The combination of artificial intelligence and wearable devices can be divided into two main categories, one is with the help of the network and the cloud, wearable devices to monitor a variety of human body and environmental data, uploaded to the cloud through the network, analyzed by the AI chip in the cloud, and sent back the corresponding data and instructions. The other category is the artificial intelligence chip integrated in the wearable device processor, so that you can not rely on the network and the cloud, directly in the wearable device to complete the collection and analysis of information to produce results. The research results of the University of Cambridge and the "Huangshan 1" chip correspond to these two types of artificial intelligence and wearable devices. On the one hand, AI wearable devices with the help of the network and the cloud have high requirements for the network and the cloud AI processor, which needs to transmit the collected data through the network with high speed and low latency, and quickly complete the computation and analysis in the cloud to draw conclusions and send them back to the user. This will enable effective monitoring of the body and timely warning of health threats. The advantages of high bandwidth and low latency of the 5G network that will be put into use in the future can strongly support the realization of this function. The continuous optimization and upgrading of artificial intelligence algorithms and the continuous improvement of processor performance also provide a guarantee for cloud-based artificial intelligence to analyze and process a large amount of uploaded monitoring data. On the other hand, by integrating the AI chip into the wearable device processor, the process of data transmission between the user side and the cloud side can be omitted, breaking away from the constraints of wireless transmission, reducing the dependence on other devices and conditions when the wearable device is in use, enhancing the degree of independent work of the wearable device, and lowering the burden of the user's use.
Unlike current common forms of AI, hybrid intelligence places more emphasis on human-machine collaboration and the role of people. That is, the role of the human is introduced into the intelligent system, forming a hybrid intelligence paradigm of human in the loop. In this paradigm, people are always part of this type of intelligent system, and when the output of the computer in the system is low confidence, people actively intervene to adjust the parameters to give a reasonable and correct solution to the problem, constituting a feedback loop to improve the level of intelligence.