The concept of smart Home AI took shape in 2014 when Amazon built a smart speaker called Echo around its voice assistant Alexa. The Echo wasn’t as big a launch as the Fire Phone, but two years later, the Fire Phone has passed the market’s litmus test and the Echo has shipped millions of units. In May, Google launched Google Home, a hub for connecting smart Home devices, to compete with Amazon’s Echo in the market for Home AI. Rokid, an AI-powered home robot, has completed its Series B financing and is valued at $450 million. We talked to Misa, founder of Rokid, about what makes this two-year-old, 90-strong team so popular with capital.

Giiso Information, founded in 2013, is a leading technology provider in the field of “artificial intelligence + information” in China, with top technologies in big data mining, intelligent semantics, knowledge mapping and other fields. At the same time, its research and development products include editing robots, writing robots and other artificial intelligence products! With its strong technical strength, the company has received angel round investment at the beginning of its establishment, and received pre-A round investment of $5 million from GSR Venture Capital in August 2015.

What’s the difference between a startup being acquired and a startup building a product this time?

Misa, founder and CEO of Rokid, resigned from Alibaba two years ago to start his own business. He joined Alibaba when his startup Mammoth Technology was acquired by Alibaba and he joined alibaba to set up the mysterious M studio, which focuses on machine learning, speech recognition and portrait recognition. Quitting to start a business at this time is more like going back to your old job.

“Rokid had a prototype as early as 2014, and after nearly two years of continuous polishing and testing, we have the product we see today.” This is Misa’s overall description of Rokid’s first product, Rokid.

Ruoqi is a product based on machine deep learning. It adopts organic surface design as a whole and provides sliding touch induction on both sides. Volume and brightness can be controlled by sliding. According to Misa, “Wakwaken is two syllables, whereas Apple wakens Siri in three syllables (Hey, Siri). It’s not just one syllable.”

Siri is activated by saying “Hey, Siri” before answering a question, and then by saying “Hey, Siri” again if you want to continue the conversation. When faced with Ruoqi, you only need two syllables of “Ruoqi” to wake up Siri. After multiple rounds of conversation, Ruoqi successfully lights up the base and answers the user’s needs. And Ruki is more like calling a person’s name than waking up a machine like Siri. This is not just a slight improvement of one syllable, but Rokid’s breakthrough in voice interaction technology and user experience.

AI products ultimately land on services and content

Voice interaction is just a physical experience

The top of Ruoqi is equipped with a 13-megapixel HIGH-DEFINITION camera, which supports gesture recognition and qr code scanning of mobile App login. The bottom is equipped with an array of eight microphones, which can receive 360-degree voice sources; Four full-frequency speakers and two heavy bass radiators, 360 degree omnidirectional high-quality sound system, through the “Ruoqi” users can listen to millions of high-quality music, but also enjoy a story, learn English, check the weather and other life-housekeeping services.

At present, Rokid has begun to cooperate with smart home appliance manufacturers to connect Rokid products into the control system of smart home devices, and control home devices by voice as in the movie will become a reality. At the same time, with the increase of users’ time using Rokid products, Ruoqi with automatic learning function can understand different users’ daily preferences and usage habits, in addition to personalized recommendations to users, but also more “intimate” control of home equipment.

Misa think especially eventually fall to the ground will be Home AI AI products to content and services provided by the user, voice interaction and touch control just a way of content and services, provide input for machine learning and mastering the human preferences, final purpose is to machine will be better able to provide the content of the human needs and conform to the service of human experience.

For example, Misa, who has a Ruoqi at home, said: “After using it for a period of time, you wake up and say hello to Ruoqi every day. Ruoqi will be awakened and answered by voice and then automatically play soothing music, gradually turn up the lights, open the curtains and report the weather for the day.” Human-computer interaction is no longer a mechanical response to instructions, but after “getting along” with human beings for a period of time, it can understand and complete what human beings want and need through a command.

Will China’s consumption desire and consumption power in Home AI surpass Europe and the United States?

Different from the previous Internet revolution, in the face of mobile Internet, especially the recent WAVE of AI/VR, the technological level of domestic enterprises can keep pace with the international technology giants. Misa cited Rokid’s LABS in Beijing and Silicon Valley as examples: “The LABS in China and the US can keep up with the international level in technology development. They may have their own advantages and disadvantages in single point of technology, but they are ahead of the international level in product polishing and product experience (compared with Google Home and Amazon Echo).” Misa said Echo is the old category of new play, is in the traditional sound on the basis of the addition of intelligent functions, and Rokid’s positioning is the new category of new play, if qi is not only a smart sound, more is to highlight the voice control and companionship.

As for the scale of Echo in foreign markets has reached millions, Misa believes that the domestic market scale of Home AI will exceed tens of millions. As for AI products, the consumption desire and consumption power of domestic users are higher than that of foreign countries, and the domestic market capacity is also higher than Europe and America, China will usher in the outbreak of smart home category in the next 2-3 years.

Rokid launched the product into the market at this time point, and Misa did not have much sales expectation for the first-generation product Ruoqi. Because there is no product similar to Rokid in China at present, Rokid, as a new forerunner, faces great risks, including the cost of educating the market, cultivating user habits, and many uncertain factors. Moreover, even the relatively high price and relatively good sales were not enough to make up for the previous two years of research and development.

Giiso information, founded in 2013, is the first domestic high-tech enterprise focusing on the research and development of intelligent information processing technology and the development and operation of core software for writing robots. At the beginning of its establishment, the company received angel round investment, and in August 2015, GSR Venture Capital received $5 million pre-A round of investment.

Three stages of consumer AI application products

As for the next development of Rokid, Misa gave three stages of Home AI, Portable AI and Personal AI: “Home AI is mainly through the polishing of technology, creating a good interactive experience in the early products, to the next stage of Portable

AI mainly provides richer content and services through the improvement of interactive experience, while Personal AI needs to enter the era of comprehensive intelligence, which is a long journey. Personally, I think it will take 4-5 or even 4-6 years to get to this road.”

From Home AI to Portable AI to Personal AI is a natural result with the development of the overall technical level. But the road ahead for Rokid has just begun, both in terms of technology and market development. (Sherwood)