With the release of the “New Generation Artificial Intelligence Development Plan”, the State Council began to deploy our country’s artificial intelligence development plan as a whole, and conducted the overall thinking, strategic objectives and major tasks and safeguard measures for the development of new artificial intelligence in China before 2030. System planning and deployment.

Judging from the national strategy, in the coming period, the combination of artificial intelligence and manufacturing will become the main means to promote smart manufacturing. From the perspective of industrial development, the artificial intelligence industry structure has taken shape, mainly including the basic support layer, technology-driven layer and scene application layer. The author will mainly sort out and analyze the artificial intelligence industry chain from these three levels.

Figure artificial intelligence industry chain three-tier structure

Basic support layer: long-term involvement of international IT giants, it is difficult for Chinese start-up companies to enter

In the field of artificial intelligence, the traditional chip computing architecture has been unable to support the needs of large-scale parallel computing such as deep learning, which requires new underlying hardware to better reserve data and speed up the calculation process. The base layer mainly uses hardware as the core, including hardware, neural network chips, sensors, and middleware for performance acceleration such as GPU/FPGA, which are prerequisites for supporting artificial intelligence applications. These hardwares provide computing power for the entire artificial intelligence operation. At present, most of them are mainly international IT giants.

Currently in the GPU field, Nvidia is leading the industrial-grade ultra-large-scale deep-network acceleration, and recently introduced the Tesla based on Volta, the first processor with a speed exceeding 100TFlops. Intel is mainly focusing on the FPGA-building industry and has introduced artificial intelligence chips that mimic the human brain. Google also introduced a second-generation TPU chip to provide chip support for its open source TensorFlow framework. In addition to these industry giants mentioned, there are many start-up companies in this area, such as China Star Micro, Cambrian and Xijing Science and Technology, but they cannot match these giants in terms of industrial layout ability and R&D strength.

Technology-driven layer: algorithm and computing power are the main driving forces, and open source is the trend

The technology layer is the core of the development of artificial intelligence, and it plays a decisive role in the application level of product intelligence. In this development process, the algorithm and computing power play a major role in promoting the development of AI. The technical layer mainly relies on the computing platform and data resources of the basic layer to carry out mass recognition training and machine learning modeling, as well as the development of application technologies for different fields, including two stages of perceived intelligence and cognitive intelligence. Among them, the sensory intelligence stage realizes the connection between people and information through sensors, search engines, and human-machine interactions, and acquires the data needed for modeling, such as speech recognition, image recognition, natural speech processing, and biometric recognition; The acquired data is modeled and calculated, and the results are obtained using the thinking functions of deep human learning and other human brains, such as machine learning, predictive APIs, and artificial intelligence platforms. On this basis, artificial intelligence can grasp the basic information input and processing capabilities of “seeing” and “hearing”, and can evolve more applied products to the user level.

At present, the domestic artificial intelligence technology platform is mainly focused on computer vision, speech recognition, and language technology processing at the application level, and the development momentum of domestic technology layer companies is also swift and violent. Representative companies include KTX, Gelingshen. Hitomi, Jietong Huasheng (Lingyun), Horizon, SenseTime, Yonghong Technology, Trump Technologies, Yunzhisheng, etc.

Scene application layer: AI and scene depth integration, a wider range of applications

The application layer is mainly based on the integration of the foundation layer and the technology layer to achieve the integration with the traditional industries, and realizes the application of different scenarios. With technological breakthroughs in artificial intelligence in speech, semantics, computer vision and other fields, it will accelerate the application to various industrial scenes.

According to different objects, application layer can be divided into two parts: consumer-level terminal applications and industrial scene applications.

Consumer-level terminals include three directions: smart robots, smart drones, and intelligent hardware. Scene applications are mainly connected to AI application scenarios of various external industries. In recent years, domestic companies have successively launched application-level products and services, such as smart customer service such as small robots and wisdom teeth services, virtual assistants such as “go out” and “degree secret”, and industrial robots and service robots are emerging one after another. Products and services are gradually landing.

Among them, IBM's earliest deployment of artificial intelligence, "universal Watson" to promote multi-industry changes; Baidu launched the "Baidu brain" program, focusing on the layout of driverless cars; and Google's artificial intelligence business is more complicated, many areas bloom, including AlphaGo, no People drive cars, smart surgical robots, etc.; Microsoft maintains leadership in language semantic recognition, computer vision and other fields. In addition, the home appliance industry has also set off an upsurge of artificial intelligence. Many home appliance companies have targeted artificial intelligence and have devoted themselves to the development of AI technology and applied it to home appliances. Since the beginning of this year, Changhong, Midea, Gree, Galanz, etc. are all making a transition to smart manufacturing, trying to establish a Smart Home and integrating artificial intelligence and smart homes more closely together.

Conclusion: Weak artificial intelligence has accelerated penetration, strong artificial intelligence is not far away

The author believes that although the field of artificial intelligence has ushered in a new period of industrial outbreak, but overall, artificial intelligence is still based on weak artificial intelligence in specific application areas. At present, the vast majority of systems, including deep learning neural networks and machine intelligence, have not yet reached the level of thinking. To continue to climb into strong artificial intelligence, the computer's cognitive ability is a must, and this is where artificial intelligence is currently trying to break through. AI originated from technology, and finally, the scene, the final need for industrial intelligence through deep integration with the industrial scene in order to achieve true landing.

Ring Type Connecting Terminals

Ring Type Connecting Terminals,Terminals,Connecting Terminals

Taixing Longyi Terminals Co.,Ltd. , https://www.longyicopperlugs.com