Reporter: Spectrum / Vicky Xiao Editor: Vicky Xiao
Wu Enda has not been for a long timeone thingSo excited. I have not slept for a few days. "He said in an interview with PingWest.
Traversing Wu Enda's past: from Prof. Stanford, to the leader of large-scale projects in the industry, to the founder and director of the start-up company, to the chief scientist of the large technology company, Wu Enda made the senior positions in the technology industry itself.
He first joined Google to create the first large-scale distributed depth learning system and the Google brain, laying a solid foundation for more in-depth study and product projects after Google and the industry as a whole; Course Education (MOOC) Website Coursera. Later, he joined Baidu as chief scientist, led his American Institute, Baidu to fight the basic research and commercial production of artificial intelligence and other leading Chinese technology companies.
Wu Enda who, from where to have no details. But when he left Baidu's chief scientist's position, everyone was concerned about a question: where would he go?
Some people think that in accordance with the routine of another large technology companies dug up Wu Enda, it was also speculated that he will join the students to create the automatic driving companyDrive.aiThe No one thought, he went back to do one of his most loving things & mdash; & mdash; & ldquo; when mentor & rdquo ;.
Today, by his in-depth study of the course& Ldquo; Deep Learning Specialization & rdquo;Officially on the Coursera line.
So far the most comprehensive, systematic and easy access to the depth of learning online courses
Wu Enda from May this year began planning, in June will announce their new projectsDeeplearning.ai, And online courses & ldquo; Deep Learning Specialization & rdquo; is the first product of the new project. This course is a total of 5 students, students can not only learn the convolution of the neural network, recurrent neural network and other basic knowledge of the depth of learning, but also in the Wu Enda and his Stanford University assistant personally under the guidance of the construction of a deep neural network, and use it Solve the problems in the real world.
"We will help you master the depth of learning, understand how to apply the depth of learning, in the artificial intelligence industry to open your career. "Said Wu Enda.
Artificial intelligence online education, Wu Enda is the initiator. In 2012, Wu Enda founded Coursera, bringing the Stanford Block CS229 "Machine Learning" course to the Internet. This class of 1.5 million people started the machine learning and mdash; this artificial intelligence main research areas, many of them have become the artificial intelligence industry, the basic talent.
But Wu Enda found that artificial intelligence and industry in recent years the rapid development of cutting-edge depth of learning technology after another, "machine learning contains a lot of aspects, many of which are now very important to help you form an overview of artificial intelligence." In recent years, the depth of learning is progressing faster, and I want to teach these courses more meaningful. & Rdquo;
This change, do not need Wu Enda, even ordinary machine learning researchers and technology practitioners can see. In-depth study is the most popular computer science category in recent years, and all around the discussion of artificial intelligence, the core of the 90% discussion is one aspect of deep learning.
Many courses, including Coursera, Udacity, etc., have in-depth study courses on many of the online science learning sites. However, most courses only focus on specific directions, providing limited knowledge, being finer and less systematic. Online has not been more comprehensive depth of the course, making high-quality depth of learning has become a teacher-oriented scarce educational resources.
And & ldquo; Deep Learning Specialization & rdquo; the greatest meaning is to fill this vacancy.
The current 5 classes in the first three classes are ready, the remaining two still need a few weeks time clip production. "Wu Enda told PingWest goods to play, and I will be teaching. There will also be two assistants in the course. The two were Kian Katan and Younes Mourri, the former or Stanford 2016 Engineering Institute of the best teaching assistant.
This course is another surprise: Wu Enda will also interview in the course of video Jeff & middot; Sindon and other struggles in the front line of absolute authority, these depth of the field of top experts and scholars personal experience and stories to share with students & mdash; & mdash ; For the initial study of the depth of the computer, the light of the contents of the value has been back to the price.
Coursera special fees (special courses), in accordance with the monthly fee system charges $ 49 per month, this course is no exception. However, during the first seven days of the course, Coursera allowed the user to have unlimited access to all content within the course, as well as a very favorable program for students with financial difficulties.
In describing the nature of deeplearning.ai, Wu Enda emphasizes that it is not a traditional "entrepreneurial company", that is, he is pursuing something bigger than business success. "I really do with Kian, Younes in thinking how we can do better education in depth, so that more people have the opportunity to learn, so do this course." "Said Wu Enda.
At present the team size of this project is not large, mainly including his assistant and some video clip staff. Wu Enda said they treat deeplearning.ai's attitude, and the goal is not to profit, but to achieve the project's self-sustaining. This class is not a product to make money. & Rdquo;
Leaving Baidu's Wu Enda undoubtedly have a lot of choices, in fact, he has not stopped. At present, he is driving a car in a car companyDrive.aiAs a director; for Coursera, he is no longer responsible for day-to-day management, but the development of Coursera and content support is very heart. But he never thinks he is a classic standard on the corporate executives. He felt his management style more like "mentor" (Mentor). "I have been very proud of the team I have trained, because they have not been able to do very well. Coursera is the case, Google brain is the case, Baidu's artificial intelligence division is the case. "He said.
Whether it is before the start of the company, to the current AI teaching, "mentor" this identity, for Wu Enda, has not changed.
The deeplearning.ai project is closely related to China
However, do not think he had to build the first large-scale distributed depth learning system in Google, he left the industry. Wu Enda to PingWest products show that he is still in the incubation & rdquo; at least two new projects. He said that through these two projects, I hope to be able to develop some of the artificial intelligence tools that can be used by millions of people. & Rdquo;
And Wu Enda hope that at least one of these projects will be related to China, although the new launch of the "Deep Learning Specialization" curriculum video is currently unable to watch the normal in the mainland, but he is hoping for the Chinese artificial intelligence community to do something. And he is also very optimistic about the development of China's artificial intelligence projects.
"The United States is more strong in basic research, and China's industry in the artificial intelligence technology into commercial products is very positive, has become a leading application of artificial intelligence countries. China State Department announced beforeArtificial Intelligence ProgramIt was very helpful and I was very excited about it. I think research and application are very important, China and the United States should strengthen cooperation in artificial intelligence. "He said.
But Wu Enda did not hatch new projects, especially in China and related projects revealed more details, but said that there will be new news in the near future.
"Artificial intelligence will take off in China faster than in the United States. I hope to stay in China for a longer period of time to support the development of Chinese artificial intelligence ecology. "He said.
Through educationAn artificial intelligence driven world
July 16, Wu Enda on TwitterBursts of tweets, Describe their understanding of the timing of the birth of technology. He believes that the 1993 Apple handheld device Newton (Newton) is very innovative, but because of the screen, battery and chip technology constraints, until the iPhone was really a revolutionary new product in 2007. Geoff Hinton, Yann LeCun et al. Proposed a deep neural network algorithm with interdisciplinary significance in the 1990s, but until recently, the computational performance has been followed up. Intelligence is once again a topic that can be discussed even by ordinary people.
And with the arrival of this opportunity, Wu Enda inWrote on Medium, Artificial intelligence will be the same as electricity, completely change the whole world.
"I hope we can build a society driven by artificial intelligence, so that everyone can afford health insurance and automatic driving car, each child can enjoy personalized education." & Rdquo;
An artificial intelligence driven society. Artificial intelligence will integrate with healthcare, transportation, finance, society, and even government, improve the efficiency of old mechanisms, or create new ones. For example, in financial security, some companies have used in-depth learning-driven computer vision for face recognition, verification of transaction security, or the use of large-scale data training neural networks, the training results used to detect potential high Risk trading.
"But this future depends only on a company can not be achieved. "Even if the huge companies such as Google, Facebook, Apple, Microsoft or Baidu, in the artificial intelligence technology into the medical, transportation, financial and other industries, but also need to cooperate with other companies." This will require people around the world to build such a grand AI system. In the future society, the depth of learning or more advanced artificial intelligence technology, may become an indispensable skills to survive and work.
This is exactly the purpose of launching deeplearning.ai, and the depth of the course is the first shot of the project. He wants to be able to become one of the millions of artificial intelligence developers "teachers" to give them the ability to learn in depth with real world problems and tools to all walks of life to develop more meaningful artificial intelligence system to achieve his ideal: to create An artificial intelligence driven world.
When talking about his goal, Wu Enda said, "I know it sounds a bit stereotyped, but I really think so & Rdquo;