Behind the expansion of products and markets, it will not blindly burn money. Smart speakers are the most natural artificial intelligence voice interaction devices at this stage. As a future IoT interaction, voice is obviously a more important way, so voice artificial intelligence has become an important battlefield for each future IoT layout. Baidu DuerOS upgraded to 3.0 in July 2018, DuerOS is a "conversational artificial intelligence operating system", tailored for future IoT, is a system platform for voice, image [including video] AI interaction, can be applied to speakers, TV, cars, and other IoT applications. In August 2018, DuerOS 3.0 activation reached 100 million, speech recognition accuracy reached 97%, and monthly active devices exceeded 35 million.
Alibaba's Ali Cloud has a wide range of coverage. The Tmall Genie is positioned as an AI product brand under the Alibaba Artificial Intelligence Lab. The first release of the Tmall Genie X1 smart speaker is positioned as an AI smart voice terminal, with built-in Ali AliGenie operating system, able to subordinate to Alibaba Cloud, focusing on voice artificial intelligence applications. At CES2019, Ali Genie doesn't just stay on the smart speakers, you can see the influence of traditional speaker companies, they began to provide intelligent voice services for BMW cars in the Chinese market. In addition, Volvo is also expected to adopt the Tmall Genie system in the Chinese market.
Xiaomi IPO on the Hong Kong Stock Exchange in the third quarter of 2018. The third-quarter earnings report showed that the Xiaomi IoT business market coverage was excellent, which should be in line with everyone's expectations for the Xiaomi IoT ecosystem. According to the specific data, the Xiaomi IoT activation device that does not include smartphones and computers reached 132 million units, and the monthly activity of Xiao AI exceeded 34 million.
Although the data is not particularly comprehensive. From the market coverage of several major brand smart speakers, each smart voice has strong technical support and tens of millions or even 100 million levels of activation. And the monthly active equipment is also more than 30 million, so the scale can no longer be considered as the user incubation stage of smart speakers. But it can be considered as the incubation stage of AI intelligent voice input application. The speaker is the most acceptable device for users to experience voice input in the market, and the future market of AI smart voice is only part of the speaker.