news

Author: Du Yuxin

Editor: Yang Jing

Editor: He Zengrong

If we were to ask which industry and field was the hottest in the past month, it would be "Carrot Run" and the Robotaxi behind it. For a while, discussions about this topic were endless, and emotions such as driverless online ride-hailing, revolution, and job-stealing were spreading among the crowd and society.

Regarding the transformation of an industry, there is another very active dimension, that is, the capital market. In the past month, related concept stocks have seen a crazy rise. In addition to promoting the theme behind RobotaxiBaiduIn addition to the more than 20% increase in stock prices in the first half of the month,Public Transportation、King Long MotorsStocks related to this concept, etc., achieved daily limit increases for many consecutive days, supporting the rare glory of the sluggish A-share market.

Volkswagen Transportation rose 300%, King Long Automobile rose 167%... According to this stock price trend, ordinary consumers don't need driverless online ride-hailing. If you grasp it properly, you can make a lot of money in minutes. It's exciting! However, the story of capital is often full of bubbles, the enthusiasm when it surges in and the sadness when it ruthlessly withdraws. More sadness is just letting retail investors take over.

Facts have proved that, except for a few individual stocks that have been boosted, the entire driverless online car-hailing concept stock has cooled down significantly since the second half of July. Many people suddenly realized that this might be just another tool for capital speculation.

In fact, this scene is no longer surprising to audiences who are well versed in the capital market and A-shares. From the original metaverse to blockchain, from the new energy boom to smart cars, including the recent low-altitude economy and driverless car projects, whichever field capital wants to be popular, that field will respond.

On the vast stage of the A-share market, hot money is like a group of mysterious dancers, dancing to the beat of the market, but it is the retail investors who are ultimately trapped or bear the consequences. In the words of netizens on the stock forum, those who have been rising for consecutive days cannot enter, and those who have been falling for consecutive days cannot exit. It seems that the majority of retail investors let the institutions cash out and leave, and the former is the core force supporting the development of the industry.

In fact, the bubble about driverless cars is far more than that in A-shares.

From the perspective of the industry, the trend of smart taxis or driverless online ride-hailing cars seems to be the inevitable result of the current technology and automobile development entering a new stage. In addition to the attention brought by Baidu Carrot Run last month, the topic of smart driving has also been fermenting in recent days.

For example, the Enjoy S9, which is backed by Huawei, recently achieved automatic operation from parking space to parking space for more than 30 kilometers in a real-world experience test, pushing domestic high-end driving assistance to a new level. As more car manufacturers adopt Huawei's intelligent driving system, the intelligent driving track driven by China's most powerful technology company will usher in new hot spots and explosions.

For example, BYD and Uber announced a multi-year strategic partnership. In addition to launching 100,000 new BYD electric vehicles in the global market, the two parties will also cooperate in the field of autonomous driving. The outside world believes that the model launched by the cooperation may be Robotaxi. As the largest car company in China, the scale effect is naturally the most powerful and is most conducive to reducing hardware costs.

Then there is Tesla's FSD, which is known as the vanguard of high-priced intelligent driving. It has announced that it will enter China at the end of this year. Tesla has also announced that it will enter the Robotaxi field. The outside world believes that Tesla may become a catfish by then, just like when it entered the new energy field, stirring up the development pattern of the entire industry and market.

In fact, Robotaxi is not a new thing. Since Google launched the research of the driverless project in 2009, the world's interest in driverless cars has heated up. Since then, many start-ups have emerged from the United States to China, and the capital market has become more enthusiastic about investment and financing in autonomous driving.

However, as funds tightened and the commercialization of driverless cars has been slow to materialize, many startups and capital investments have been unable to make ends meet. This includes Argo AI, which was invested by Ford and Volkswagen, which went bankrupt and liquidated, and Ibeo, the pioneer in the field of lidar, which filed for bankruptcy. More driverless startups are beginning to go bankrupt.

It can be said that autonomous driving has gone through a period of expansion, then hit a wall, and finally entered a period of rational return. This cycle has once again pushed the autonomous driving track to the forefront.

Why is driverless driving suddenly back on track? Perhaps the most fundamental reason is the advancement of technology, the computing power and algorithm improvements brought by high-end AI chips, and especially the milestone of “end-to-end” transformation, which has enabled driverless driving to evolve from original rule-based control to deep intelligent learning and decision-making.

What is end-to-end? In fact, it is a concept in deep learning. The logic behind it is more inclined to the learning thinking and decision-making mechanism of the human brain. An AI model only needs to input raw data to output the final result, and no additional modules need to be introduced for processing in the middle. With the support of powerful computing power and the feeding of more data, the iteration of the intelligent driving model will become faster, and even reach the driving level of humans.

Chips, software and algorithms can promote the development and progress of this field. It seems that this is a very promising field. According to statistics, the scale of my country's driverless market is expected to reach 383.2 billion yuan in 2024. Such a large-scale hot commodity will naturally attract capital.

In addition, external actions that follow the advancement of driverless technology are also driving the development boom in this field. Recently, the "Guangzhou Intelligent Connected Vehicle Innovation and Development Regulations (Draft for Comments)" was released, which intends to support the application of intelligent connected vehicles in transportation hubs such as highways, airports, ports, and travel service scenarios such as urban buses and taxis.

The core content of the draft regulations emphasizes the application and promotion of intelligent networked vehicles in key transportation hubs and urban travel services, as well as the gradual opening of roads throughout the city where autonomous driving and human driving can be mixed. This will not only help improve traffic efficiency and service quality, but also lay the foundation for the commercialization and popularization of intelligent networked vehicles.

Many information and facts point to the fact that driverless cars are becoming the core area of ​​current technological advancement. The upsurge of capital investment is just the longer-term logic behind it. The real value it brings to consumers and society remains to be observed and discussed. Including the hotly debated operating price of driverless cars, the price increase that has already occurred, and even the possibility of a routine in the future like shared bicycles and power banks, where capital is first used to attack the market and then harvested. No one knows what the situation will be like for driverless cars.

Regarding driverless cars, many people previously had an opinion: buy an driverless car, use it yourself normally, and let the car go out to work and pick up passengers automatically when not in use. Not only does it not cost much, it also makes money for the owner... This opinion made many uninformed listeners feel incredible. It turns out that driverless cars can be played like this, which is really eye-opening.

In fact, this vision may be realized under a single social system, but I personally believe that this situation is absolutely impossible to achieve under a complex and diversified social system and scenario.

Based on NIO's Li Bin's investment and understanding of the shared bicycle and ride-sharing businesses, he believes that driverless taxis cannot replace traffic. A city can only accommodate a certain number of operating vehicles, whether they are driven by people or not. This quantity means that driverless taxis will never become a borderless and scalable business like software cloud services.

In other words, the real potential of this track is limited. The capacity of urban roads is limited. There are diversified means of transportation such as airplanes and high-speed rail, and people's travel demand for automobile products is also limited. The field and track of automobiles cannot have absolute and unlimited expansion like software, mobile phones, and shared bicycles. Behind this is the relationship between the upstream and downstream industrial chains, which has already determined the overall capacity of the automobile industry. Even if there are changes, they are only changes in a very small part of the field.

Take the current trend of driverless cars as an example. Whether it is the high-end intelligent driving represented by passenger car companies, or the commercial operation of driverless cars represented by Luobo Kuaipao, both solutions are in the early stages, or even in the primary exploration stage with no future in sight.

Although the ADS 3.0 system represented by Huawei can achieve automatic operation from parking space to parking space, and the driverless car represented by Baidu can achieve fully automatic unmanned operation, the handling of special circumstances, regulations, ownership of rights and responsibilities, the degree of acceptance of such things by different people, and even possible risks in the future are still a difficult problem that lies on the road to development.

For example, in April this year, Tesla FSD once again "indirectly caused" a fatal traffic accident. The traffic accidents and social topics caused by Baidu's driverless cars have made the so-called driverless and high-level intelligent driving full of contradictions in future development.

It is said that only by seeing through the true essence of intelligent driving can we truly understand the direction in which intelligent driving will move forward. According to Li Bin, the real social value of intelligent driving is to liberate energy and reduce accidents. "It is not to eliminate the jobs of today's hard-working private car drivers and taxi drivers. The real value is to reduce accidents, liberate some of the energy of driving, make driving less tiring, and be safer." Behind it all depends on how much the manufacturer is willing to pay to improve the user experience.

Of course, the outside world's enthusiasm for high-end intelligent driving can also be seen as an attack by many new forces in the new round of competition for automobile selling points. After all, from Tesla to Huawei, from new forces such as Wei, Xiaoli and Li Auto to many intelligent driving suppliers, as automobile homogeneity becomes more serious and more new selling points are needed to stimulate consumption, the intelligent driving competition has ushered in a new round of efforts.

On the evening of July 30, Xiaopeng Motors held a press conference on AI intelligent driving technology. He Xiaopeng said that Xiaopeng Motors has launched the first mass-produced end-to-end intelligent driving model in China, becoming one of the only two car companies in the world to achieve mass production of end-to-end large models. Not long ago, after test-driving Tesla's latest FSD, He Xiaopeng judged that 2025 would be the Chat GPT moment for fully autonomous driving.

At the NIO IN 2024 Innovation Technology Day on July 27, NIO released its intelligent driving technology architecture NADArch2.0 and its first self-developed intelligent driving chip. NIO said that NADArch2.0 has been upgraded to an end-to-end architecture that introduces a world model at the algorithm level, generating driving decisions directly from raw sensor data, reducing information loss in traditional solutions.

Coincidentally, at the 2024 Smart Driving Summer Conference on July 5, Ideal Auto released a new autonomous driving technology architecture based on end-to-end models, VLM visual language models and world models, and launched an early bird plan for the new architecture.

Traditional car manufacturers are also unwilling to lag behind. BYD said that their entire intelligent driving team has 4,000 people, of which 3,000 are engaged in software development. The basic salary of the autonomous driving team can reach 1 billion yuan a month. Mercedes-Benz announced that it recently became the first international car company to be approved to conduct L4 urban and highway autonomous driving tests in Beijing, and will carry out tests on designated roads in Beijing.

More news also points to the fact that car manufacturers are trying their best to seize the initiative in this field in the involutionary automobile race, hoping to add some selling points and popularity to their own brands. However, as for autonomous driving and advanced intelligent driving, it is at most a dessert, not the whole of future cars, and no one even knows what proportion it will play in people's car purchase and use decisions.