news

한어Русский языкEnglishFrançaisIndonesianSanskrit日本語DeutschPortuguêsΕλληνικάespañolItalianoSuomalainenLatina

*This article is from the 8th issue of "Banyuetan Internal Edition" in 2024

The self-driving taxi "LuoBoKuaiPao" was launched in Wuhan, Hubei, and quickly sparked heated discussions. The price ranged from 4 to 16 yuan per 10 kilometers, attracting many citizens to queue up to try it out, and also made more people realize that self-driving will come sooner than expected.

This groundbreaking experiment is not only a test of the technology's footprint, but also an impact of technology on the existing labor force structure and industrial structure. In addition to stakeholders, the public's attitude has reached a delicate balance: they are afraid that autonomous driving will not come, but they are also afraid that autonomous driving will be messed up.

New technologies that free hands, improve efficiency, and save attention will surely bring about social progress and an increase in total wealth. But what about people who make a living by driving? How should this account be calculated? This is not a single-choice question that means choosing "technology" means giving up "people".

How to choose a sock knitting machine or a sock knitter?

Today's "radish" shock wave is similar to the "Luddite Movement" more than 200 years ago.

One day in the 19th century, the first batch of fast and efficient hosiery machines entered British factories, and without any warning, a group of hosiery workers were on the verge of being "optimized". The hosiery workers struggling on the poverty line took their anger out on the machines because of their livelihoods, and entered the factory to smash the hosiery machines.

How to choose between hosiery machines and hosiery knitters? History has given the answer. Britain, which chose hosiery machines, was the first to complete the first industrial revolution and completed primitive accumulation far ahead of other countries. Today, hosiery machines, a new technology that was born more than 200 years ago, not only did not increase the unemployment rate, but instead gave inventors of various machines the opportunity to create boldly, and also allowed us to wear more beautiful and warmer socks.

In the past one or two hundred years, there have been countless choices like this, and history has made them one by one. We did not throw away instant noodles because of the chef's job, nor did we stop takeout because of the instant noodle manufacturer's job, nor did we stop driving cars because of the coachman's job, nor did we prohibit the installation of cameras in smartphones because of Nikon and Canon's jobs, so we have more and more colorful modern life.

The "Carrot Run" driverless vehicle is driving on the road in Pingshan District, Shenzhen. Photo by Liang Xu

At present, every step of artificial intelligence is on the tip of the public's mind. Seeing the screenshot of the "7 kilometers 5 yuan" order shared by the first-time experience passenger of "Carrot Run", everyone can't help but sigh: "Why bother to draw a number?" "Shared bicycles are just like this"... And for operating companies, the dividends of autonomous driving are even more needless to say: not only can they work around the clock, but they also don't need five social insurances and one housing fund, and can work overtime anytime and anywhere, which can be called the "king of cost-effectiveness".

Just as smashing the machine is the stress response of a sock knitter after seeing the sock knitter, people who regard autonomous driving as a scourge don’t need to do so.

"Carrots" lead humans to self-cultivation

Today, the fast-running "radish" also has his own journey.

The complexity of autonomous driving from trial operation to commercial closed loop tests the availability, reliability and controllability of the technology.

Wuhan citizens who experienced the driverless taxi called the car "苕苕". "苕" in Hubei dialect means "stupid": "When it is cut in, it will brake and stop suddenly even in the middle of a busy road. It's very stupid." said a Wuhan citizen after experiencing it.

This means that at this stage, human drivers still have an advantage in flexibility and mobility, and there are still many skills that Carrot needs to improve. To achieve a closed loop of autonomous driving, three clear questions must be answered: Where am I? Where am I going? How do I get there?

As a special type of "robot" and a mobile "car robot", autonomous driving requires precise perception, analysis, and execution. Sensors, like human eyes, are responsible for collecting information about the external environment during the car's driving. When the information is transmitted to the computing unit, the on-board chip, like the human brain, uses the decision-making algorithm to give the appropriate control strategy. Then, based on the control strategy of the analysis layer, the on-board chip controls the car's throttle, brake, and steering wheel to complete acceleration, deceleration, steering, and other actions to complete the movement.

This process is simple and clear. However, if we want to achieve "real-time" and "precision", we need a technology ecosystem that can affect the entire system and "hardware and software must work".

Currently, the evolution of autonomous driving is tied to the capabilities of artificial intelligence. Recognition, analysis, decision-making, and control, each link is inseparable from the precision of artificial intelligence algorithms.

Advanced intelligent driving requires advanced algorithms and relies on high-quality data. The problem is that traffic accidents often occur in black swan events with extremely low probability, but it is these extreme scenarios that have the most important data, yet are the most difficult to obtain. The high-performance end-to-end models currently being installed in vehicles by new energy and new forces rely even more on large amounts of high-quality data. In the foreseeable future, model capabilities will remain a key factor affecting the final outcome of autonomous driving.

So, is everything done by training the model? It's not that simple. On the diagonal line of vehicle-road collaboration, there is also a competitor called single-vehicle intelligence. Vehicle-road collaboration has many requirements for the environment and a high threshold, while single-vehicle intelligence only requires one vehicle. Once this route matures, it is more likely to reach a wider area.

If it takes a long time for single-vehicle intelligence to mature, autonomous driving under the vehicle-road collaboration route will have the opportunity to move from intelligence to sharing, which will subvert the travel market and reshape the pattern of automobile consumption.

When a higher proportion of car purchases come from travel platforms rather than individual households, the design, production, sales, and insurance of vehicles will be a completely different picture. Will the right to speak and the bargaining power in the industry chain belong to artificial intelligence companies, vehicle manufacturers, travel platforms, or new roles that have not yet been born?

Leaving the Cave and Heading for the Promised Land

Autonomous driving is not the first application of artificial intelligence to the topic of "sock knitting machines and sock knitters", nor is it the last. Unmanned supermarkets and unmanned warehouses have begun to be realized, and unmanned distribution, unmanned mines, drone airports, and unmanned farms will also be realized one by one in the future. Changes will only become bigger and faster. The unstoppable efficiency revolution makes the wind outlets become fleeting.

Similar to the range of new energy vehicles, humans will also enter the "long-range era". This "long" does not only refer to the length of life, but also represents a long-term perspective.

Before the insecurity strikes, you might as well ask your next generation: Are they really willing to continue driving online ride-hailing cars? Just like knitting socks with your hands in the age of machines, holding on to horse-drawn carriages in the age of cars, and swimming in the age of ships, repeating the lives of the previous generation? Will the next generation have their own new dreams?

Technology never stops to wait for anyone. Even without autonomous driving, the online car-hailing market is too crowded. The number of online car-hailing drivers is visibly increasing. The acceptance rate is lower and the average customer price is falling.

If, under such a trend, we still hold on to the obsession of "refusing to change" and cut off the possibility of technology "making the pie bigger", it would be like the primitive man who was reluctant to leave the cave. Although he avoided the wind and frost outside the cave, he also missed the lush water and grass and the wider world.

Banyuetan reporter: Zhang Manzi/ editor: Zhang Xi

Editor: Chu Xiaopeng / Proofreader: Zhang Ziqing