In order to help Chinese developers, Google is also very worried

On September 20, taking advantage of the afterglow of the World Artificial Intelligence Conference, Google held the 2018 China Developer Conference (Google Developer Days) at the Shanghai World Expo Center. Compared with last year, Google held the China Developer Conference two months earlier this year. In terms of content, it focused more on technology updates and application development itself.

Obviously, this is a real event for developers.

At around 9:30 a.m., the conference officially began. Liang Xinping, senior product manager of Google AI, took the stage and said that Google is committed to promoting the development of artificial intelligence and machine learning in the Chinese market. To this end, Google has established the Google AI China Center and launched a free machine learning crash course in China. In addition, it will also hold a Google AI machine learning application winter camp.

Anna Jin, a software engineer at Google AI, said that since its release, TensorFlow has been downloaded 17 million times, with 39,000 code submissions and 1,600 contributors; in China, it has been downloaded 2 million times and has more than 70,000 followers on its WeChat public account.

At this conference, Google made some new announcements for developer tools, improved applications, and machine learning. Specifically:

• In terms of application development, Flutter Release Preview 2 was launched, which is also the last preview version before the official version 1.0. Flutter, a development tool, supports Android and iOS platforms. In fact, there are more developers using Flutter in China than anywhere else in the world, including apps such as Xianyu and Now Live. In Flutter Release Preview 2, Google increased its investment in consistent development, focusing on the performance and stability of iOS; and invested heavily in building Cupertino Widgets, with which developers can develop applications that meet the appearance and functionality requirements of iOS users.

• Wear OS By Google has launched a series of new features and redesigned the application experience and UI. The new features allow smartwatch users to display the most important messages and notifications, rich health and fitness information, and support smart replies. In addition, in China, Google and Mobvoi have conducted in-depth cooperation on the Chinese version of Wear OS By Google. For example, the Chinese version is equipped with Mobvoi Chinese Assistant developed by Mobvoi, the application store and watch faces jointly developed by the two parties, etc. Google said that China is already the second largest market for Wear OS by Google in the world.

• For the development of ARCore, Google has launched a Chinese official website: developers.google.cn/ar, which can help Chinese developers better develop AR applications.

• Based on the previously popular Guess the Painting app, the "Play with Friends" mode was launched. In multiplayer mode, users can draw with up to 7 friends at the same time, see their friends' drawings in real time, and check the scoreboard after each round; Google said that this new feature is designed to help Chinese users experience AI in a more interesting and social way.

In addition, Google also introduced the latest developments in Android 9 Pie, Firebase, Android Jetpace, etc. to developers at the conference. In the exhibition area at the conference, Leifeng.com also visited the Google AI, Android 9 Pie, Wear OS By Google, ARCore and other exhibition areas.

Of course, for developers, in addition to providing the most basic development tools, Google has also done a lot of work in application promotion, application export, overseas localization, etc. In order to find out specifically how Google helps Chinese developers, Leifeng.com conducted an exclusive interview with some developers at the conference to get a glimpse of what Google has done.

Take LingoDeer, an app that focuses on helping foreigners learn Chinese, Japanese, Korean and other minority languages, as an example.

LingoDeer founder Wang Zhulong told Leifeng.com that LingoDeer App was actually aimed at overseas users from the beginning, so it tried to acquire users on Facebook and Twitter platforms in the early stage, but there were still many obstacles. By chance, after LingoDeer learned about and tried some functions of Firebase, the backend of Google Play, it found that it could help it better understand and analyze user data and user behavior, improve its understanding of users, and thus help the APP reach users more accurately and get feedback.

As for user acquisition, Wang Zhulong said it was very difficult. After LingoDeer was first released on Google Play in September last year, it unexpectedly gained a group of loyal fans and showed it on YouTube. However, thanks to the quality of the app itself, this APP was recommended by Google Play one month after its launch, and was even shortlisted for the Android Excellence apps list announced by Google in July 2018.

Wang Zhulong said that with zero promotion costs, LingoDeer has received more than 3.5 million downloads, nearly 170,000 reviews, and an overall user rating of 4.91 points.

Compared with LingoDeer, another developer interviewed by Leifeng.com - Chen Mingquan, who graduated from the Computer Science Department of Zhejiang University - has benefited a lot from Google in terms of technology and localization.

In August 2015, Chen Mingquan created an app called "Xingse", which can realize intelligent plant recognition based on image recognition and deep learning technology, and is known as the "flower recognition magic tool". Chen Mingquan told Leifeng.com that he initially used AWS machines to realize backend computing, and also tried frameworks such as Facebook's Caffe and Amazon's MXNet on the model; later he tried TensorFlow and called some of its instruction sets to try, and found that the latter could achieve an 80% improvement in recognition efficiency, so he decisively chose TensorFlow and Google Cloud.

Moreover, when the "Xingse" APP was globalizing and localizing overseas under the name of "PictureThis", Google also provided a lot of help. For example, "Xingse" used ancient Chinese poems to introduce plants, but the Japanese version of PictureThis used traditional Japanese haiku, which seemed more localized. This localization strategy came from the help of the Google Play team. For example, the Google Play Japan team invited local Japanese users to experience and provide feedback on the first localized version of PictureThis. In the end, these local users also provided PictureThis with a lot of more authentic and localized plant introduction content.

Of course, in addition to these two developers, Google has also provided a lot of help to other Chinese developers; data from Google shows that the number of Chinese developers increased threefold in 2017, and the revenue of Chinese developers on Google Play increased by 50%.

Compared with the Google China Developer Conference in December 2017, this conference did not seem to be as important in terms of product and event releases, but judging from the atmosphere felt on the scene by Leifeng.com, Chinese developers' enthusiasm for Google has only increased. This also confirms Leifeng.com's judgment that this is, after all, a grand event for developers.

Of course, in the existing multi-faceted relationship between Google and the Chinese market, developers also play an extremely important role. In a sense, the AI ​​for everyone that Google has been advocating absolutely requires the support of Chinese developers. This mutual relationship based on mutual needs is undoubtedly one of the ties between Google and China.

Why should we learn from Xiaomi when doing IoT?

Meituan is listed! Its market value surpasses Xiaomi and JD.com. Wang Xing thanks Steve Jobs

Detailed explanation of Tencent's AI layout, 2B business may rise