After spending an afternoon with Du Qiu yesterday, the cold feeling between the two disappeared a lot. When he saw him walking into the box, he casually introduced: "Qiu, this is Dr. Huang Xuedong from Microsoft's Redmond Research Institute. I met by chance in a hotel cafe yesterday. I asked him to be a technical consultant for me temporarily. Don't you mind?"

It was him indeed...

Du Qiu had heard Huang Xuedong's academic lectures before traveling through time. At that time, he was already the world's top voice recognition technology expert. He was a dual academician of two top computer associations, IEEE and A, and was also Microsoft's only outstanding Chinese technical engineer. The word error rate of the speech recognition system he led was only about 6, which was comparable to professional stenographers. The results were used in smart voice projects familiar to ordinary netizens such as bg, rtana, skype, and Xiaobing chat robots.

In addition to technology, this person's social circle is also very wide. Microsoft founded the China Research Institute in Beijing in 1998, and later it was upgraded to the Asia Research Institute. There were four deans, including the first Kai-Fu Lee, the third Chen Xiangyang, and the fourth Hong Xiaowen and Huang Xuedong, who had deep connections, because they both studied at Carnegie Mellon University in the United States, and they were both students of Turing Award winner Raj Reddy. Kai-Fu Lee and Hong Xiaowen were also two of them who joined Microsoft after Huang Xuedong lobbled.

"Dr. Huang, welcome, why don't you notify me in advance or He Lixing?" Du Qiu shook hands with Huang Xuedong enthusiastically and said, "When I was in Seattle last month, I wanted to visit you, but unfortunately, because the work schedule was full, I couldn't make up for the regrets at that time. Today I finally made up for the regrets at that time."

"If you want to say regrets, it is actually my regret. I only found out afterwards that Mr. Du came to Seattle. He did not have the full friendship of landlords at that time, and it was very rude."

Huang Xuedong used Chinese, and Jennifer couldn't understand it, but he guessed what it meant. He was a little surprised and asked, "Do you know each other?"

"Manager He from the Snapdragon operating system department is an alumni of Dr. Huang. With his introduction, Dr. Huang and I have contacted each other by email before, but have never met him." Du Qiu sat down opposite Jennifer and asked, "Did He Lixing know that Dr. Huang has come to Yuncheng?"

He Lixing studied for a bachelor's degree at Qingcheng University, and then went to Carnegie Mellon University to study for a master's degree. Huang Xuedong studied for a master's degree at Qingcheng University and did postdoctoral research at Carnegie Mellon University. The two are dual alumni. Although they have been a few classes apart and have never studied or worked together, this relationship has been enough to establish friendship in a foreign country. After Du Qiu traveled through time, he has always wanted to recruit outstanding talents from studying abroad, so naturally he would not miss this kind of connection. He had contacted him a long time ago through He Lixing.

"It was Xiao He who invited me to come." Huang Xuedong is from Hunan and has a strong Mandarin accent. In order to take care of Jennifer, he switched to English and said, "I went back to my hometown a few days ago to attend a friend's wedding. I saw news about the Internet conference in the newspaper, so I called Xiao He. I heard that he introduced a lot about Snapdragon. I was curious and took the train from Tanzhou yesterday afternoon to take a look."

"So that's the case. Manager He will be having dinner with experts from the Chinese Academy of Sciences. I will give him a holiday in the afternoon and let you reminisce about the past." Du Qiu's hungry chest was pressed against his back, raised his chopsticks and greeted the two of them for dinner, and said, "Jennifer, how did you know Dr. Huang?"

"In 1992, Dr. Huang was responsible for developing a speech recognition project called Sph-ii, which won the best in the Darpa selection. When I was a reporter, I read reports about him, so I recognized him after meeting him in a hotel cafe yesterday."

Such a coincidence...

Darpa is the abbreviation of the U.S. Department of Defense Advanced Research Projects Agency. This institution has played a very important role in the history of IT because it invented the world-changing technologies such as the Internet and user graphical interfaces under its funding and management.

Sph is a project that Kai-Fu Lee started during his PhD in Carnegie Mellon University. After graduation, he was taken over by Huang Xuedong and Hong Xiaowen, and developed a better second-generation product. In 2016, this project still exists and has evolved to the fourth generation. It is the best open source voice recognition system in the world.

In 2016, speech recognition was a very popular computer technology. Du Qiu had never been engaged in research on this aspect, but he was not ignorant. Because speech recognition is closely related to neural networks, and neural networks are closely related to distributed computing, he knew some of the basics. Otherwise, he would not have listened to Huang Xuedong's academic lectures in particular, so he showed off his skills and talked to him about speech recognition technology.

"Now, speech recognition mainly uses the Hidden Markov model, and the Markov model is probably like this..."

"There is a computer scientist named Fukushima Hanhiko in Japan who proposed a model of a convolutional neural network in 1980. In 1998, Dr. Young Lechun from Bell Laboratory further improved the theory based on it. I think this model is very useful in speech recognition..."

"Mr. Du also has research on neural networks? You are right. Neural networks have great prospects in speech recognition, but there are currently two flaws, which lead to poor results. One is that the hardware performance is insufficient and the computing volume cannot keep up. The other is that the number of training layers of neural networks is too small, and the data that can be processed is very limited..."

Both of them were technical mates, and they were having a lot of chatting. They treated Jennifer as a decoration. She silently ate a lion's head, two small pieces of roast duck rolls, and drank half a bowl of lily ribs and vegetable soup. She couldn't help it and said, "Two gentlemen, can you take care of me and stop talking about computers?"

"Sorry, I didn't expect that Mr. Du also knew a lot about voice recognition, so I couldn't help but chat for a few more words." Huang Xuedong raised his chopsticks and said with a smile: "I won't talk anymore, concentrate on eating and being an interpreter."

"Okay, Jennifer, I'll talk to you about investment topics you care about." Du Qiu apologized, pushed the turntable on the dining table, turned the squirrels close to him over the yellow croaker, and said, "This is a special dish in the hotel. It tastes good, there are few fish bones, you can try it."

"Is that so? Then I'll have some."

"If you find Dr. Huang as a consultant, you should be very clear about the technical details and commercial significance of the press conference this morning. I won't go into details." Du Qiu poured beer for Huang Xuedong, and said: "The Internet is an information medium. Although the technology is new, the business model is old, similar to newspapers, television and movies."

"You said the same thing to different people yesterday afternoon, and the content is king, right?"

"That's right, the money earned by the New York Times and Time magazine is much greater than the profits of paper mills and printing factories. The money earned by the Time Warner and BS is much greater than the profits of projectors and TV set manufacturers. For Internet companies, the development prospects of content manufacturers are much greater than tool software manufacturers, so I think Yahoo's future is far stronger than Netscape."

Jennifer, who was clumsy and picking up fish with chopsticks, stopped and looked up and asked, "Where is the Snapdragon?"

"We are not tool and software manufacturers, we are platform service providers." Du Qiu vaguely avoided this topic and said: "The Internet has no national boundaries, but there are language and cultural restrictions. The most popular language in the world is Chinese and the most widely used language is English. Therefore, the most profitable content manufacturer in the future will only be produced in China and the United States. If you want to be an excellent investor, in addition to Silicon Valley, you should also open up the Chinese market as soon as possible."

Huang Xuedong asked after hearing this: "There is no company in China that specializes in venture capital, right?"

"No, but it will be there soon, because I plan to set up a private venture capital fund." Du Qiu raised his wine glass and smiled half-truely and tentatively: "If Dr. Huang is willing to return to China to start a business, he must contact me, and I am willing to give him all his help."

"Thank you Mr. Du for your kindness, but my personality is suitable for research and not for entrepreneurship."

"That's a pity." Du Qiu and Jennifer also touched the cups gently, then took a sip of beer lightly, and said: "Jennifer, I have ideas, funds, and technology. The only thing I lack now is the investment rules, processes and management knowledge. Are you interested in helping me?"

"Me? I just entered the venture capital industry in May. I don't know much, so I'm afraid I can't help much." Although Jennifer refused so, he asked interestedly: "But if the investment amount is relatively small, I am willing to use private funds to cooperate with you and accumulate some experience."

"The cost of starting a business in China is far lower than that in the United States, and the funds need not be used. I have three projects in my hand, and the investment amount is not large. The most one is only about 500,000 US dollars. If you are interested, I can arrange for you to have an interview with the entrepreneur."

"Can you talk about it now?"

"The first project was founded by Snapdragon employees. The main business is to build websites for government departments and large and medium-sized state-owned enterprises. We released the Yuncheng hotline the afternoon before yesterday. Yesterday, we received dozens of consultation calls and order intentions. There are hundreds of cities and thousands of state-owned enterprises in China. This is a project with very low risks and great potential."

Jennifer asked a little confused, "Isn't this a direct conflict with Snapdragon's business?"

"I don't like dealing with government officials, especially the commission orders related to design and technology, which are very troublesome because they don't know anything, but they like to point fingers." Du Qiu smiled sarcastically and explained: "Website construction is not the focus of Snapdragon's development, but this business requires a lot of people. Keeping it will make the company bloated and chaotic, so it's better to go out independently and encourage those employees with innovative and adventurous spirit to develop freely."

"You are such a strange person." Jennifer said without asking the specific amount of investment, "But I agree with you, doing business with the government is a very troublesome thing. What are the other two projects?"

"There is a postdoctoral fellow at MIT who is preparing to return to China to start a business, build a website called China Online, and use the Internet to provide Chinese business information to overseas investors."

"The information is paid, right?" Jennifer asked with a sharp business mind. "What are his advantages in this kind of business Dow Jones, Bloomberg and Reuters?"

"His academic qualifications are advantages. Returned overseas students are very popular in China, especially famous schools like MIT." Du Qiu briefly introduced Zhang Chaoyang's resume and said: "Last month, he accompanied the president of MIT to visit China and was received by Vice Premier Zhu. This experience is very useful in China and can allow him to enter the circle of high-level society and obtain information that news agencies such as Dow Jones and Bloomberg cannot access."

Jennifer looked at Huang Xuedong and saw that he nodded and said that it was indeed the case, and said, "Okay, this is a good project, and there is another one?"

"There is another project that is an e-commerce website that serves small and medium-sized enterprises, but the partner I chose has not decided whether to do it or not. Let's talk about it later."
Chapter completed!