With the right scenario, even driverless cars can be commercialised today.
Through the time of a child, artificial intelligence has experienced “three ups and two downs”. Now, artificial intelligence is in its third wave. Among them, we have seen the gradual application of deep learning algorithms, speech recognition, computer vision, autonomous driving and other technologies in the industry, as well as the government’s strong support for the ARTIFICIAL intelligence industry. For example, not long ago, The State Council of China issued the first “national” document for ARTIFICIAL intelligence — The Development Plan for a New Generation of Artificial Intelligence, which put forward directional opinions for the next development plan of artificial intelligence industry, including the “three-step” strategy, the expectation of market size and so on.
The wave of artificial intelligence has arrived, and Lee has identified four ways players can play
“From the perspective of innovation Factory, AI will go through four waves. The first wave is Internet AI, which is blessed with large amounts of annotated in-scene data, such as Google and Facebook in the US and BAT in China. The second wave is based on business scenario channel data, which is used by companies to generate value based on business processes. The third wave is to do the AI+ work of traditional enterprises, to capture and apply the data of the real physical world that was not online in the past to create new applications, such as smart stores and so on. The fourth wave is the WHOLE AI movement, like robots, unmanned vehicles and so on.” Kai-fu Lee, founder of Innovation Works, said at the China Artificial Intelligence Summit 2017 (CAIS 2017) held at the Nanjing International Expo and Convention Center yesterday.
In this passage, Lee describes four waves in the development of ARTIFICIAL intelligence. But if we look at it from another perspective, isn’t it four different ways of playing in the realm of artificial intelligence? At the same time, not only does Lee tell us how ai players play, but from an investor’s point of view, he also evaluates these four approaches, or four different fields.
“The first wave has already happened, and there are already many 2C applications like Facebook, wechat and Toutiao,” he said. “There is no room for players to start their own businesses. The great opportunity of the second wave of AI+ lies in the combination with traditional industries, such as the investment of the innovative workplace, the pursuit of one technology and so on. The third wave is the opportunity for traditional firms; The fourth wave is quite difficult, such as autonomous driving, and they need to continue to make technological breakthroughs, but some scenarios are feasible.
The industrialization of artificial intelligence technology, the first thing to do is to find an application scenario
Deep learning algorithms, speech recognition, semantic analysis, computer vision… These are all familiar technologies in the field of artificial intelligence. Based on these technologies, we have also seen artificial intelligence products and applications such as autonomous driving, robots, and face-scanning payments. Among them, computer vision, speech recognition and other subdivision industry has a considerable scale, can be said to achieve industrialization.
In contrast, automatic driving technology and so on in the industrialization of many weak, we can not help but wonder, in the industrialization application of technology, how should players do? Compared with speech recognition and other artificial intelligence technologies, autonomous driving may need to continue to work hard on recognition accuracy, but is the current industrialization really not feasible?
In fact, the answer is no. As for how to achieve industrialization, before the rigid requirements of technology and performance, it is very important to find a suitable application scenario.
“AI is all about the scene. What China’s AI industry lacks is not technology or talent, but a good scene to land on. The definition and segmentation of the scene drives the technology, and conversely the technology drives the scene.” When discussing the dual driving forces of ARTIFICIAL intelligence, Megvii founder and CEO Inchi said.
Take face recognition as an example, in the B-end market, security, authentication and other fields are the scenes where it realizes value and provides services. Such as unmanned again, although still need technical breakthrough, in also has the shackles of the law, but this does not prevent driverless cars in closed or regular sex scenes of lower speed in the fall to the ground and industrialization, as controlling the potential of science and technology, founder and CEO Wu Gan sand park park, in the warehouse forklift truck and other special type.
Wu Qiang, vice president of Horizon Robotics and chief cloud architect, and Yu Zhichen, founder and CEO of Turing Robotics also have their own views on this. Wu Qiang said that the ground line robot is committed to embedded processor, but this is just the basic technology, only around the application scenario to customize the solution can accurately meet the needs, to form a large-scale landing, and then achieve industrialization. As for Yu Zhichen, based on his experience in the industry and application scenarios, he concluded that only by digging deeply into the scenarios can the value and services of ARTIFICIAL intelligence technology in actual scenarios be realized.
We can see that nowadays speech recognition, computer vision and other technologies have been quite mature in terms of performance, but from the perspective of industry, the application scenarios of these technologies are only limited to a few areas. So, to a certain extent, those fields may have been developed very mature, but the industrialization of these technologies needs to continue to open up more application scenarios.