Baidu’s Robin Li Introduces Firm’s Feedback-Driven Innovation

At the 2022 Baidu Create occasion on Tuesday, agency CEO Robin Li delivered a speech about feedback-driven innovation and its affect on China’s actual financial system, exhibiting the tech large’s expectations for the longer term golden decade of synthetic intelligence improvement.

In the previous 12 months, there was much less dialogue about “promising opportunity” and extra consideration given to the well being of sure development indicators.

From a graph of world GDP per capita up to now 2,000 years, it may be seen that within the first 1,800 years, the change was not apparent. In the final 250 years or so, it has elevated sharply, pushed by a number of scientific and technological revolutions. It was technological improvements which have pushed nice development.

Robin Li identified that the image of the fourth scientific and technological revolution is the deep studying algorithm that will carry effectivity enchancment and financial development. Major improvements associated to deep studying, together with automated driving and clever dispatching methods within the fields of hydropower and power, can have an important social influence like innovations reminiscent of vehicles and the web.

Baidu’s Feedback-Driven Innovation

Robin Li gave precedence to discussing “feedback-driven innovation.” Some scientists have executed a thought experiment – how lengthy will it take a blind man to resolve a Rubik’s Cube if he rotates it as soon as each second? The reply is 13.7 billion years. But if each time the blind man turns the Rubik’s Cube, somebody provides him suggestions, telling him whether or not he’s nearer to the purpose or farther away from the purpose, he would take two and a half minutes to complete the mission. This is the magical energy of suggestions.

Therefore, innovation can solely be realized when we now have the chance to enter the market and get suggestions from customers and prospects.

Baidu has a whole lot of sensible expertise in “feedback-driven innovation” in its enterprise improvement.

For instance, Baidu’s Kunlun synthetic intelligence chips are very superior as a result of they’ve been optimized for Baidu’s search service for 10 years. Baidu’s search service responds to billions of actual customers’ wants day by day, and conducts 1 trillion deep semantic reasoning and matching day by day, which may present the truest and most well timed suggestions, thus forcing the optimization of deep studying frameworks and chips.

Baidu has revealed orders from Apollo Go, its autonomous driving companies platform, each quarter since a 12 months in the past. The giant order quantity signifies that it may get essentially the most helpful suggestions. Baidu seen from social platforms that net customers in lots of locations hope that the unmanned car journey service may be extra accessible, which signifies that folks’s acceptance of automated driving could be very excessive, and business operation could also be sooner than anticipated.

Baidu’s enterprise scope covers chips, frameworks, fashions, and purposes. From the high-end chip Kunlun, to the deep studying framework “PaddlePaddle,” then to the big mannequin “Wenxin,” there are key self-developed applied sciences that may generate varied suggestions in any respect ranges.

Baidu’s Influence on the Real Economy

First, the brink for creating AI know-how is decreasing.

PaddlePaddle is an open supply deep studying framework developed by Baidu. At current, it has gathered 5.35 million builders, served 200,000 enterprises and establishments, and created 670,000 fashions, which has constructed a affluent deep studying ecology. It permits builders to construct AI purposes like constructing blocks. Software is important for innovation.

As the technical base of many purposes, the massive mannequin can successfully combine pure language processing, pc imaginative and prescient, clever voice, and extra, and may be optimized together with eventualities of assorted industries and companies, thus eliminating the standard fragmentation improvement mode.

Second, AI empowers the event of the true financial system.

Robin Li regards clever transportation as an clever dispatching system that may management site visitors move via clever site visitors lights, thus bettering effectivity. Through the clever transformation of the transportation community, site visitors effectivity may be improved by 15% to 30%. Baidu’s clever transportation scheme has been applied in 63 cities. Li predicts that the clever transportation scheme could make China’s first-tier cities not want to limit car purchases earlier than 2027, thus activating car use. Before 2032, the congestion drawback can principally be solved by bettering site visitors effectivity.

At the Baidu Create occasion final 12 months, Robin Li mentioned, “With the continuous lowering of the threshold of technology application, creators will usher in the golden decade of artificial intelligence.” But the method of realization won’t be straightforward. This is the attribute of technological improvement.

Opportunities and Challenges During Development

In January 2000, Robin Li returned to China to begin a enterprise. Just two months later, the web trade shook and the world’s market worth evaporated by 8 trillion yuan ($1,184 trillion). At that point, many individuals selected to surrender and people who endured made the golden decade of web improvement.

Ups and downs like it will nonetheless occur at some point within the AI area. The silver lining is that previously 12 months, AI has modified its route each on the technical stage and on the business software stage.

At the technical stage, AI has moved from understanding content material to robotically producing content material reminiscent of footage, texts and movies.

The most consultant change in enterprise purposes is automated driving. In the previous, everybody thought that L2 via L5 got here step-by-step. But Robin Li believes that after L2 comes L4. Because it’s tough to outline the accident legal responsibility of L3, it takes longer to popularize it. As of September final 12 months, Baidu’s L4 automated driving service orders exceeded 1.4 million in additional than 10 cities reminiscent of Beijing and Shanghai, and all unmanned business operations had been opened in Chongqing and Wuhan.

SEE ALSO: Baidu’s Robin Li: Next Commercial Stage of Autonomous Driving Is L4, Not L3

However, the challenges are nice. The digital transformation of many fields in the true financial system has not been accomplished, and digitalization itself has not led to a big enchancment in effectivity. It will take time for intelligence to penetrate extensively. Therefore, the commercialization of AI must be explored for a while.

On Baidu’s AI portray platform, a picture is robotically generated by AI with “crisis” and “hope” because the key phrases. There is not any easy crusing. For nice enterprises and nice creators, triumph comes after difficulties. Difficulties will stimulate innovation, and innovation is the true driving pressure of development.