To AI, combining IoT is a necessary step

- Apr 25, 2019-


Seriously, the history of AI can be traced back to IoT for decades. However, compared with the latter, in view of the difficulty of technical research and development and the difficulty of landing, AI's journey is very difficult, and it has been one after another. Only in recent years has achieved breakthrough progress in the technical level, and the public awareness continues. While improving, I found the possibility of technology landing.


At present, because of technological breakthroughs, the daily AI technology products that we can access include fingerprint recognition of mobile phones, face recognition security inspection of transportation hubs, shopping malls and shopping robots. Although some AI technologies or products are sometimes not satisfactory in terms of performance, they can even be called “mental barriers”, but it is undeniable that AI technology is now able to see people compared to several peak periods in the past. To the future.



However, as a new technology, AI will inevitably encounter the problem of “difficult to land”.


According to data from iResearch, the global AI market will be approximately 119 billion yuan in 2020. In the next 10 years, artificial intelligence will be a market of 200 billion US dollars. From the data point of view, AI's market prospects are recognized, but can AI alone really afford such a big market? Obviously it is not feasible.

For AI, as Alibaba's chief intellectual networking scientist Ke Zhen said, the lack of IoT AI will become a chicken rib, and the lack of interconnected AI will be more similar.



To give a simple example, take the AI product smart speaker that is currently discussed more, for example, it is a typical application of AIOT. As far as smart speakers are concerned, its "intelligence" is mainly embodied in two aspects: instruction reception and execution - in order to receive instructions well, smart speakers need to be able to capture and understand consumer voice commands, including speech recognition, natural language. Processing AI technology; in the process of implementation, it is necessary to assume the responsibility of a housekeeper, "command" other equipment to respond accordingly according to the instructions, this time I need to play.



Without the integration of IoT, the smart speaker can only exist as a completely independent individual. At this time, IoT's smart speakers can't be combined, such as checking the weather, putting songs, buying things, etc. Can only rely on content on the Internet. In addition, in the pursuit of personalized services, the services that can only be output by smart speakers that can only provide pure content services are “not comprehensive”. Under such limited circumstances, the value of AI technology obviously cannot be maximized.



It can be said that without the addition of IoT, AI technology or hardware sometimes cannot be well integrated into the landing scene, thus facing the problem of landing difficulties.



For AI, the combination of IoT and IoT is a necessary measure to maximize value and seek better landing. As the industry has said, AI has entered the second half of development, and the combination with IoT will help AI open an important channel for real landing.