you have to keep your hands on the steering wheel, which is kind of a squirrel design.

[00:09:17]>> Yeah.

[00:09:20]>> So, there's voice control in this car, but it only speaks Mandarin.

[00:09:25]>> Oh my gosh. It's >> um >> We need to make a left coming up. And it's it's fine.

[00:09:30]>> It found an opening, but it felt very aggressive the way it took that opening.

[00:09:34]But maybe that's just the way you have to drive here. Um, the massaging seats are on, so they actually feel really nice. But maybe I need to be paying a little more attention instead of falling asleep.

[00:09:49]So, we're doing the identical loop, but with NGP versus FSD.

[00:09:56]>> Yeah. I'm most curious on how it handles stopped roadway situations where, you know, the Tesla seemed like it played it a little too nice and people were getting upset.

[00:10:07]I want to see if this actually goes around obstacles a little more aggressively.

[00:10:15]>> So, here >> right right here it right around the scooter.

[00:10:28]I know with this system too, I can interact with it and if I need to turn, it won't completely take me out of NGP. Where with Tesla, if I was to turn slightly, you'd automatically be booted out. It'd be a disengagement. where this one I can kind of interact with it a little bit more which I now understand would be nice in this kind of environment.

[00:10:53]>> This is where it was I think here we had to make it turns. Yeah. This was where the Tesla had Yeah. Here comes some challenges right here. People >> on their phones >> on their phones. People don't care.

[00:11:06]But I mean >> it it went through without the hesitation that Tesla had >> to where we got honked at quite a bit at that intersection. So here it kind of went with the local culture of thread the needle.

[00:11:21]>> Yeah.

[00:11:23]All right.

[00:11:26]>> People on the street >> tunnels. We have some shadows.

[00:11:31]As we go into this tunnel, it's going to get dark. Again, these are vision based systems. We're taking a right.

[00:11:39]Okay.

[00:11:45]>> The Tesla also kind of came in through this lane. I remember >> this was the road where the Tesla just kind of drove right through the center of it, which is what we're doing. It's very similar approach.

[00:12:07]We're going to take a sharp left turn.

[00:12:09]You can't see what's on the other side right now. There's no LAR in these vehicles either. So, it's making that turn. Thinking about it.

[00:12:21]It's dark now. It's an environment.

[00:12:23]Notice the headlights just turned on for a second.

[00:12:35]Okay, we appear to be slowing down a little bit right here. And I believe it's because of the corner and it can't see what's happening. And now we have scooters and people look like schools getting out right now.

[00:12:51]Some students.

[00:12:53]This was also parked on the side street earlier. Maybe not this exact one, but >> was able to go around it. No hesitation at all.

[00:13:09]I'm scared to jinx myself if I say like it's doing a good job, but so far, knock on wood.

[00:13:19]Okay, here's one of those intersections that I don't quite understand because >> it's it's cruising.

[00:13:27]>> Everything is going at us at all directions.

[00:13:34]>> Do you see how this vehicle kind of continues moving through? At least our our experience so far. It doesn't just um it continues with the flow of traffic. So now this guy's trying to make a left.

[00:13:50]>> So far it definitely feels like it's been trained in this chaotic environment.

[00:13:55]>> Yeah. Now if this was a human driver, I would imagine the honking would have started to give this guy even more urgency to scoot through.

[00:14:04]Look, there's somebody behind you now trying to navigate >> with a much I don't even know if there's enough room.

[00:14:10]>> Got a larger vehicle, too, it seems.

[00:14:12]>> And they're trying to get around.

[00:14:13]>> And they're all They're all looking at us like, "Why are we not trying to squeeze through?"

[00:14:16]>> No, it is trying to >> No, it's >> So perhaps I should Okay, I'm going to put my foot on the accelerator.

[00:14:22]>> Yep.

[00:14:23]>> It's not very similar to the Tesla then in that scenario where >> it was more patient than a human.

[00:14:34]Okay, people on the right. Feel like it got closer to those people than Tesla would.

[00:14:42]>> It also has obstacles on the left to try to dodge >> every turn. I'm like, how is it going to handle this? How would I handle it?

[00:15:02]>> It's just wild to me that people Oh gosh, we're so close.

[00:15:06]>> That was close.

[00:15:08]>> The people would ride these scooters without a helmet.

[00:15:10]>> I know. I'm thinking the same thing.

[00:15:12]>> It's like knowing how many cars you have to get within centimeters of it. Oh gosh, this guy is going the wrong direction in my lane.

[00:15:23]>> All right, I don't >> Look at this like four-way roundabout without the roundabout.

[00:15:29]>> Now turn left >> and no lane markings.

[00:15:31]>> Charging stations on this the curb.

[00:15:33]That's an interesting little >> You don't see those in the States.

[00:15:37]>> I believe there is a company that was putting them on.

[00:15:39]>> I remember hearing about them, >> but not there yet.

[00:15:45]Now, when full self-driving first came out, didn't was it someone maybe it was the Chinese that said it would never work on these kind of roads?

[00:15:54]>> Yes, it was some some CEO of a Chinese car maker, maybe Leato or one of them that said that it's a lot more chaotic in the back roads of China. Um, and I mean, we're getting to see >> a level of chaos that you definitely don't I mean, maybe in New York City, but certainly not to this extreme where you've got, you know, scooters jumping out at you. And I think both of us have established just having a version 14 FSD on our brand new Model Y and a version 13 and our Model X that's similar to this. Out of the hundreds of cars we've tested and the dozens of self-driving comparisons we've done between all of them. Even the older model Tesla in version 3 is still couple of years at the very least ahead of the next best.

[00:16:41]>> Yeah.

[00:16:41]>> Self-driving. So, I'm super curious to see here in China how Xbang does compared to this version of FSD.

[00:16:50]>> This is NGP. It stands for nextgen pilot. It's the FSD equivalent. All right.

[00:16:57]Need to focus.

[00:17:00]Oh, wait. NGP is doing that for me.

[00:17:03]Okay. Students on the right.

[00:17:07]I don't know what this is on my left.

[00:17:09]Scooters all around me.

[00:17:11]one of the most chaotic environments you could possibly test a self-driving car in. And we're doing it.

[00:17:24]>> And while Xbang CEO is a big fan of Tesla, saying he's personally tested its FSD V14 in Silicon Valley and called it near level 4, he's so confident Xpang can beat Tesla's FSD in China's chaos today, he invited Musk to come see it for himself. But what happens next completely caught me off guard.

[00:17:42]>> And I test cars all the time. I test the self-driving tech that we have available in the US. And Tesla is definitely the top. It's the best.

[00:17:55]But I've never been able to test a Chinese-made cars self-driving.

[00:18:02]>> See, it wants to go around this person, but there's just not enough room. and a and a human might just create room by making that guy uncomfortable. Um, the Tesla's being very polite.

[00:18:16]>> Now, that would be a scenario where it would be nice to be able to stay in full self driving >> and kind of, you know, edge it forward a little bit.

[00:18:25]>> Yes.

[00:18:26]>> All right. So, now we are stopped, but I feel like we could go. I'm not sure why.

[00:18:30]It's >> Somebody's going backwards. People are mad at you for not going back.

[00:18:33]>> I know. They're They're honking at me saying, "Hurry up." Okay, I think I need to go. I think I need to go. All right.

[00:18:40]>> I did put my foot on the accelerator.

[00:18:43]Didn't boot me out of full self-driving.

[00:18:45]So, you can speed up and slow down with full self-driving, but if I was to stop or turn the steering wheel, it would take me out of full self-driving.

[00:18:55]>> Can you just take a moment and appreciate the complexity of what the car was going through in that environment? And again, right here, super small streets and people just darting out in every which direction.

[00:19:06]This is >> It thinks faster than I can think.

[00:19:09]>> Yeah. Well, as a Tesla fan, I'm really impressed to see it performing in this environment. But yeah.

[00:19:16]>> Okay. Why are we not moving again? It's, you know, it has to take a second to be able to do the reasoning before it can react. And it definitely takes longer here on these roads than it does in the US. Yes, >> these roads are a lot more complex.

[00:19:32]>> Oh, wow.

[00:19:33]>> I'm not sure what happened there.

[00:19:35]>> The kind of Are we at point?

[00:19:38]>> Not quite yet.

[00:19:39]>> Not quite yet.

[00:19:40]>> I think it's just unsure.

[00:19:42]>> It's kind of giving up on us here. We're not >> Okay, I think I have to take over cuz I'm going to get hooked up.

[00:19:47]>> We weren't at our way point. The waypoint would have been here. That's interesting. Okay, so I would continue straight. Let's go continue trip. Am I supposed to go straight still or am I supposed to turn there?

[00:19:59]>> I don't think you're cuz that's closed off.

[00:20:02]>> Okay.

[00:20:02]>> People are not happy.

[00:20:04]>> Continue.

[00:20:04]>> Yeah. I don't see continue on the waypoint though. Um when I hit the dock.

[00:20:09]>> Now make a U-turn.

[00:20:10]>> Oh, I think I missed it.

[00:20:12]>> No, because it was up on a >> No, I we were supposed to take a right there, but I couldn't tell.

[00:20:16]>> Is that accurate? Were we supposed to turn there or continue here?

[00:20:20]>> No, it's going to go.

[00:20:21]>> Making a U-turn, but it's too narrow.

[00:20:23]>> It's too narrow. All right.

[00:20:27]>> Oh, there's a car coming the wrong direction.

[00:20:30]>> And the car wants us to make a U-turn because I think I missed.

[00:20:35]>> It's going to I think it's going to continue the right way. Maybe not.

[00:20:38]>> Maybe. Let's see if it can figure out to reroute us.

[00:20:40]>> I mean, these streets are so abnormal for what we're used to.

[00:20:46]>> Yeah.

[00:20:46]>> Now, make a turn.

[00:20:47]>> It It wants to go there, but the map has >> It wants me to do your turn.

[00:20:51]>> Is it possible to do it? I will try.

[00:20:54]Gonna have to manually do that. That's scary.

[00:20:57]>> Are you gonna go the other way, Kim?

[00:20:58]>> Yeah, I have to do a U-turn going the other direction here. I'm gonna have to figure it out.

[00:21:04]>> Um they said they got insurance on us, right?

[00:21:10]>> Yeah. Yeah, we did.

[00:21:11]>> Okay.

[00:21:13]>> I think I'm going turn on my Okay.

[00:21:17]>> It's nice that we have all the cameras outside so people can tell we're up to something.

[00:21:21]>> Yeah, I might. I don't know if it's going to be able to make this narrow return.

[00:21:24]>> Make a turn.

[00:21:25]>> I'm trying, guys. I The problem is I'm used to I'm not driving like a >> I'm not driving like the Chinese backups, though.

[00:21:33]>> I'm definitely sorry.

[00:21:36]>> Oh, these scooters. How do you drive with them?

[00:21:39]>> They're insane.

[00:21:40]>> You're fine. You're fine. You're fine.

[00:21:41]All right, you're fine.

[00:21:42]>> Okay, got it. Oh, but then this guy.

[00:21:43]>> Oh, wow. He didn't >> really like they're just >> Wow.

[00:21:47]>> Oh my goodness.

[00:21:48]>> You're good. You're clear. Oh, this guy's trying to come through. Seriously people, >> if if there is a millimeter, they will squeeze.

[00:21:54]>> That is insane that they would even try.

[00:21:58]>> I feel like there needs >> You did a good job. I may have hit a couple people in that scenario because I would not have expected people to come through those spots.

[00:22:08]>> Good job.

[00:22:09]>> Wow.

[00:22:09]>> Now turn left.

[00:22:12]>> Okay.

[00:22:13]>> The scooter population here is insane.

[00:22:16]And they don't follow any kind of rules.

[00:22:19]No, >> they just, you know, whatever way they don't even they'll drive the wrong way.

[00:22:23]>> So, this guy is nice. He's letting you go.

[00:22:25]>> Yeah.

[00:22:26]>> And FSD will >> That was very nice.

[00:22:28]>> Destination is on the left.

[00:22:30]>> All right. So, this is our next way point.

[00:22:32]>> So, will it hit the >> Let's see.

[00:22:34]>> Yeah, I think it will.

[00:22:36]>> Interesting.

[00:22:38]>> So, again, >> it didn't slow down the way it should.

[00:22:40]>> Slow down the way it should, >> which we've noticed in version 13 on our Tesla at home, it doesn't slow down where 14 does.

[00:22:47]>> Mhm. So, that might be part of it, >> right?

[00:22:55]>> I guess it's going to do a U-turn and come back down that road to go on the next waypoint.

[00:23:00]>> Okay, let's see this U-turn on FSD.

[00:23:06]>> I don't think I've ever done a U-turn on FSD.

[00:23:10]>> I don't know if it's um Yeah, this might be a Yeah, I don't know.

[00:23:15]Let's give it a second and see what it does. Maybe it's looking for a better spot.

[00:23:20]>> There's no one here, though. Honestly, >> there's no one here. I had plenty of chances.

[00:23:23]>> Like, why is it not Is it going to >> Cuz the map makes it look like a big wide U-turn.

[00:23:29]And that would have been a much better spot for it.

[00:23:32]>> I want to give it a second. Let it breathe and see what it does here.

[00:23:37]>> It's trying to find a little area to tuck into.

[00:23:41]>> Okay. So, we'll do this. Okay.

[00:23:46]So, >> it's definitely taking its time to think about it.

[00:23:51]>> I'm really enjoying seeing this car struggle and try to figure out.

[00:23:54]>> I know. I've never seen it struggle like this before, but >> it has.

[00:23:59]>> I'm like so tempted to take over >> because like people don't they don't they'll just go around the road more. I thought we're supposed to be doing like a U-turn the other direction.

[00:24:09]>> What is it doing?

[00:24:10]>> Okay, I'm going to have to take over.

[00:24:11]This is like I think the car is confused with what?

[00:24:15]>> So, there's nobody there.

[00:24:15]>> There's nobody here. This is a lot easier than the last U-turn I did.

[00:24:20]>> Way point was down by the gate. And then I guess in in a perfect world, it should have turned around and continued on that same road we were on.

[00:24:27]>> Okay. Driving so slow on this road.

[00:24:29]>> It is. And it's on hurry mode. I think see if we go to autopilot. Yeah, it's on the fastest profile.

[00:24:37]>> I'll just get ready >> now. Turn right.

[00:24:40]>> It's going slow enough here to >> Okay, I did see it. It registered.

[00:24:53]I think it saw her coming.

[00:24:56]Oops.

[00:25:12]Okay, they just kind of came out of nowhere, too. They were hiding behind.

[00:25:19]>> All right.

[00:25:22]>> It would be exhausting for someone like us to try to actually drive on these streets and navigate to a destination on your own. I don't know how people >> because there's just thousands of decisions every minute >> just to keep you alive or keep someone else alive around you.

[00:25:52]>> Okay, there's a guy crossing the street.

[00:25:59]What's funny is when you and I were crossing the street for normal drivers, it definitely we didn't find too many people that would wait this long for us.

[00:26:07]>> I know.

[00:26:08]>> As you can see, nobody's waiting for them. They're just going through them.

[00:26:10]>> So, even in hurry mode, it's not fast enough. I'm going to give it a little help here. My foot's on the accelerator, but it's still in full self-driving.

[00:26:25]Okay.

[00:26:25]>> Okay. Cars in our lane coming straight at us.

[00:26:28]>> In our lane, >> I guess scooters.

[00:26:31]>> Yeah.

[00:26:32]>> Oh, this is insane.

[00:26:38]>> Oh my gosh.

[00:26:41]See all that. Do you see how that scooter is riding around people?

[00:26:44]>> Yeah. We have a bike rider. Now we are in lane.

[00:26:48]>> Opposite lane.

[00:26:49]>> I mean, I guess it is driving like the locals would in those scenarios.

[00:26:54]It's a lot more courteous than the local drivers.

[00:26:57]>> It is more patience for sure.

[00:26:59]>> Um, yeah, there's scooters on either side of me.

[00:27:07]>> I'm sure the cars behind me are getting annoyed, but they can tell that we are in self-driving because of the blue lights, and they're probably very annoyed.

[00:27:20]They're like, "Oh, it's one of those cars that follows every rule."

[00:27:25]>> Okay. So, I feel like Chinese drivers would have just gunned it and gone for that.

[00:27:32]>> I mean, look at this. Look how many >> Yeah.

[00:27:34]>> red light runners you'll see on >> on non fourwheels.

[00:27:38]>> And I say that, but you kind of have to be an aggressive driver here or else you will never get anywhere because everyone else drives like that.

[00:27:47]So, as you'll notice, the rules definitely don't apply to to scooters.

[00:28:01]It seems to be confidently navigating this scenario with all these people and scooters and cars coming right at us.

[00:28:16]Yeah, the streets are extremely narrow and the obstacles are extremely frequent.

[00:28:22]So, I mean, in these like blind corners, the human driver would not be as cautious as this. I mean, you saw the scooter just go full speed around the corner.

[00:28:33]>> Um, And you've got an old lady with >> Yeah.

[00:29:05]>> some sort of a dolly that is completely on the road.

[00:29:09]>> Yeah.

[00:29:10]>> She has no interest in moving her dolly.

[00:29:14]>> Oh my god. I got scared.

[00:29:16]>> Was that you or >> That was me. That was me. I got a little bit nervous because the dolly is facing.

[00:29:23]>> Is it still in self-driving?

[00:29:24]>> No, now I'm driving.

[00:29:25]>> We can restart it.

[00:29:26]>> How do you restart it?

[00:29:27]>> Pull down twice.

[00:29:29]>> I just wanted to get around her cuz that dolly is >> I don't think people behind you noticed that dolly's there.

[00:29:35]>> And GPA back. Okay, so it's back on.

[00:29:38]Sorry, I got a little nervous.

[00:29:40]>> Yeah, the the people behind you have not noticed this lady with a dolly does not care.

[00:29:48]Uh like these. It's so scary.

[00:29:55]>> It I feel like every turn is like you're on a ride at Disney or Universal and it's the type of ride where you almost want to like close your eyes cuz it's so scary, but you think like, you know, you're going to get hit, but then you like go underground or something. But with this one, you're like, "No, this is really happening." And >> this guy's on his on his cell phone onehanded scooting.

[00:30:22]So, we're picking up some speed here.

[00:30:24]>> Mhm.

[00:30:24]>> Look at this. There's kids crossing the road coming up. Oh my gosh. It's not slowing down. It's just threading.

[00:30:32]>> You go between wanting to slam on the brakes and wanting to see what it will do and how it will handle this scenario.

[00:30:39]It's a constant like battle in my head like how far do I let it go?

[00:30:43]>> I'm noticing situations where it definitely drives more like a human like the locals and then other situations where it does what the Tesla does and it becomes extremely cautious.

[00:30:54]>> It's a matter of trust though because this is new to me. It's also a newer system.

[00:30:58]>> Yes.

[00:30:59]>> Which is very impressive that it's been able to do what it can do.

[00:31:02]>> Yeah.

[00:31:04]I mean, even already, I would I would already go out and say that no other system outside of Tesla we've tested in the United States, >> no, >> has has this level. This seems if you were >> kind of put them side by side so far, I would say it's very similar to what, >> you know, we always talk about the gold standard of self-driving.

[00:31:24]>> This is definitely, you know, right up there with Tesla full self-driving. So far, they've been very equal. And what I think is really impressive is that most car companies that I've worked with in the past, they don't want me to compare to Tesla.

[00:31:41]>> We made that observation.

[00:31:42]>> I right away said, "No, you're going to test it with Tesla's full self-driving.

[00:31:48]Compare them. That's what they want me to do because they're that confident."

[00:31:53]Which is impressive. Like that says something right there.

[00:31:59]>> Oh my gosh. This guy is completely in our path.

[00:32:03]>> Did you break at all?

[00:32:04]>> No, I didn't. That was the car. It did a great job.

[00:32:07]>> By now, you can see how capable Xping's VLA 2.0 is. Arguably, even more in this environment than Tesla's FSD. It anticipates stops and flows around obstacles in a way we've rarely seen outside of Tesla.

[00:32:20]>> Interesting.

[00:32:22]>> So, we are threading the needle right now, but it's doing a, you know, it's more confident than I am in this scenario. There's a guy on the street that hopped off a bike.

[00:32:34]>> In a lot of ways, I'm I'm relieved that it's a confident driver because I don't know if I would be. I I definitely wouldn't be.

[00:32:43]>> So, now it's got to get all the way to the right.

[00:32:46]>> There it goes.

[00:33:04]I feel like in this video I want to be able to compare side by side Tesla versus this, but at the same time I feel like we need to release the full drives of both both vehicles in this chaotic environment cuz people need to see what it's like. Whoa. We are very close to that car.

[00:33:27]>> Oh my gosh, there's a bike right next to us.

[00:33:32]>> I mean, the situational awareness that's needed.

[00:33:35]>> Yeah.

[00:33:36]>> To know that there's a car touching your headlights almost up there where there's a bike touching your tail lights over here. And then to be able to watch out for both of them.

[00:33:50]>> Okay. So, it definitely can think faster than I can.

[00:34:03]>> All right. So, I will look for a good opportunity to test out co-pilot up here.

[00:34:09]>> So, what's co-pilot again?

[00:34:10]>> Co-pilot is where you can move the vehicle within NGP and it doesn't take you out of NGP. Okay.

[00:34:17]>> So, you can interact with it. So there's a different setting you have to engage.

[00:34:22]>> I think you can just do it. So if a scooter is going by you and you want to react faster, >> you could turn.

[00:34:29]>> So right now there's a bus stopped here.

[00:34:31]So if you wanted to change lanes yourself, >> I believe I can.

[00:34:34]>> You won't disengage it.

[00:34:36]>> So I'm turn on my blinker right now and then like see.

[00:34:39]>> So okay, >> I just turned >> and if you let go, you're resuming. Wow.

[00:34:43]>> And then I resumed. That's brilliant cuz in the Tesla it's pretty stiff and you feel a sharp disengagement.

[00:34:51]>> Exactly. That would have disengaged full self-driving but here I was the co-pilot.

[00:34:57]>> Okay.

[00:34:58]>> That's that's what it's called.

[00:35:01]>> I like that. That's a really great feature.

[00:35:03]>> Lucid has that with itself driving. I've seen some other cars do that as well.

[00:35:08]>> The prom starts with P >> D H U D H U D >> H U D.

[00:35:22]>> You did it.

[00:35:24]>> You speak Mandarin.

[00:35:25]>> You didn't know all kinds of things about me. Oh, this is great.

[00:35:28]>> Yeah. Can you see it?

[00:35:29]>> I can. It's great. It's very like vibrant.

[00:35:39]Do you see how the screen tilts towards me? Look it.

[00:35:46]Okay, we are turning.

[00:35:50]Here is our U-turn.

[00:35:52]>> Do you see our scooters?

[00:35:54]>> Wow. Wow.

[00:35:57]>> Oh my gosh.

[00:35:58]>> They just don't end.

[00:36:00]>> I mean, they outnumber cars. There is no way that the Tesla would have been able to do that that quick. For sure. We would have been back there still waiting.

[00:36:10]>> Yeah.

[00:36:10]>> On the scooters before >> going out. This definitely reacts faster.

[00:36:16]>> Yeah. And has Scooty senses.

[00:36:20]>> Crazy that it's able to do that too without LAR.

[00:36:23]>> Yeah. And you'll notice in China, they've gotten apparently our previous driver was telling us that for about 10 years now, this technology where they'll tell you how many seconds this light's going to be green for and how many seconds the next light that is red, for example, has remaining on red before it turns green.

[00:36:41]>> So, you had about 30 seconds of green left. And >> and even some of the lights had the counter.

[00:36:46]>> 82 seconds. Yeah, the lights themselves will have a counter on them, but right now we've got 78 seconds before that light turns red. And he laughed at us when we were like making such a big deal out of it because he said that they've actually had this in China on cars for like 10 years.

[00:36:58]>> Oh, look at this.

[00:36:59]>> Wow. This car is >> Oh my I'm shocked there aren't more.

[00:37:05]>> BC3, not a BC4.

[00:37:08]>> I'm shocked there aren't more accidents because everyone is playing a game of Tetris with their bodies.

[00:37:16]>> So, it wants to do a U-turn here. I don't see how it's going to do that.

[00:37:23]How? There's no >> It should have gone there.

[00:37:26]>> Was it supposed to go that way?

[00:37:28]>> I think it was.

[00:37:29]>> So, it kind of missed its own turn. Now >> they left in 200 m there left.

[00:37:37]>> Okay. I would >> Now we're tracked by >> um H.

[00:37:47]>> Okay.

[00:37:49]>> Yeah. It's going to try to go.

[00:37:52]>> Try to see if it just let it go.

[00:37:55]>> Yeah. I mean, if it if it thinks it's >> Oh my gosh. It did not expect that car to come there.

[00:38:01]>> Is it supposed to do a U-turn here on a highway?

[00:38:05]>> At some point further up. I think there'll be another way.

[00:38:08]>> Okay.

[00:38:11]Take a deep breath.

[00:38:13]Take a deep breath.

[00:38:16]>> Now they're left.

[00:38:17]>> Oh my goodness. That's full self driving you crazy.

[00:38:21]>> It is.

[00:38:23]>> It is so stressful.

[00:38:26]>> I mean, we've seen Tesla's presentations on autonomy day and, you know, AI day where they show these uh long tail cases and the extreme cases that they have to train on and the footage they use where a duck, a family of ducks, you know, goes in front of that pales in comparison to some of the things that we just encountered. I've never seen take every single >> longtail case we've experienced self-driving for the tens of thousands of miles we've self-driven. I don't think it'll compete with the last 15 miles of self-driving.

[00:39:01]It's signaling right, but it needs to go up. Maybe >> this the way point.

[00:39:05]>> It's another one of those.

[00:39:06]>> Now make a >> It should be going up there. Yeah. Oh my gosh, her shoes came off.

[00:39:12]>> Her shoe just came off.

[00:39:13]>> Oh man. Oh no. She hit the bump and almost lost her balance that she's got a whole bunch of eggs in one hand.

[00:39:18]>> And >> a what a sweet little boy with ice cream in his hand.

[00:39:23]>> I wouldn't put the shoe on for her.

[00:39:25]>> Miss our son.

[00:39:26]>> Okay. So, I had to disengage again. I think I've lost count with the amount of times I've had to disengage at this point.

[00:39:33]>> I think we're about between four and five. Yeah. But do you count those disengagements? Because those are kind of like way point disengagements.

[00:39:40]They're not >> those are not >> Yeah. But I think at least four were I mean these streets we have no words to describe like we call it roundabout or four-way stop. This is like a 16-way roundabout but it's neither.

[00:39:56]>> It's just they're not we don't have roads like this in the states.

[00:39:59]>> And then to factor in the bikes and the scooters and the you know helmetless riders and >> could you imagine a Whimo >> something with like a pre-mapped >> a premapped system >> system would not be able to handle anything like this. That's true.

[00:40:13]>> So, we are turning left.

[00:40:15]>> Really no well marked lanes in these areas. I mean, the convergence of 24 vehicles.

[00:40:23]>> All right.

[00:40:26]And then, >> and they're going the wrong direction, too. That one just came. This >> the bicycle.

[00:40:34]>> On the bicycle. Gosh, >> this will be the most difficult part.

[00:40:40]>> Okay.

[00:40:41]>> Okay. So, two hands.

[00:40:49]All right. I was just told this is the most difficult part of the trip right here. Two hands on the steering wheel.

[00:40:59]>> I mean, yeah, the people walking across the street, but there is no urgency. There's no fear.

[00:41:06]>> There's no concern.

[00:41:08]They >> Well, they're used to it.

[00:41:10]>> They're used to it. I mean, look at this.

[00:41:12]>> Look at the sky is just in the middle of the road.

[00:41:13]>> Suicidal.

[00:41:14]>> It really feels that way.

[00:41:17]>> Looks like >> What is happening? Oh my god.

[00:41:21]>> But I need to I need to move right now.

[00:41:23]I have to move because >> Okay, so we're going to disengage.

[00:41:26]We're creating an hazardous situation.

[00:41:29]Be nice.

[00:41:30]>> It It took too long to think about what was going on and I had to >> look at the map. Look at the center screen. Oh my gosh.

[00:41:37]>> Okay, I'm back in now. It's like raining ants on the center screen.

[00:41:43]>> There's so many pedestrians.

[00:41:46]>> Okay. I Oh gosh. Sorry. I disengaged there cuz I got scared. But I think the car would have >> been able to do it, but I just got nervous. I don't want to hit anybody.

[00:41:55]>> Just re-engage that. I want to see this.

[00:41:58]>> I feel so tense.

[00:42:00]>> Mhm.

[00:42:02]>> I don't even want to look down in it cuz I feel like I need to be paying attention.

[00:42:05]>> On the left.

[00:42:06]>> Okay. Your next way point's 300 m on your right. So you'll just Oh, don't don't disengage.

[00:42:14]>> I'm going to Yeah. Not quite yet.

[00:42:16]>> Is it stopping by itself? Is it sensing?

[00:42:18]>> Yeah.

[00:42:19]>> I think it's also seeing scooters around us. So, it's wanting to take those into account before it accelerates.

[00:42:26]>> Yeah.

[00:42:31]>> Okay.

[00:42:35]Deep breaths.

[00:42:41]I think I'm going to go to the hotel spa and relax after this. Oh, I like want to close my eyes cuz it's so scary.

[00:42:52]>> Um, I'm I'm speechless. I've never seen anything with this much complexity.

[00:43:00]I >> mean, look at this.

[00:43:01]>> Look at this. This is the >> Holy cow. This is the suicidal street.

[00:43:07]>> I mean, it's right on the turn and the suicide attempts begin.

[00:43:13]>> Japan has the suicide forest and this has gosh >> the suicide street.

[00:43:20]>> Look, they start before it turns green.

[00:43:23]I mean, rules don't apply.

[00:43:27]No fear for your life.

[00:43:31]No fear.

[00:43:35]Reminds me of like when I was a cheerleader. There was a song or not a song, a cheer. It was a beat.

[00:43:41]Aggressive. Got to be aggressive. Got to be e A gear.

[00:43:48]Okay, they're all coming.

[00:43:49]>> But did you see it start moving with a countdown on two? So, it actually anticipated the light turning green, >> which I imagine is how it's been trained with people in this environment. They don't wait for for green. You can't wait in this environment or you'll be waiting forever. You'll never get anywhere.

[00:44:06]>> Okay, this >> Okay, I didn't I didn't interact.

[00:44:12]>> That was it. Okay, >> that was it. That was it. I was had my foot about ready to break, but it it was able to do it.

[00:44:19]>> It would have gone until that person committed >> to stepping in front of it, which again is how a human probably would react in this environment. Maybe not a human, you know, state side, but in this environment, you just got to thread. They thread.

[00:44:40]>> I was actually more scared in the Tesla and full self-driving. I can't believe I'm saying that, but I was more scared on this road with the Tesla than I than I feel right now cuz I feel like it's, you know, equipped to be able to handle this kind of environment.

[00:44:57]This has got this has got to be one of the most complicated streets on the planet when it comes to the edge cases that are thrown at it.

[00:45:08]>> I mean, the center screen says the entire story.

[00:45:12]>> Actually, not even because I feel like there's more going on out here that's even on here. It's basically like driving in autopilot in Time Square.

[00:45:22]Yeah, >> I think this is more intense than Time Square, though.

[00:45:28]>> It is more intense >> because there's the amount of rules that are technically being broken with, you know, with the bikes and scooters just crossing, going through red lights.

[00:45:38]>> I really don't know how it's doing this.

[00:45:42]>> Wow. Look at this. It's like hundreds of scooters going every direction, coming right at you, people walking.

[00:45:51]They just have no regard for their life.

[00:45:54]>> Phone. Yeah.

[00:45:55]>> At all. They're just >> the mirrors of the scooter hit that lady's arm.

[00:45:58]>> Do people not sue each other in this country?

[00:46:02]>> I don't think that's a thing.

[00:46:06]What an American question, Kim.

[00:46:10]>> This car is cutting in.

[00:46:13]Okay. Oh my god, we are so close.

[00:46:20]I was not expecting this at all.

[00:46:23]>> It would be fascinating to get some of the vehicles in the states that we have access to and run the same exact >> They wouldn't even make it out of the parking lot.

[00:46:32]>> Yeah, there's no way they wouldn't do it. They wouldn't be able to do it.

[00:46:39]>> See how it's not slowing down? But clear as day, it sees it.

[00:46:42]>> Mhm.

[00:46:43]>> But it's almost like, I'm going to do what I need to do, and you need to figure it out.

[00:46:48]>> And I'll only stop if I calculate.

[00:46:52]>> It's driving like >> It's driving like a human drives here.

[00:46:56]>> Yep.

[00:47:02]>> I mean, look at this. Look how many scooters.

[00:47:05]>> Yeah. We're We're going on to oncoming lanes to get rid of possible scooter collision chances.

[00:47:18]And it's just wild for someone like me who didn't grow up anywhere like this.

[00:47:22]I've never driven in this kind of environment. I've never seen roads like this before.

[00:47:29]And for it to be able to handle this, it's just like crazy. I wouldn't be able to I would not want to drive here without this.

[00:47:41]>> I don't know how it's able to take in so much information and react so quickly.

[00:47:47]>> It's all I mean look people are crossing the street >> at the same time and it's going with the flow of >> these people.

[00:47:55]>> This happens for example with this sort of people crossing in our neighborhood.

[00:47:59]It would stop in the United States. Um, we're here.

[00:48:03]>> Stops as soon as it sees the person that it thinks might be >> getting close to.

[00:48:08]>> But in this environment where there is so much chaos, >> you can't >> it's going with the flow.

[00:48:14]>> I know.

[00:48:15]>> And that's the reasoning element when it comes to self-driving that we've seen kind of take off in recent months is that you need reasoning. Um, >> but even Tesla in hurry mode, which is what we were just on, like look at right now. It's like >> it wants to get through. It wants to get through.

[00:48:34]And this, I guess, technically is its lane. I don't know why their scooters coming up. Is it his lane?

[00:48:38]>> It is a lane, but they're >> the arrows facing.

[00:48:41]>> It's a one way, but they're >> they're the ones they're the ones that are at fault.

[00:48:46]>> Yeah, they're going the wrong way.

[00:48:50]>> The majority of people are >> going the wrong way.

[00:48:53]>> Going the wrong way.

[00:48:54]>> They're going the wrong direction.

[00:48:56]They're This is the suicide road.

[00:48:58]>> Suicide road.

[00:49:06]Oh my goodness.

[00:49:11]These guys.

[00:49:14]This is probably one of the most impressive um experiments of physical AI that I've ever done, that I've ever seen in the last two hours. Being able to test it with Tesla and then test it um right here with Xpang has just been incredible.

[00:49:34]>> Yeah. I mean the whole physical AI perspective that people that engage with GPT on a daily basis and kind of think of how incredibly powerful and capable that is and then to take it to a physical realm and talk about the lawnmowers, the pool cleaners, the vacuums that we have in our house. Mh.

[00:49:51]>> Um, those are impressive, but there's still hiccups where they get caught in the corner of a rug. And to get caught in a corner of a rug in your living room and to be able to achieve what we saw this vehicle achieve in this sort of an environment where it's literally a life or death decision making that's been happening, you know, around us is is one of the most, like you said, impressive things that I've seen with physical AI.

[00:50:13]>> I think that's why these companies are not just vehicle manufacturers, they're AI companies.

[00:50:24]Okay, so this guy is wanting it and it was aggressive enough >> aggressive enough to keep him from coming through.

[00:50:30]>> And now look, it's trying to navigate around this lane cutter. Look at this.

[00:50:34]It's a quad lane cutter. He's cutting across four lanes and it went right around it. This is another one of those indescribable intersections where it's not a four lane or you know it's just I don't even know how many >> lanes and >> there's got to be at least 12 >> three four's five >> five intersections each with at least three lanes coming out of them.

[00:50:59]>> Mhm.

[00:50:59]>> So you you extrapolate 15 lanes that are crisscrossing >> scooters that don't follow direction.

[00:51:04]People on foot bikes >> everything.

[00:51:10]And there's your trash.

[00:51:13]The little scooters. Yeah.

[00:51:18]>> So, see, I wouldn't have even noticed that turn green cuz the light is half a mile away.

[00:51:24]>> You think >> it's being obstructed?

[00:51:25]>> Yeah.

[00:51:27]>> You think you've seen rush hour and then you've really seen rush hour.

[00:51:33]>> It's funny. I think so far I would give the edge to Xpong when it comes to how it's handled the different variances of going around people there. You know, initially had similar pauses >> on the road, but then over the last hour or so, I think it's definitely passed it up as far as being more aggressive.

[00:51:58]>> And you can't like fake this type of thing. You can't like say, "Oh, this was a setup. There's no way >> from the outsers's perspective, you think of being more aggressive or being more cautious in Tesla's case as possibly that being a positive. Um, in certain environments that might be the case. I think in this environment being extra cautious led to additional honking. The cops at one point were flagging us to move or you had to disengage >> and make that turn. Um, so in this environment being more aggressive, you know, >> you need to be more aggressive >> and assertive on those roadways is is more important.

[00:52:34]>> It's almost like you need something above hurry mode with Tesla.

[00:52:38]>> Mhm.

[00:52:39]>> Well, on a 13, I guess, if once you get into the 14, you have Mad Max and beyond just the 13 that is available in this country >> had that.

[00:52:48]>> It would be very interesting to try this out with Mad Max. Back in the Tesla, FSD drove well overall, but its navigation clearly lagged behind Xpang, even sending the car down roads it shouldn't have taken.

[00:53:01]>> This is incredible.

[00:53:04]>> Look at this. We've got >> lush and green right here, though.

[00:53:06]>> I know. We've got a cart of some sort stopped with now gravel road around a bridge >> around construction. I don't even know where we are. This is now as off the >> It feels like we're on like a back alley.

[00:53:25]>> Are we supposed >> Is this a waypoint or are we >> Are we off the waypoint?

[00:53:30]>> I don't know.

[00:53:31]>> Somehow we went off the waypoint. Oh my gosh. Now we're in a construction zone.

[00:53:34]You might want to stop it, but I don't think we should be here.

[00:53:36]>> Is this This isn't the right way, is it?

[00:53:39]>> But there was a couple where it just stopped in the middle for no reason. Or >> So that was a mistake where the navigation took us here. Okay. Oh, look at this. Oh my god.

[00:53:49]>> I'm going to take it off.

[00:53:50]>> That was a definite disengagement.

[00:53:52]>> Yeah, I had to disengage because >> I'm surprised it um >> Is this a one-way road or >> I would have seen >> that one.

[00:54:04]>> I thought I would have seen that cuz I mean you and I saw that from pretty far away.

[00:54:08]>> Okay.

[00:54:09]>> Okay. That guy's letting you go.

[00:54:10]>> Uh maybe not.

[00:54:11]>> Yeah, he's letting you go. That's nice.

[00:54:16]There are some nice people that are not self-driving but are still nice.

[00:54:19]>> Thank you.

[00:54:24]>> Yeah, this is very >> All the street cleaners are coming up this road and it's a narrow road, >> but I mean this is exactly what a human would do. Come >> on to the side and wait for your space.

[00:54:36]>> Is it going to go?

[00:54:37]>> Um, no. I'm No, it's not.

[00:54:38]>> I'm going to give it a little confidence here.

[00:54:41]>> Why is it scared to go? The bush is I think the bush is moving. We're giving it.

[00:54:45]>> I had to put my foot on the accelerator twice to kind of assure it to go forward there.

[00:54:52]>> In 200 m, turn left.

[00:55:00]>> Oh, >> okay. So, there's no reason for it to slow unless you know what it saw. These >> I bet you Yeah, there's another disengagement.

[00:55:09]>> I did.

[00:55:10]>> So, it's it's seeing things that maybe you and I are not seeing. So it's stopping and then when you kind of scan your surroundings you realize why because there's just >> you know >> Yeah. There's things happening behind us all sides of us like >> I mean right up in front complete blockage here with a a truck with what looks like >> trash truck >> maybe truck hit the side of the wall.

[00:55:34]>> Oh gosh. Did it run into the wall? It possibly is.

[00:55:38]>> I have no idea.

[00:55:39]>> And now the honking begins.

[00:55:42]And >> okay, so we've got >> There should be something wrong.

[00:55:46]>> Yeah, this is the ultimate. I mean, look at this cam.

[00:55:49]>> Do we go around it? Can we go around it?

[00:55:51]>> Yeah, we can because we don't >> I mean, everybody else is. Yeah. So, maybe Oh, the car is signaling. Is that you or the car?

[00:55:57]>> That's the car.

[00:55:57]>> The car is deciding. Maybe it needs to Let's see if these bikers clear if the car decides to move.

[00:56:02]>> I know there's another car coming out on the >> So, the car knows that it should start going around. what's happening, but it also sees other cars.

[00:56:12]>> So, it turns out the truck driver had realized his cargo wouldn't clear the low tunnel, so he had to try to turn around on this super narrow road.

[00:56:22]>> I mean, edge case after edge case.

[00:56:24]Seriously, >> this is unbelievable. So, the car remained engaged and dealt. I mean, the trash is rubbing against the walls now.

[00:56:33]>> Now, turn left.

[00:56:36]We're trying to turn left.

[00:56:38]>> Oh gosh, he's hitting a street sign. Oh, he's going to rip that sign right off.

[00:56:42]He is. Oh no. Oh, there goes trash. And more trash.

[00:56:48]>> Wow. Oh, is it going to hit right here?

[00:56:50]>> Oh no. He's got Oh, look at the cables on the top.

[00:56:55]>> Oh no. Look at this.

[00:56:59]The the street sign is >> with others like >> this is an extreme.

[00:57:05]>> Look at this intersection.

[00:57:07]>> It's unbelievable.

[00:57:10]The center screen is going bonkers. By the way, and like the people are just chill. Let go.

[00:57:19]>> Now I should go. And it's not going though.

[00:57:21]>> But it's trying to make a left though, Kim.

[00:57:22]>> Oh, it is. Oh my god.

[00:57:23]>> It's not being aggressive enough. And the cop the cop is saying go. Okay, maybe disengage.

[00:57:29]>> The cop is telling you to make your left.

[00:57:32]>> I'm scared.

[00:57:33]>> Okay.

[00:57:35]>> Can I go?

[00:57:37]>> You just >> I have to.

[00:57:40]>> So >> after this scooter >> after this? Yep. Be careful for your left though.

[00:57:46]>> Okay.

[00:57:46]>> Okay.

[00:57:47]>> Okay.

[00:57:48]>> Wow.

[00:57:51]>> Wow. Okay.

[00:57:55]So, if anything, the Tesla is a little tooid like timid for Look at this. Look at this.

[00:58:01]>> Like, go.

[00:58:02]>> What is this styrofoam truck?

[00:58:06]>> Oh, and that's barely staying in there, too.

[00:58:09]>> Okay. And all the scooters are It doesn't want to go.

[00:58:12]>> I again, it's just too timid where a human in this environment.

[00:58:17]>> I have to put my foot on the accelerator right now to get it to go.

[00:58:21]>> It doesn't want to drive. And everyone's honking. They're honking. They're like, "What's your deal?" Okay, so this is your next way.

[00:58:27]>> Destination is on the left.

[00:58:29]>> So, you can see how Tesla's hesitation and lack of direction created more issues at times than how a local driver may have handled it.

[00:58:37]>> There's another person just stopping on the side of the road.

[00:58:40]>> There's >> Oh, wow. Hazards are on. So, this car has to decide.

[00:58:44]>> Yeah.

[00:58:45]>> And cars.

[00:58:45]>> Now, the car behind me was able to get over.

[00:58:48]>> Yeah.

[00:58:48]>> Now, let's see if I am able.

[00:58:50]>> Do you want to wait for this? This is >> I'm going to do the co-pilot here.

[00:58:56]>> And >> nice. So, if you let go, are you back in? You're Yeah, you were never this.

[00:59:01]You were just controlling it briefly.

[00:59:03]>> That was really nice to be able to just co-pilot.

[00:59:07]>> That is a feature I could see Tesla bringing or adding. That should have been there from for a while now >> because disengaging honestly is kind of jarring unless you use the brakes to lightly tap cuz if you disengage by hand, you're >> jerking completely out.

[00:59:22]>> Um, and that shifts the car, rattles the car a little bit >> and it seems safer.

[00:59:28]>> Mhm.

[00:59:29]>> Yeah. This is the area where Tesla navigation failed and we ended up on some old dirt road and who knows where.

[00:59:46]This was the place the Tesla got stuck, by the way, on this little grass.

[00:59:51]>> This has peopleird and tight. Yes, >> that was so tight.

[00:59:55]>> Yeah, >> but it's still moving. We're still going.

[00:59:57]>> Yep. Tesla went over that bridge for some reason. It wasn't supposed to. It got stuck behind that.

[01:00:02]>> Yeah.

[01:00:05]>> Here we go. This is the the bridge that trapped the trash truck last time.

[01:00:10]>> Yes, you're right.

[01:00:13]>> I wonder if we'll see more debris.

[01:00:14]>> Oh gosh.

[01:00:16]>> There seems to be more chaos down there.

[01:00:18]>> Yeah.

[01:00:22]>> So, there's signage that tells you it's 3.3 m.

[01:00:27]So, like about 12 13 feet.

[01:00:31]>> These scooters are just >> These scooters have no patience.

[01:00:38]30 seconds of time saved over life.

[01:00:47]So this is the turn that the cops were yelling at the Tesla for being too shy.

[01:00:55]So >> I mean this is insanity this turn.

[01:01:00]>> I don't blame the Tesla because this is crazy. There it is.

[01:01:04]>> This is what the traffic jam was. Was this street is insane.

[01:01:14]>> These people are staring.

[01:01:16]>> I know. I mean, there's just So, there's that cop. Let's see if >> this is Yeah, the cop was saying you have to go.

[01:01:23]>> This has got to be the most complicated street on the planet.

[01:01:29]I mean, >> so in the US, you would have someone in the center directing traffic, >> but >> there is no direction. It's just >> no. He's like, I don't want to be in the middle of this chaos. There's no way.

[01:01:43]>> He's sort of there, but like we're not moving at all. I don't know if I should just go or what. I feel like I'm going to have to go.

[01:01:49]>> You might have to.

[01:01:50]>> I think I'm okay.

[01:01:50]>> Go co-pilot this.

[01:01:52]>> I am going to definitely have to co-pilot or we will never get anywhere.

[01:01:56]And this guy is hitting the trees as he comes through.

[01:02:03]>> Deep breath. And I'm just going to go for it.

[01:02:06]>> I'm proud of you, Kim.

[01:02:08]>> I would not wish this on my enemies.

[01:02:10]Should be careful with people on your left.

[01:02:13]>> This person's >> I'm trying.

[01:02:19]>> No, don't don't ignore them cuz you're going to make yourself hit people. Just ignore it. Go only when you feel like you are safe to go.

[01:02:27]>> I just have to go. I just have to.

[01:02:30]>> Gosh.

[01:02:32]>> Almost there. Almost there.

[01:02:34]>> You made it.

[01:02:36]>> I did.

[01:02:36]>> You just have to force your way in.

[01:02:39]>> Wow.

[01:02:40]>> There's no other way.

[01:02:42]>> Even the physical AI was like, "Sorry, dude."

[01:02:44]>> Yeah.

[01:02:45]>> This is you.

[01:02:46]>> So, we're back in NGP. Very useful, actually.

[01:02:49]>> Wow. Look at this, Kim. We >> got a massive semi. coming through in an ultra narrow path.

[01:02:56]>> Yeah.

[01:02:58]>> And I mean it it wouldn't even fit if we tried to go through. Let's see what it does. So a human here would would estimate that I got to late, you know, hold on for this guy to come through >> before I go.

[01:03:10]>> We're still edging forward slightly.

[01:03:12]>> But I think it hopefully looks like it calculated its room.

[01:03:18]There's another jwalker.

[01:03:24]Wow.

[01:03:26]Good job, guys, for for finding this street.

[01:03:30]>> You guys found the most complicated street on the planet.

[01:03:35]>> I like how it sped up there. There was no car there and it was sped up to get through.

[01:03:41]>> It does. I've noticed that, too. It has areas where it feels very confident going fast.

[01:03:49]Again, >> there's no lanes here. There's like not no lane markings whatsoever.

[01:04:16]All right. So, the bus We either wait for or we go around.

[01:04:22]>> I don't think I think we wait for it.

[01:04:24]>> He's trying to signal left though.

[01:04:27]>> No.

[01:04:29]>> Okay.

[01:04:48]the scooter's here.

[01:04:50]>> I know.

[01:04:51]>> Yeah. Now, for me, if I were to compare Tesla's full self-driving and Xbuns and GP side by side on a scale of 1 to 10, um, they were both extremely impressive given the incredible amount of challenges that were thrown at them. The Tesla, I would probably say seven out of 10. Um, just because of the hesitancy in certain areas. And you can see clear as day. this right here. It just kind of threads the needle and goes with what other cars in this environment would do. Um, I would say this edged it slightly. Probably an eight out of 10.

[01:05:30]Um, and once you look at the challenges that were put in front of it, I think an eight out of 10 is is one of the best scores you could ever imagine.

[01:05:38]>> I would fully agree with that. It's a little hard to compare 100%.

[01:05:44]>> The environment >> because the environment is different.

[01:05:46]We're also on a version of full self-driving that is different than what we have in the US. It'd be, you know, if it was version 14, we had Mad Max, all that.

[01:06:02]>> Okay.

[01:06:07]>> What's the price point of the P7? The starting price, >> it would be around $30,000 US equivalent.

[01:06:14]>> Okay. which is probably where the Model 3 starts here, right? Probably around 30,000. But this feels far more premium for sure.

[01:06:22]>> If it was sold in the US, it would be like 100% tariff and >> but still at $60,000. If it was 100% tariff, >> this is what I would expect. Honestly, this would be a very high 60,000.

[01:06:35]>> Yeah, it's a lot more premium with the massaging seats, the ambient lighting.

[01:06:39]You have a little display in front of you. There is a head-up. Xbay includes full self-driving in the price, while Tesla takes a different route, charging $99 a month for access. This was one of my biggest aha moments for us. Tesla has led for years, but that gap isn't universal. What XBank has done in just a few months is seriously exciting for the future of this tech. If you made it this far, hit subscribe. It helps us keep bringing you access like this, and we'll catch you next time.