Beach smart speakers (on): "New World" or "mirage"?

This article is produced by NetEase Smart Studio (public number smartman 163). Focus on AI and read the next big era! Text / Xiaoyan Amazon Echo's hot directly detonated the domestic smart speaker market, the entire intelligent hardware industry chain has shaken. However, in the process of leaping forward, Chinese consumers do not seem to buy it. So, can voice interaction become the next generation of interactive technology? Smart speaker capable of voice interaction entry-level products? What is the next general public hardware after the mobile phone? What are the current smart speaker services? How big is the market in the future? First, the industry is booming, with ice and fire For the first two years, for the Internet and hardware manufacturers, you would lose a piece of cake without smart hardware; and this year's statement is that if you do not make smart speakers, you will be behind an era. This is because all people are sure that voice interaction will become the next generation of human-computer interaction. The smart speaker is the first entry-level product for voice interaction. To the east of the Pacific, Amazon’s Echo shipments have exceeded 10 million units, accounting for 70% of the U.S. market. The US smart speaker market has risen. The success of Amazon Echo has led Google, Microsoft, Facebook and other foreign giants to follow suit. On the other side of the Pacific Ocean, in China, tech giants such as Baidu, Ali, Jingdong, Lenovo and Xiaomi also began to make arrangements. Especially after Xiaomi released the smart speaker of 299 yuan, the entire industry chain has been shaken. According to a supply chain source, due to the layout of major Internet companies, many domestic OEM factories have received tasks across the board. Amazon Echo Series Speakers "The Amazon-powered smart speaker market can be replicated entirely in China, and the Chinese market will surely be able to reach tens of millions or even billions of shipments." Wei Qiang, who has been pursuing the smart speaker market for two years, said to NetEase Intelligence. Wei Qiang is the general manager of Linglong Technology. This company, which was jointly invested by Jingdong and HKUST, launched its first smart speaker product, the 叮咚Sound Box, as early as August 2015. “From the market capacity, the Chinese market is certainly larger than the US market. From the perspective of smart speakers, the US’s popularization time is one to two years earlier than China.” Song Shaopeng, who fought for five or six years in the speaker market, also stated that The same point of view. Song Shaopeng is the CEO of start-up company Sugr. When the WiFi speaker emerged in 2014, his team launched the well-known sugar cube speaker. When the wave of WiFi speakers passed, Song Shaopeng keenly observed the success of Amazon Echo, and decided to transform the solution integrator, currently providing authentication solutions for the Amazon Alexa platform. "Without the enclosure of the speaker, we formed a complete solution in the form of IC+software." According to Song Shaopeng, Sugr is the only solution for Alexa certification. This solution includes microphone processing (such as noise reduction, Echo and other technologies), Alexa complete system, speaker's industrial design, structural design, acoustic design, user experience and APP design. At present, there are many smart integrator solution providers in China, many of which are for the Alexa platform. The domestic market is booming, so that they have begun to shift their goals to the domestic. Baidu betting on AI also smelled business opportunities. Jing Ye, general manager of Baidu’s Secret Business Unit, revealed that Baidu’s program integrators for Shenzhen’s smart speakers, especially those provided by Amazon Alexa A box shell, as long as the development kit into it, you can easily make a smart speaker. Jing Hao hopes that these brands and solution providers will use the DuerOS to bring the hardware produced by Alexa directly to China. Almost all people think that the demand for smart speakers in China will be blowout. However, the fact is that the sales volume of the smart speaker market in China is sluggish, and consumers do not buy it. According to a rough estimate by agencies, the monthly sales volume of smart speakers in China is currently less than 20,000 units, with annual sales of up to 200,000 units. Deeply ploughing the smart speaker market for two years, the squeaky speakers have annual sales of only about 100,000 units. The industry boom and user icy have formed a strong contrast. It seems normal to Xie Diaoxia. “The smart speaker market in China will not start at a high speed and the maturity will not be so high. This is because smart speakers play in Chinese families. The role is far from important in the United States." Xiaomi launched 299 yuan smart speaker As the CEO of Haizhi Intelligence, a startup company specializing in natural language understanding, Xie Dianxia has lived in the United States for a long time. He analyzed the usage scenarios of Echo as follows: "Echo's users are mostly housewives and they spend a lot of time at home. Echo's usage scene is mostly in the kitchen. American housewives usually listen to background music while cooking. This is because the American kitchen is mostly Open kitchens, and their cooking is quiet compared to Chinese frying and frying and is suitable for listening to music." Xie Dianxia believes that, by contrast, China does not form a social group for housewives. At night, because socializing is more often not at home, Chinese families generally do not have open kitchens. Most of them are watching TV. These few reasons led to fewer speakers in China, and the usage scenarios and time periods were all compressed. "Even if Chinese manufacturers can make smart speakers more beautiful and cheaper than Echo, and content services are better, but as a single product speaker, its success rate and speed are not as good as those of the United States." Xie Dianxia told Netease Intelligence. “Because of statistical issues, the current sales volume of smart speakers in China may be less than this figure (monthly sales of 20,000 units), but there is no denying that Chinese companies will use the money to pull this amount out, but how many consumers will go? Buy, this question I am skeptical." Liu Rui said in a summary. As Netease's director of artificial intelligence products, Liu Rui has a profound understanding of smart speaker technology and user experience. He believes that the idea of ​​copying Echo in China does not work. China's smart speakers also need to do more polishing in product form and service. The old tree blossoming technology has provided a smart speaker inheritance plan for many domestic technology companies. Its CEO Zhu Junwen has many resource integration capabilities in the field of smart hardware. He views the initial market of smart speakers in China as follows: “This wave of intelligent speaker market will be divided into two big blocks, one can be called the high-end market, the price is about four or five hundred yuan, this piece is all major Internet companies are the main; the other is one or two hundred dollars or even more Low-priced smart speakers replace the traditional bluetooth speaker market." “But smart speakers certainly cannot reach the sales and status of mobile phones, and it is already remarkable that they can reach tens of millions of dollars,” Zhu Junwen added. Xie Dianxia believes that like millet AI speakers will hit the price of 299 yuan, in fact, wants to cut the cake of traditional Bluetooth speakers, "If you can convert this part of the stock market is also good. In the second half if you see a lot of smart speaker prices, or It will not be a surprise when there are a large number of one or two hundred smart speakers on the market." “The next 6-12 months are critical for the smart speaker market in China. There will be many brands and products at this time, which can play a role in educating the market.” Wei Qiang predicts the domestic speaker market. As for the smart speaker will become the next smart phone, or the next smart bracelet problem, Wei Qiang said he did not answer, "When the bracelet wind blowing, mainly the back end of the service is not done. Speaker also has this early This kind of problem has not been able to satisfy the user's service. To make the smart speaker become a user's necessities, it is necessary to create a killer application." Second, voice interaction is an irreversible tide The heavy investment behind smart speakers is the expectation of these companies for a major change in voice interaction. Siti Chi CMO Long Mengzhu often runs at the forefront of the smart voice market. She believes that Amazon's Echo is a fortuitous product, but Alexa is a necessity. Long Mengzhu believes that Internet or voice technology companies are not willing to drink alcohol, and most of their abacus behind them is to promote their voice interaction technology with the sound of speakers. "We don't know how to do that on the PC's entrance. Is it the voice control that powers on and off? But it will bring about cognitive changes to the user. Another interesting point is that we have a partner who wants to The voice interaction was added to the toilet and the voice was used to control the toilet to flush.” Long Meng Zhu said half jokingly. Today's smart voices, like the rise of touch screens in the past, changed all the places to touch interactions, and blindly changed all operations into voice interactions. From the current point of view, the trend of voice interaction is an irreversible trend. However, voice interaction is a completely new way of interaction, and the use scenarios and products are still being explored. Baidu COO Lu Qi once said that the key to landing artificial intelligence is to find scenes and business models, make the ultimate experience, and quickly iterate. Voice interaction itself is a big entrance, but the entrance of voice interaction is not a smart speaker, and now nobody knows. But for sure, smart speakers are the first generation of voice interactive products. Xie Dianxia believes that the smart speaker market in China is not as optimistic as we thought, but it does play a big role in the early days of voice interaction. "But the activity, stickiness and retention rate of smart speakers in China may not necessarily be as high as in the US market." Wei Qiang said that the smart speaker is an entry-level product. His profit point is not in the hardware itself, but in the future of the back-end content services. Making money from hardware is not our main goal. “Smart speaker We have been doing it for two years. The product experience is the most important thing. On the one hand, it is to improve the user’s understanding of the product. On the other hand, we need to do a good job of the background service.” But in Long Mengzhu's view, smart speakers shouldn't be called smart speakers. “One day, maybe a manufacturer has made a robot-shaped smart speaker. It broke out? It's just a robot or a speaker, if you guys Now I would like to post concepts, I would rather define it as a new smart product, weakening the concept of smart speakers. The key is to find accurate audiences and scenes.” Long Mengzhu said. Zhu Junwen thinks that there are currently three entry-level products for voice interaction: smart speakers, smart story machines, and smart home appliances. “The three product cycles are different, and the speakers are in the market cultivation period. Children’s smart story machines have entered rapid development this year, and smart home appliances will be able to complete market education next year. In addition, the demand for smart lights may also rise.” According to Zhu Junwen's point of view, the outbreak of smart speaker sales in China is still next year. "The focus of future competition is on the smart cloud platform, in which the Internet company is an important force. The entire industry hopes to have a complete cloud platform solution to do." Zhu Junwen said. Baidu's DuerOS will focus on the use of scenes in families, cars, mobile phones and other rapid iterative scenes. Jing Hao believes that the demand for voice in these three scenarios is gradually warming up. The demand for home and car is from 0 to 1, and the voice assistant on the mobile phone has existed before, and its subsequent role will become more and more important. "TV is definitely an important carrier for future speech interactions." Jing Ye believes that with the ability to have speech, the interaction and vibrancy of users and devices will increase dramatically. Liu Rui is optimistic about both home and car scenes. “Voice interactions are used in smart home scenes, which require input, output, and an information presentation center. This information presents the center and a larger screen is required. It may be TV. Car scenes may also need better network, such as the popularity of 5G." According to Xie Dianxia, ​​smart speakers are an emerging market in China, and TVs, refrigerators, and air conditioners are a stock market. "There is a chance that the intelligent voice interactive TV will run faster." Xie Dianxia said, "The future will really play an important role in the family, may still be television, and I also look at smart lamps, children's stories, etc." Third, decentralization: talk about smart home control theory In Xie Dianxia's opinion, from the perspective of the dimensionality of human-computer interaction development, the greatest contribution of speech interaction is to provide a control dimension. "What is the difference between current mobile phone apps and web pages and software on PCs in the past? The main difference is that app has been upgraded. This dimension is a lot of sensors, microphones and cameras." Xie Dianxia explained, "Because of the GPS sensors, There are companies like Drops, Uber, and Ofo. Based on this, plus voice and microphone, there will be WeChat and WhatsApp. If you add a camera, you will have Instagram." According to Xie Dianxia's point of view, when APP was promoted, there was no company or product in the PC era. With voice interactions in the future, another dimension is added, which is the control dimension. Xie Dianxia believes that the sensors in the past are just feelings. Now that the sensors are controls, the imagination is very great. The shape of the products in the future will change dramatically. Smart speakers are just the tip of the iceberg in a big era. However, there is a pit here: smart speakers should not be the central control device to control other home appliances, but should be decentralized. As Zhu Junwen said, before Echo, there was a lot of people talking about central control. The core of this is the central control model, but Amazon’s Echo is decentralized. "The voice interactive robot is the future development direction, the future voice interaction will be Always On (real-time online), all the appliances in the home should have the ability of voice interaction." Xie Dianxia so arrived. Liu Rui believes that the voice collection terminal of the future should be distributed in various parts of the household, just like Echo Dot. "The truly intelligent experience is that it should be distributed in all corners of the family and be decentralized." In other words, smart home control is still a beautiful imagination. "Echo's most important function to impress people first is to turn on and off the light. The feedback is very strong and very fast. Positive feedback is a very fundamental aspect of human nature." Liu Rui said that there is a point in Echo's fire that really is because of smart homes. However, if China's smart speaker market is betting on smart home control, it is currently a very risky thing. Wei Qiang believes that home internet of things is a process and must form a unified set of standards. Although smart homes are a trend, the iterative update speed is too slow and the path of implementation will be very tortuous. Zhang Peng, who is responsible for Panzhiora's project, believes that using audio to control home is the future trend, but the central control is not necessarily to use speakers to present, can be any type of existence. However, in Song Shaopeng's view, smart speakers do not necessarily have to control the home to become an entrance, as long as they can carry streaming media, information playback, information acquisition and distribution, it is already an entrance. "Before the smart home was really realized, all the predictions were just a kind of speech." Liu Rui said that the best way to predict the future is to join in and create it. At least, from the current point of view, smart home is actually "the emperor's new clothes." Pay attention to NetEase smart public number (smartman163), get the latest report of artificial intelligence industry.