Qianjiang University, School of Mathematical Sciences.

Under the guidance of Dean Cai, Gu Mojie was received by two big bulls.

In addition to the expected acquaintance Academician Qiu, there is also President Zhu Jingwu of Hong Kong University of Science and Technology.

HKUST and Qianjiang University have long-term exchange student cooperation, and it is not surprising that President Zhu came to visit Qianjiang University in China. What's more, Zhu Jingwu also has a double identity - he is the son-in-law of the old Professor Chern Shiingshen, so theoretically he and Academician Qiu can be regarded as a brother and sister.

A few days ago, after holding a memorial service for Mr. Chen, Academician Qiu and Zhu Jingwu wanted to go back to the United States and Hong Kong. However, there are no direct flights to the United States from Jinmen, and they all have to go to the Shanghai River, so the two of them coincidentally visited several universities in the Yangtze River Delta area, which can be regarded as a send-off.

With Qianjiang University's attainments in science and engineering, the interviewee bore the brunt.

With Gu Mojie's current status, he would not let those top bulls-eyed babes come to see him, so he could only be said to have brought them with him. But now that I've seen it, I can always benefit a lot. After all, Gu Mojie is now a "representative of the industry" in the field of artificial intelligence and deep learning, and he has to add a "one".

Principal Zhu Jingwu is engaged in physics and materials, and he and Gu Mojie are the first time they have met, but they only have time to get acquainted, and there is no industry topic to talk about.

President Zhu politely asked Gu Mojie if he was interested in getting an exchange student place and going to HKUST for graduate studies in the future. Gu Mojie also sincerely said that he would consider it in the future, but there are too many things at the moment, and it is really inconvenient to go to Hong Kong.

In contrast, Academician Qiu and Gu Mojie have a lot of topics to talk about. After Academician Qiu got older, he seemed to be more concerned about the industry dynamics in the field of algorithms, even if he didn't understand the technical details, he also had a thorough understanding of the cutting-edge trends in the industry.

We talked for a while. Academician Qiu preached to Gu Mojie earnestly:

"Xiao Gu, I see that your focus is still too narrow! Deep learning, so a good general direction, how many applications can be expanded. You were a person who drank saliva at the beginning, but after more than half a year, you are still working on cloud input and cloud security. Although I don't understand the business world. I don't know how to operate, and I also know that you have to attack on multiple fronts now, and it is to seize the opportunity.

I AM IN THE UNITED STATES, AND I AM CONCERNED ABOUT THIS TREND EVERY DAY, BUT GOOGLE HAS SPENT MORE THAN HALF A YEAR WORKING HARD TO MAKE A 'PICTURE SEARCH ENGINE', AND I HAVE SEEN THAT THEIR INTERNAL BETA VERSION IS VERY EASY TO USE; FACEBOOK HAS ALSO MADE 'FACE RECOGNITION TECHNOLOGY'. It was put into official use just a few days ago.

Deep learning algorithms, you only use them to 'make the computer automatically recognize sentences more accurately', is it overkill? Chinese should also do a little bit of 'let the computer automatically recognize pictures and recognize sounds more accurately' technical application! Don't keep following the back of the Americans' butts and see what others are doing to keep up. Young people, it is good to be responsible. ”

These words are a bit pedantic, and they do not take into account the constraints caused by Gu Mojie's current lack of funds, but they can be regarded as inspiring Gu Mojie a lot.

Gu Mojie himself knows if he is willing to be like Mark Zuckerberg. Do not hesitate to estimate the market value of your company. And then take hundreds of millions of dollars in investment at once, and then reduce his own equity to less than 50%, then he can grow as quickly as Zuckerberg. But he just doesn't want others to take his ride. Reluctant to cheapen those bankers.

Regarding Academician Qiu's suggestion, he did not refuse in person, but sincerely said that he would consider and investigate immediately. Dean Cai also said some words of encouragement on the side, after all, it is a combination of industry, education and research projects supported by the school, and Dean Cai also hopes that Gu Mojie will be able to give back to the school after he becomes bigger.

……

After the meeting. Gu Mojie went back to think about it behind closed doors, and called up the "Graphic Master" function that Hatsune Niang came with and checked it carefully.

On the Hatsune Musume body, the "graphics master" has powerful graphics processing and playback capabilities, far beyond what can be done on today's PCs. THE ACHIEVEMENTS THAT GOOGLE AND FACEBOOK HAVE JUST ACHIEVED ARE NOTHING IN GU MOJIE'S VIEW.

Not to mention searching for the web source, image name, and source file of an image based on the fragment of an image. Even if it's based on a picture. It's easy to search for video resources on the web that contain this image. With the current deep learning algorithm programmers in his hands, coupled with the friendship and outsourcing resources of Jeff Hinton and Yan Lei, as well as the imitation advantages brought by the later development, it will not be a problem to transplant.

However, Gu Mojie just couldn't find a channel to monetize this technology and collect money. If there is no profit, the technical cost of transplantation alone will make Gu Mojie feel deeply dragged down.

This kind of thing, you can't think of the result by thinking about it on your own.

TWO DAYS LATER, GU MOJIE CHANGED HIS MIND AND COMPILED THE RESULTS OF GOOGLE AND FACEBOOK IN THE FIELD OF IMAGE SEARCH USING DEPTH ALGORITHMS INTO AN INTRODUCTION DOCUMENT, WHICH WAS DISTRIBUTED TO THE COMPANY'S EXECUTIVES AND PARTNERS, SO THAT EVERYONE COULD BRAINSTORM AND SEE IF ANY INTERNET COMPANIES IN CHINA MIGHT BE ABLE TO USE IMAGE SEARCH-RELATED TECHNOLOGIES.

GU MOJIE RELEASED A SMALL RUMOR: WHAT GOOGLE AND FACEBOOK HAVE JUST DONE NOW, HE CAN ALSO DO. He is also the first batch of users of deep algorithms, and has this R&D resource, but no one in China pays for this technology.

Gu Mojie's intentions are similar to the cooperation with Google in San Francisco last summer: at that time, Gu Mojie was ready to meddle in cloud security, and the method was to set up a new offshore company, Gu Mojie outstanding technology, Google to pay for it, and then each of them took shares. Now, too, he hopes to get a domestic company that is willing to pay for a deep algorithm project in image search

When this rumor was released, Gu Mojie's first target customer was Baidu - in Gu Mojie's view, on the American side, Google took the lead in doing this, so Baidu, which is closely following Google in domestic affairs, will definitely follow suit. In addition, I heard that Baidu is seeking to list on the NASDAQ this year, and it should emphasize its technological leadership, which should be quite reliable.

For this reason, Gu Mojie specially contacted Liu Feitong, who had been friends with him for two years, but the result was not ideal.

Baidu was not interested in getting involved in image search engines in 2005. In other words, the main direction of Boss Li's funds is not here, and the urgency of today's domestic netizens for this demand is not enough, even if the search engine company develops this new feature, the additional traffic brought is not attractive enough.

On the phone, Liu Feitong repeatedly rejected Gu Mojie: "Mr. Gu, I know the sincerity of your cooperation. Actually, I really want to get involved in an image search engine project, but I can't really help you with that. The price on your side is too high, and Mr. Li can't accept the technical outsourcing of this offer. ”

Gu Mojie repeatedly emphasized how cost-effective his offer was: "Please tell Boss Li that I just want an investment of 600 million yuan, and we can set up a new company to run this project." Baidu paid 600 million, which can account for 60% of the shares, which means that the total valuation of the new company is only 1 billion.

In terms of the realizability of the technology, if you feel that it is not guaranteed, I can sign an additional VAM agreement and use the parent company of Hatsune Network Technology as a guarantee. Google has invested hundreds of millions of dollars in developing an image search engine. THE $200 MILLION THAT FACEBOOK RAISED IN THE FOURTH QUARTER OF LAST YEAR WAS ALSO MAINLY USED FOR 'FACE RECOGNITION TECHNOLOGY'. The epoch-making progress at the search engine level will cost at least hundreds of millions of dollars in development costs, which you know best, and my offer is not too much. ”

"I'm sorry, our financial statements this year are under a lot of pressure, in order to go public, for the sake of the price-earnings ratio, we can't make such a long-term investment."

Liu Feitong finally hung up Gu Mojie's phone. Gu Mojie also knew that there was no need to call Li Yanhong himself, and any company would not take a big risk on the eve of listing.

Just when Gu Mojie was disappointed, an unexpected customer came to the door.

A week before the Spring Festival, cousin Zhou Lixin drove to Gu Mojie's house to visit the door, Gu Mojie thought it was just an ordinary relative, but he didn't expect that there was still business hidden behind it.

The friendship between Zhou Lixin and Gu Mojie is naturally nothing polite. After putting down the gifts for his uncle-in-law and aunt-in-law, Zhou Lixin directly pulled Gu Mojie into the study, closed the door and asked:

"I heard that you are now peddling an image material search solution that imitates Google's 'image search engine' every day? In terms of price, as long as one-third of the R&D cost on Google's financial report? ”

"Almost, you can save me some face, if I want to do it, I must have my own innovation, and it is not purely a copycat Google. Google will definitely invest more than $300 million in developing this thing, and I just need someone to give me $100 million in cash. Of course, I still have a small stake in the new company. ”

Zhou Lixin cautiously confirmed some functional problems with Gu Mojie, mainly to understand what Gu Mojie might achieve in what way. After getting an affirmative answer, Zhou Lixin seemed to have made a lot of determination, patted Gu Mojie's shoulder and said:

"It seems that I'm helping you get buyers -- if you can make a different image material search solution than Google, someone will pay you $100 million."

Gu Mojie perked up, but he was a little puzzled, and asked, "Who?" Baidu refuses to ask for this thing, who else in China has such boldness to follow Google to do this kind of technology? ”

"Stupid! You don't want to think about who your brother is working for now. Of course, my boss might be interested. ”

"Boss Ma? Isn't Ma Feng doing e-commerce? He's not a search engine, so what do you want this thing for? ”

"Who told you that Boss Ma is an e-commerce business? Ahri Group is essentially a data company, but we are C2C, and we don't sell things personally.

Let's take a look at this plan -- I wrote it by combining your technical description with our Taobao demand analysis. I have submitted the draft and suggested that Boss Ma add a function called 'Find the same model' to Taobao, which requires the use of your technology. Boss Ma has also agreed, specifically, you can talk about it yourself in a few days, the brother can only help you get to this point. (To be continued.) )