How can artificial intelligence be applied in reality? Google is the key

The annual Google I/O developer conference arrived as scheduled on the 17th. Compared with previous years, this year's Google Developer Conference is more like a routine appearance in a transition year. Compared with last year, Google's strategic direction of "from mobile first to AI first" has not changed at all.

During the keynote speech session on the first day of I/O, Google did not display or release any dazzling black technologies as it did in the early years. Instead, it applied artificial intelligence technology to Google's existing product lines and released and improved some specific functions to enhance the user experience.

Although the whole process was a bit boring, and Google did not come up with the amazing black technology and dazzling technology as people expected, it has to be admitted that in reality, Google is still the best company among the current mainstream technology companies in promoting artificial intelligence technology to the application level. Google is using artificial intelligence technology more pragmatically to solve the actual "pain points" encountered by many users in the process of using its products.

The product line is fully AI-driven

At the IO conference that day, Google CEO Sundar Pichai pointed out that Google's core mission, which is to organize and sort out information about the world, has become clearer in the era of artificial intelligence. In this context, Google is rethinking its existing product lines.

The above picture clearly shows that Google's artificial intelligence technology is already driving Google's most important products including search, Google Cloud Services, Google Maps, Youtube, Google Photos, etc., and Google's artificial intelligence technology applied to specific product applications appears in the form of entities such as "Google Assistant".

With the help of the artificial intelligence technology behind Google Assistant, Google has re-optimized the user experience of almost all product lines, making these products more humane and smarter for users.

For example, the Google Home smart speaker with Google Assistant built in can understand the context of conversations between people and it, making the conversations closer to natural language. Google Home can also display information in real time on the screen of the entire user ecosystem, whether it is the user's smartphone screen or the user's smart TV. At the conference that day, Google Assistant even landed on the iPhone terminal, directly competing with Apple's Siri voice assistant.

It is worth mentioning that Google has also opened the developer tools of Google Assistant and has carried out extensive cooperation with third-party hardware brand manufacturers. This idea is actually the same as the previous Android, forming a broader ecosystem in an open form, and the core is still driven by Google's artificial intelligence technology.

Judging from the overall theme emphasized by Google at the conference that day, there was no change in the strategy of "transitioning from mobile first to intelligence first" from last year. The biggest change is that Google's commitment to "AI first" is being fulfilled through a more detailed and complete product experience.

In addition to the intelligent upgrades to its existing product lines, Google also released some new products at the conference that day, including the Google Lens application that uses artificial intelligence technology for image recognition. The application uses Google's image recognition function to identify specific information in the images captured by the camera. For example, by pointing the camera at a restaurant on the street, Google Lens can identify the name of the restaurant, previous user ratings, and a series of other information already in Google's huge database. The application can also recognize information such as phone numbers in pictures, allowing users to make direct calls.

Another highlight of Google's release that day was the release of the cloud-based second-generation Google deep learning chip TPU. Google CEO Sundar Pichai emphasized at the conference that TPU was used more in deep learning recognition in the past, while GPU was used more in artificial intelligence model and data training. In the future, TPU will provide powerful functions in both model search and recognition.

At the conference, Pichai proposed the concept of Google's artificial intelligence framework (Tensorflow) + everyone. In fact, this is similar to the "Democratize AI" proposed at the Google Cloud Conference a few months ago. The key point emphasized is that in the future, artificial intelligence will not only be something that large companies can do, but that small and medium-sized developers can use the artificial intelligence technology framework provided by large companies to create specific products and applications. The significance of this is actually tantamount to a technological revolution that promotes AI in an all-round way.

Google is still the company that does the best job in pushing artificial intelligence to the application level

Although from the perspective of ordinary users, the optimization of many functions and product updates launched by Google that day may not arouse the interest of too many people. From the user's perspective, perhaps many functions should have been provided by Google or other large companies long ago. But in fact, from the perspective of technical implementation, applying artificial intelligence technology to the product level is not an easy task. For Google engineers, perhaps a specific and small product function driven by artificial intelligence requires as much energy and time as developing a new product.

Therefore, at the Google Developer Conference or other developer conferences, such cognitive differences can be clearly felt. For ordinary users and even the media, Google's developer conferences are becoming more and more boring. There are no parachutes falling from the sky and no weird-looking glasses. Google is becoming more pragmatic, constantly optimizing and improving its products, which it excels at, to enhance user experience.

The basis for all these efforts made by Google is that its product line has deeply influenced a wide range of people. At the conference that day, Google announced that the total number of monthly active devices running the Android system has exceeded 2 billion. On this basis, Google products with more than 1 billion users include search, Chrome browser, maps, Youtube, Google Play App Store, Google Mail, etc.

At such a scale, the rapid development of artificial intelligence technology in recent years is actually a rare opportunity for Google. By using Google's artificial intelligence technology, Google can actually further expand its territory based on the original huge ecosystem. For example, Google Assistant is cooperating extensively with third-party manufacturers to promote its applications to a wider range of devices, and Android Auto is also cooperating with more car manufacturers to enter more car center consoles.

From the perspective of the entire industry, artificial intelligence has actually become the next focus emphasized by all major technology companies. However, like Google, Microsoft, Amazon, Facebook, etc. face similar problems, which is how to truly promote artificial intelligence technology to the level of application products that can reach users.

The industry has actually reached a critical juncture, where users have unprecedentedly high expectations for artificial intelligence and believe that AI can accomplish everything and change everything. However, from the perspective of companies, there is still a long way to go in the transition from artificial intelligence technology to application productization.

At present, major technology companies have found a consistent direction, which is to emphasize voice as a new form of interpersonal interaction. As a result, smart speakers including Amazon Alexa, Google Home, and Microsoft Cortana in cooperation with other hardware manufacturers have emerged.

From a technical perspective, the voice recognition technologies of various companies are similar and the differences are not that big. An expert in voice artificial intelligence technology told Tencent Technology that in fact, current smart audio products all use relatively mature technologies, and artificial intelligence voice recognition technology is already far ahead in academia. However, when it comes to the application level, it involves user experience issues. For companies, in terms of the trade-off between using the latest technology to improve recognition accuracy but increase latency and using mature technology, accepting relatively low accuracy but obtaining extremely small latency, companies often have to use more mature and reliable technologies from the perspective of user experience.

At present, in the field of voice human-computer interaction, Google's advantage lies in that it has firmly occupied the most advantageous position in the mobile Internet era. The Google ecosystem based on smartphones, smart TVs or other mobile smart devices has been widely formed and difficult to shake. At this time, it becomes natural to launch a new human-computer interaction interface based on artificial intelligence technology.

From the perspective of practical application, Google is actually the most successful company in the industry in pushing artificial intelligence technology to the application product level. Most of the specific products under the Android ecosystem currently used by the majority of users have already applied Google's artificial intelligence technology. Although from the user's perspective, the experience brought by Google's artificial intelligence technology is not perfect. For example, the image recognition technology that has been applied in the Google Photos application still has the problem of low accuracy when identifying objects in specific pictures. In addition, the recognition rate of Google Voice Assistant still needs to be improved, but it is undeniable that Google has taken a solid step towards comprehensive AI and is still in a leading position in the industry in this field.

As a winner of Toutiao's Qingyun Plan and Baijiahao's Bai+ Plan, the 2019 Baidu Digital Author of the Year, the Baijiahao's Most Popular Author in the Technology Field, the 2019 Sogou Technology and Culture Author, and the 2021 Baijiahao Quarterly Influential Creator, he has won many awards, including the 2013 Sohu Best Industry Media Person, the 2015 China New Media Entrepreneurship Competition Beijing Third Place, the 2015 Guangmang Experience Award, the 2015 China New Media Entrepreneurship Competition Finals Third Place, and the 2018 Baidu Dynamic Annual Powerful Celebrity.