Technology invades the modern world

Chapter 387 Typical Style, The Art of the Deal

"Is it the L5 autonomous driving we understand, or the L5 autonomous driving that the media has portrayed and shaped in the public's mind?" Musk asked.

The world's first L5, a product never before seen, is, as Musk just experienced, smoother than any currently available or laboratory-developed autonomous driving system. But Musk still wants to know exactly what kind of L5 it is.

The sudden emergence of the L5 is somewhat similar to a sixth-generation fighter jet. In the past, everyone had fifth-generation fighter jets, but suddenly a sixth-generation fighter jet appears. I can't say that it isn't, but I don't want to admit that it is.

"Understood? Shaped?" Lin Ran repeated before asking, "You mean it's the kind of autonomous driving technology that, as the media portrays it, will never make a mistake once it reaches Level 5?"

Musk nodded seriously: "That's right. In the media, it seems like technology can solve all problems. Once L5, the highest level of autonomous driving, is achieved, it will never make a mistake, it will always drive correctly, and there will absolutely be no safety accidents."

We know that humans make mistakes, and so do autonomous vehicles.

L5 simply means that it is less likely to make mistakes than a human and can adapt to driving needs in any scenario.

But it still makes mistakes.

The car continued to drive quietly on the streets of Shenhai, attracting numerous onlookers and photographers along the way.

Some passengers who knew about it even stuck their bodies out of the windows to take pictures, trying to see the people inside through the glass.

Car reviewers on Weibo are in an uproar.

This time it's really exciting.

"I had always known that Huawei had a vehicle project led by their 2012 Labs. I thought it was Zunjie, and I figured it would probably incorporate a lot of new technologies."

Now it seems she was probably a silly girl!
Holy crap, this is such a pleasant surprise! To be honest, I've always believed that the world's first L5 will come from a Chinese company. We have the best market, the most diverse road conditions, the most varied scenarios, and the most competitive companies.

But I still didn't expect it to happen so quickly. It's not even the end of 2024 yet, and we already have L5.

When will it be implemented? Which models will be equipped with this autonomous driving system? Will it start at a million units or will it be available on all models?
And most importantly, how will other automakers play the game?
My first concern was for XPeng. I was really worried about them. They have always been the most radical car company that advocates pure visual solutions. What will they do now?
Upon closer examination, one will realize that it makes no difference; this is an indiscriminate attack on all car companies.

Once Huawei implements HarmonyOS in all its HarmonyOS-enabled vehicles, whether you use pure vision or a combination of sensors, you will directly face the impact of this L5 autonomous driving system. This is a game-changer, and the bosses of all non-HarmonyOS-based car companies are probably having a sleepless night.

On the contrary, Lei Jun may be under less pressure, as they sell more emotional value and are not particularly known for autonomous driving.

However, I believe that Huawei will not go too far and will not take over the entire market. In fact, it may even become a true Bosch and sell this solution to all Chinese car companies.

This requires wisdom and coordination from Beijing. When considering China as a whole, I believe it is a sweet trouble.

Finally, I have to complain. "Silly Girl" is such an awful name. I know it's a tribute to Magic Phone, and at my age, I've probably seen Magic Phone when I was a kid. But when it's officially released, please, please, please change the name.

To offer some emotional support, I'll never laugh at Huawei's naming convention based on the Classic of Mountains and Seas again. Compared to Silly Girl's, the Classic of Mountains and Seas is way more sophisticated.

"I finally know what that test vehicle that was exposed on the road before was. At the time, everyone was joking that some company was crazy enough to design a model similar to the Li Auto MEGA, or even a smaller version of the Li Auto MEGA."

Now I know, it turns out to be Burning God. Deep Red truly deserves to be called the No.1 domestic artificial intelligence, and this has been proven once again with the first L5 autonomous driving.

Musk was probably feeling bitter. He thought this trip to China would be his last visit to Tesla before taking office as NASA administrator. He might be working at NASA from now on and wouldn't be able to come to China or see the Gigafactory there for a long time.

But to everyone's surprise, there were no surprises, only shocks. Whether it was the intense, deadly dialogue throughout, or the final Level 5 autonomous driving experience, Musk is probably having a sleepless night.

“My perspective may differ from others; I focus more on the technology itself.”

Is it L5? I don't know, but if Burning God says so, then I believe it must be. Crimson doesn't manufacture cars, nor does it need to. Burning God himself doesn't need to rely on this kind of way to attract attention.

Then let's talk about the points observed during the entire live broadcast. First of all, everyone noticed that when they started to move from a red light to a green light in some narrow lanes, there were many electric bikes and delivery motorcycles weaving in and out of their path.

Traditional autonomous driving systems are often overly conservative, causing them to get stuck in place and afraid to move.

But this car didn't hesitate; it made a smooth start decision, even exhibiting a behavior similar to an experienced driver pressing the accelerator to grab a window.

This demonstrates not just avoidance, but prediction and game theory. The model clearly made a high-confidence prediction of the movement intentions of surrounding vehicles. To put it bluntly, the model itself is very confident, which is consistent with Huawei's past aggressive and confident approach to autonomous driving.

Then, when crossing a complex intersection, I noticed that as I entered the intersection, there was a large truck on my left and pedestrians crossing the road illegally on my right.

The car didn't just wait passively; instead, it made a small, forward-probing motion, something traditional perception systems struggle to do. This shows that it not only relies on vision but also combines radar signals to probabilistically model targets that are obscured.

From my professional perspective, I believe this is the difference with L5: unlike L2/L3, it doesn't get stuck in the "stop when you can't see" situation; it dares to make decisions even with incomplete information.

Finally, there was the overtaking and lane changing. When they entered the inner ring road from Baoshan and got onto the elevated road, Ran Shen and his team encountered a slow-moving taxi. The system decisively changed lanes and overtook it.

This behavior could also be done in previous autonomous driving systems, but overtaking was a high-risk decision in older autonomous driving systems, which generally required strong rules and a large safety redundancy.

When I saw it, I specifically checked the speedometer. According to my speedometer reading, the "Silly Girl" completed the overtaking maneuver in just 1.5 seconds through trajectory prediction and real-time game modeling. This is yet another demonstration of the style of a human driver.

I think that's why we find it so smooth, because its improvements are comprehensive, truly like driving with an experienced driver, rather than the jerky feeling we experienced with autonomous driving in the past.

Finally, the only thing I'm curious about now is how Silly Girl behaves in extreme environments, something that wasn't demonstrated this time.

On Weibo, "Shaniu L5" has already climbed to the top of the trending topics.

Automotive bloggers and social media influencers have been sharing their opinions.

The comments section on Huawei's official website is mostly filled with ordinary netizens asking when the product will be released. Many users who are waiting to buy it have even said that they will not buy it until the product is released.

With L5 already available, no matter how much each company hypes up their intelligent driving systems, I still have to wait.

For car salespeople, it's like the sky has fallen. How are they supposed to sell these cars now? Especially for brands that emphasize intelligent driving as their main selling point.

Turning her gaze back to the car, Lin Ran smiled after listening: "Elon, you're already panicking."

Even humans make mistakes, so how could our L5 autonomous driving system possibly be the L5 that the media is promoting?
Of course it will make mistakes, of course it will err.

We need to further clarify some concepts, namely, what kind of autonomous driving can be called L5 autonomous driving.

Personally, I believe that if its error rate is lower than the human accident rate across all scenarios, then it can be considered Level 5.

If we set even higher standards, then in every type of scenario—highway, city road, extreme weather, etc.—its error rate must be lower than the human accident rate, and its mortality rate must be lower than the mortality rate caused by human drivers. Silly Girl is in this stage.

Even with the strictest requirements, I believe the pinnacle of Level 5 is the ability to drive better than any human driver in any scenario. This is very difficult to quantify, and we are still working towards that goal.

Musk said softly, "L5.1, L5.2, L5.3?"

Lin Ran: "That's my personal opinion."

Max's trip to China can be described as a disappointing one, though not entirely disappointing, but rather half disappointing.

The appearance of the silly girl made him realize that things had changed and that the emergence of self-driving cars would affect Tesla and even the global automotive industry.

Musk also wondered if this meant that Crimson had taken another crucial step on its path to General AGI.
This made Musk even more disappointed. Why didn't this Chinese entrepreneur open source like other Chinese companies? Alibaba and ByteDance open source their technologies, but Shenhong never mentions open source. How can today's young people forget such a great internet spirit as open source?

Musk thought this on his private plane back to Florida.

He wasn't going to Washington; Big T hadn't moved into the White House yet. He was going to Florida, to Big T's ranch there.

Musk spread out the notes, folded like a map, on the long table.

"China's conditions were more straightforward than I expected."

There are four lines of text on the paper:
America: It puts pressure on the 4V semiconductor industry to relocate advanced production capacity to America, and leads American companies to hand over high-end production capacity to Japanese and Korean companies, gradually hollowing out the economic lifeline of 4V.

China: Through tariffs and preferential industrial policies, it will precisely target 4V companies, deprive the capital market of support, transfer lunar semiconductor-related technologies, provide lunar north pole landing technology, and cooperate in research on lunar semiconductor processes.

Cooperation and coordination: On the surface, the two sides are in opposition to each other, but in reality, they each get what they need, avoiding directly revealing the traces of cooperation.

In exchange, America must acquiesce to Japan exporting semiconductor-related technologies and equipment to China.

Musk drew a heavy line on the last line. He looked up and said, “They’ve made it very clear that this isn’t about tariffs, it’s not about markets. What China wants is to unshackle them in the semiconductor technology field.”

They need to be able to directly purchase the key components of the semiconductor industry from Japan: materials, equipment, processes, and everything else related to semiconductors.

China wants a key to unlock the door.

“Interesting chips, interesting trades,” Big T said.

The prospective Secretary of State, sitting next to Big T, said, "I don't think we should make such a deal with China."

If Japan opens its semiconductor exports to China, it could close the gap with America within ten years.

Musk corrected, "Sorry, based on what I observed during my trip to China, it doesn't need to be ten years, five years is enough."

A ten-year forecast is too optimistic.

A middle-aged white man in an army uniform wrote three words on a whiteboard: 4v—America, 4v—Continent, and Japan Tech—Transit.

"The first two are transaction structures that we can accept, but the third is the problem."

He turned to look at Big T, "If we tacitly agree to exporting neon semiconductor technology to China, it's like putting a knife in their hands in global industrial competition."

An executive from IBM said, "Moreover, there is no gray area in this area."

This is the core issue of export restrictions.

If we give even the slightest tacit approval, and they seize upon it, the entire Congress, the media, and even our allies will be after us.

Big T tapped the table, his tone somewhat playful: "But the stakes they're offering are quite high."

By turning 4V into an empty shell, what we ultimately gain is high-end semiconductor production capacity, high-end jobs, and a large amount of cash.

John Morgan Sr. laughed, “Five hundred billion, Mr. President, at least we can squeeze five hundred billion dollars directly from 4V.”

In the long run, through certain operations, we can continuously reap the benefits of 4V, and the profits will not be less than one trillion.

This doesn't need to go to Congress, it doesn't need to go to the national treasury.

His tone was seductive.

John Morgan Jr. joined the Donald Party and collaborated closely with Lin Ran. When Big T showed signs of winning, the Morgan family brought John Morgan Sr. into Big T's core team.

John Morgan Sr., on the other hand, was a master of the art of eating, drinking, and having fun, and was also very good at flattery. He was an exceptional golfer, after all, he had been kicked out of the core family business since the 90s and was determined to live the life of a playboy, so these were all essential forms of entertainment for him.

John Morgan Sr. was in New York and was, in a sense, an old acquaintance of Big T, so he quickly integrated into this small circle.

Being a direct descendant of the Morgan and Adams families through marriage gives the name a significant advantage in America.

John Morgan Sr. undoubtedly felt it was time to do it; China breaking through the blockade was the future, and making money was the present.

The diplomatic advisor opened a folder: "That's why they've brought neon to the forefront."

What they want is not for us to sell technology directly, but for Japanese companies to export directly to China.

They offered a very suitable perspective: Japan's technology is not the most advanced, and it still lags behind semiconductor companies like ASML and America, so they felt that this condition was acceptable to us.

On the other hand, from Yanjing's perspective, as long as America turns a blind eye and stops forcing Tokyo to block the exit, they can bypass America's restrictions.

This is actually the most sensitive point.

There are two factions in Japan: one is pro-American, and the other is pro-China.

If Tokyo senses we're relenting, those industry factions will immediately push for equipment and material export licenses.

The incoming Secretary of State strongly objected: "That's the core lifeline. We can't be draining our own 4V while building bridges for them."

The industry consultant hesitated: "If 4V is completely hollowed out, it could make America the only country with a truly complete semiconductor supply chain in the short term."

However, if China acquires Japanese technology, their catch-up speed will increase exponentially.

Big T turned his gaze to Musk, "Elon, what do you think?"

Musk is currently enjoying great popularity and is the most prominent figure in the Big T team.

"I think it's entirely feasible to hollow out the 4V."

But people will think that we are loosening the reins on China.

What I want to say is that we should also obtain lunar landing technology from China.

The current competition is on two fronts: space and semiconductors.

We traded time in semiconductors for time in space, and the deal that China attached was the wealth that 4V had accumulated over the past decades.

This is the essence.

I believe we need to reach a cooperation agreement; it's uncertain how much of the previous president's trillion-dollar lunar fund remains.

My opinion remains the same: we shouldn't compete with China for the moon; we should go directly to Mars.

We'll acquire lunar landing technology from China and build a small base on the moon to maintain America's dominance in the space industry. In reality, we won't invest too much energy in that; our focus will be on going to Mars, building a Martian base, and terraforming Mars.

Musk's subtext is that cooperating with China gives the American people a sense of victory, ensures that we can still win the election four years later, and ensures that we can maintain our dominant position in the aerospace field through propaganda.

Then all resources were poured into Mars, and we won completely.

Winning in theory and winning in practice are two separate paths.

Big T said in a muffled accent, "They don't want us to sign documents publicly, they just want us not to stop the Japanese."

From their perspective, this is simply tacit agreement, and in the international system, tacit agreement is sometimes more effective than policy.

The long table fell silent.

Big T finally spoke up: "For A and B, we will implement the '4V core extraction, porcupine strategy'."

"Keep Insurance C as well."

He raised his head, his eyes carrying the stage presence of an election campaign: "As for China's demands, we will not pay them, we will not unlock it."

Then he changed his tone: "But we can pretend we're talking."

He pointed at Musk with his finger: "You continue to be their ears, give them hope, but don't give in."

We need to make them think there's a chance. That way, they'll continue to cooperate.

Then try to get them to hand over the moon landing technology. Elon, you can pick some less critical semiconductor technology from Japan and give them a little something to start with.

But we will never hand over the truly crucial key to them.

A typical oversized T-shirt style.

They want all the benefits, but they don't want to give away any of the bargaining chips.

(End of this chapter)