The United States restricts AI exports, China says: Not affected!

On June 6th, 65438+/KLOC-0, the U.S. government regulations restricting the export of artificial intelligence software to China came into effect. According to this regulation, all target recognition software applied to smart sensors, drones, autonomous driving, satellites and other automation equipment will be supervised. As the largest AI country in the world, the United States issued an injunction against China. On the surface, I'm afraid it will have a great impact on the development of smart cars in China.

However, is this really the case?

The regulations promulgated by the state are indeed mandatory, and it is not difficult for the state to ban a company's technology. However, the restrictions on AI export this time involve not a technology and a company, but the integration and integration of different technologies between companies in two countries.

Take an event in 12 century as an example. At that time, Venice was the center of glass manufacturing in the world, and its glass products were unique in the European market. 129 1 year, in order to ensure the leading position of its products, the government issued a ban to move all glass workshops outside the island and refused to be snooped by other outsiders. At the same time, various regulations have been formulated to prevent glass masters from going abroad or leaving the industry. However, looking back on the incident itself today is a joke!

Even at that time, Venice's strict ban didn't really control the glass technology completely at home, let alone China and the United States in the information age.

The development of a country needs constant innovation, and so does any technology. To some extent, blockade technology is like shutting the country out. Give another simple example. Circuit boards look beautiful and even full of artistry, but behind them, you may need stupid bronzing machines, bulky digital machines and various durable heavy machinery. This shows that in today's society, many industries have formed industrial chains, and the reaction caused by simply blocking a technology is chain-like, which may lead to the lack of an industrial chain system.

At least in the short term, the AI blockade has no real impact on China. At present, there are four directions related to AI informatization in China, namely AI vision, AI voice, GIS software and intelligent driving. Among these four, AI voice and intelligent driving are deeply related to cars, so I will discuss them in these two directions.

The first is AI voice. At present, the most representative enterprise in China is Iflytek. Since its listing in 2008, Iflytek has been deeply involved in artificial intelligence technology 19 years, among which AI voice is very representative. From 20 10, Iflytek began to set foot in the automobile circle. Up to now, more than 1000 vehicles and more than17 million front loaders have been delivered, covering more than 90% of domestic independent and joint venture brands.

More importantly, Iflytek has the international leading core technology of artificial intelligence, which is mainly reflected in the word "independent research and development", which is precisely the AI part banned by the United States. To put it simply, Iflytek independently developed AI related to the US blockade, and according to Iflytek, it has long been on the US restricted purchase list, so this blockade has not affected it.

To some extent, intelligent driving is a sub-field with the largest replacement space for industrialization in China. How many car brands in China are not equipped with L2-level autopilot assistance system? And if you look closely, you will find that the proportion of independent brands will be higher than that of joint venture brands. This is because there are no price and technical obstacles in the localization of L 1-2 intelligent driving system in China. It has become a real mass-produced product, as common as glass in this century, and not so difficult.

Secondly, even compared with overseas technology, China, as an AI power second only to the United States, has the same strength as foreign countries in L 1-2 autopilot assistance system, and even has advantages in development cost and supporting services because of its localization advantage. As for the more sophisticated L3 or even L4, the pace of algorithm research by domestic and foreign competitors is basically the same.

In the short term, China will never worry about any related AI blockade, but in the long run, point-to-point blockade is the consistent practice of the United States to prevent the spread of domestic technology. From the chip, to the control of Watson system, to the software, the blockade of the United States has never stopped, and the area is getting bigger and bigger. Therefore, the fundamental way to deal with the blockade is to vigorously develop independent research and development of AI technology.

However, the discipline development of AI technology is not easy. First of all, in terms of cost, AI needs a lot of high-precision hardware, which is the cost of money; Carrying out a course often requires 7-8 researchers to participate, which is the personnel cost. There is no doubt that domestic development in this area is not as mature as that of the United States. Therefore, in recent years, China has actually benefited from the "academic evaluation system" of the United States.

For example, we often say that returnees are gilded. Since we don't know whether it is a golden turtle or a rotten copper turtle, how can we judge whether it is genuine or not? At this time, we can check their published papers. For example, a returnee who has published a paper in a magazine such as Nature or Science is undoubtedly a top talent. This is a transparent, fair and open academic system, and it is also the most lacking academic system in China at present. Thanks to this system, we can better distinguish between gold and scrap iron. At present, the United States is quietly restricting China's academic research in sensitive scientific and technological fields. In the long run, it will definitely not be conducive to domestic academic development, resulting in a serious brain drain and a scientific research fault. Therefore, we won't be affected in the short term, but in the long run, there is only one way to make ourselves stronger if we don't want to be influenced by others.

(Image source network, invaded and deleted)

This article comes from car home, the author of the car manufacturer, and does not represent car home's position.