The renminbi in 04 has not yet begun to enter the appreciation channel. 1 million US dollars, the actual amount is 7.5 million yuan.

Such a large sum of money is only the buyer's initiative to offer, without the coaxing of the bargaining stage. And what this offer wants in exchange for is just Gu Mojie's co-signature on a string of papers, no matter how you look at it, it's a little too incredible.

With this incredulity in mind, Gu Mojie took more attention to read the review that Professor Jeff Hinton will publish next month.

Some of the content Gu Mojie had already dabbled in in the scattered reading of the previous week, but at the moment he just felt more integrated. There are also some contents that are unexpected, which is quite deafening.

After watching it for two hours, a strange countercurrent of thinking rose in Gu Mojie's mind, just like a martial arts master who practiced supreme internal skills, struggling on the verge of going mad before opening up the second vein of Ren Du.

Maybe it's a bit too fantasy to say this, but Gu Mojie is no stranger to it, because after he was reborn, he had experienced this kind of experience once.

It was a year ago, when he flew to Beijing for the first time to negotiate Baidu keyword bidding ranking for the "universal input method" he wrote at that time. Because of a coincidence, I got the interest of Boss Li of Baidu and interviewed Boss Li for half an afternoon.

It was in that conversation with Baidu Boss Li that Gu Mojie was awakened by Boss Li's words, and a cognition suddenly popped up in his mind: in the next year or two, there may be a major reshuffle in the domestic antivirus software market, and new products with a completely free breakthrough model may sweep this market.

After returning from Baidu, Gu Mojie confirmed the meaning of this situation through communication with Hatsune Niang: this is a reconstruction of quantum memory fragments. In other words, it was originally at the moment of Gu Mojie's rebirth, because some "future history" was destroyed by the butterfly effect. It was no longer inevitable, so that part of the memory was erased when he was reborn. But once all the conditions necessary for the occurrence of these histories are met again, and these things become inevitable and certain in the future, then it is possible that this part of the memory will be re-integrated.

At this moment, the stirring feeling in Gu Mojie's mind is the same as the one he had in Baidu a year ago. It's just that the medicine primer that time was Boss Li's head stick to wake up. This time, the introduction to medicine is Professor Jeff Hinton's paper.

After the turbulent waves in Gu Mojie's mind subsided, he only felt sweating profusely, but there were three words left in his mind: cloud computing.

Looking back at Jeff Hinton's paper, Gu Mojie suddenly felt that he was much more familiar with it. Many of the ideas that had been learned in previous lives were regrouped from the fragmented state. As for the prospect that Jeff Hinton tried to make, Gu Mojie knew even more in his heart than Jeff Hinton himself.

Gu felt that he could sum up in a few words the future direction of artificial intelligence envisioned by Jeff Hinton.

Before the advent of Jeff Hinton's theory, mankind had always believed in the means of realizing artificial intelligence in the future. It will be the kind of model that Alan Turing proposed more than 50 years ago -- the computers built by humans are getting more and more powerful, and finally a super terminal has been developed, which can understand all human language, perceive the same keen sense of sight and hearing as humans, and can "understand" the meaning of pictures and musical notes as humans do.

This misguided belief came to a head in 1997 when IBM created the "Deep Blue" and defeated the then world chess champion Kasparov with it: Look, computers can really think. Will it be far from building a machine with artificial intelligence?

In the midst of the optimism, Jeff Hinton happened to be the one who kept his composure. He envisions the opposite of the problem.

Jeff Hinton has always felt that although it is possible for the superterminal to eventually "be able to truly understand everything" like humans, it is too far away, and the current level of hardware will only be possible to reach that point for at least a few more decades.

Before that. Human beings should have a better shortcut, by connecting countless less advanced single computer terminals with a network, so that the mistakes made by each node can be learned from other nodes, so that the self-learning and evolution of computers is not limited to summarizing the mistakes they have made. Rather, it extends to "summarizing the mistakes of all the same kind." Then, it is possible for human beings to enjoy the benefits of artificial intelligence decades in advance, when the hardware is not particularly powerful.

The network topology model mentioned in this scenario was originally called "convolutional neural network", and in order to realize the continuous optimization learning of "convolutional neural network", a series of internetwork feedback algorithms are required to process the results, and the series of algorithm sets are called "deep learning algorithms".

Jeff Hinton is a superstar who is in the top five academic positions of all mankind in the field of "convolutional neural networks", and in the specific research direction of "deep learning algorithms", he is the world's first.

It may be a bit boring to describe it like this. But later generations, if they look back from the results, they can see how great this guiding ideology is:

Since Jeff Hinton summed up the complete theory of "deep learning algorithms", within only one year, Google has successfully developed a "picture search engine", so that the search engine is no longer limited to searching for text in the link, but can search for an image - because Google's search engine uses the idea of deep learning algorithms, so that computers can also learn "how to read the meaning of a picture".

SIMILARLY, FACEBOOK, WHICH WAS STILL IN ITS FORMATIVE YEARS AT THE TIME, INVENTED FACIAL RECOGNITION TECHNOLOGY BY BORROWING FROM JEFF HINTON'S IDEAS. FROM NOW ON, PHOTOS UPLOADED ON FACEBOOK SPACE WILL POP UP WITH THE PERSON'S NAME ONCE THE READER CLICKS ON THE FACE. AND AS LONG AS THE USER DEFINES THE NAME OF A CERTAIN FACE ON A CERTAIN PHOTO, THEN IN OTHER PHOTOS UPLOADED BY THE SAME USER, AS LONG AS THE SAME PERSON APPEARS, EVEN IF THE MAKEUP AND EXPRESSION ARE SLIGHTLY DIFFERENT, FACEBOOK CAN RECOGNIZE IT.

TWO OR THREE YEARS LATER, THIS TECHNOLOGY WAS INTRODUCED TO CHINA, BUT CHINESE PEOPLE COULDN'T USE FACEBOOK, SO WHEN MOST PEOPLE FIRST SAW THIS AMAZING TECHNOLOGY, THEY SAW IT IN QQ SPACE, SCHOOL INTRANET, RENREN AND OTHER PLACES.

Sogou company, because with the help of the guiding ideology of this algorithm. Invented the "Sogou input method" that can evolve itself...... However, in this time and space, there is obviously nothing to do with Sogou, because Gu Mojie has already taken the lead.

All the scientific and technological innovations in the fields of "fuzzy processing" and "machine deep understanding" are the initial sparks that come from Jeff Hinton's academic theory.

And these things have been brewing and growing for several years in later generations. The result was a trendy concept called cloud computing.

……

Gu Mojie naturally doesn't remember those specific examples of later generations, but the essence of thought that has been refined has never been so clear and clear.

After thoroughly understanding all this, Gu Mojie finally realized why Jeff Hinton sent his right-hand man and protégé to China at any cost, begging him to co-sign the future string of papers.

In the history of the original time and space, these reviews and prospects of Jeff Hinton. It wasn't proposed until October 2004 because Jeff Hinton himself hadn't been able to figure it out too much, and many of the ideas lacked practical support.

IT WASN'T UNTIL 2006 THAT JEFF HINTON ACHIEVED HIS ACADEMIC SUCCESS AND WON IMMORTAL STATUS -- IT WAS IN THAT YEAR THAT AFTER JEFF SINTON'S ACADEMIC SUCCESS, GOOGLE AND FACEBOOK BECAME THE FIRST APPLICATORS AND BENEFICIARIES OF THIS THEORY ALMOST INSTANTLY.

Now, Gu Mojie's "Hatsune Input Method" and "Hatsune Translation". But it became the touchstone that inspired Jeff Hinton. Jeff Hinton put forward his theory more than half a year in advance, and collected enough material to prove his argument two years in advance.

"Hatsune Input Method" is only a trivial piece of software. However, Hatsune Input Method already has a series of key nodes that need to be emphasized by "deep learning algorithms" such as "being able to use convolutional neural networks, learn from the error information of other neural nodes, and efficiently self-correct".

If it's just a "Hatsune input method", then it can also be explained as a coincidence.

But Gu Mojie just launched a "Hatsune Translation" at the same time, which also has the key nodes to reflect those thoughts. That's not just a coincidence, it can only be explained that Gu Mojie is also knowledgeable.

So. Unless Jeff Hinton no longer engages in research in the field of deep learning algorithms, he can only woo Gu Mojie, let Gu Mojie declare that the two are a cooperative relationship, and let Gu Mojie appear as an "industrial practitioner under the guidance of Jeff Hinton's thoughts".

There is no way around this. Otherwise, Jeff Hinton's academic standing in history textbooks will be tarnished in the future. Leave regretful blemishes.

Jeff Hinton was willing to pay $1 million, not for profit, but for fame. Jeff Hinton may be like those Silicon Valley madmen, who treat money like dung, but he can't escape the chase for fame.

What he wants is to be immortalized in the history books.

Gu Mojie and Jeff Hinton, one seeking profit and the other seeking fame, seem to have a strong complementarity of cooperation. The question at hand is just to see how much the name Jeff Hinton wants is worth and what conditions Gu Mojie can squeeze out of him.

Of course, Gu Mojie also knows the truth of stopping in moderation and accepting it when it is good.

After all, the other party is just a true scholar engaged in basic scientific research. Neither bureaucrats nor profiteers. If the price is too high and others can't get it, it's not interesting. For true scholars who do not rely on their status in the rivers and lakes and rely on real talent and real learning to win their reputation, Gu Mojie still respects him in his heart.

After thinking about the cause and effect, Gu Mojie looked at the time, it was seven o'clock in the evening, and there was still some work for him to do off-site work. So he made two dispositions at once.

First, he called Sheng Wei and asked Sheng Wei to help him hack the low-secret peripheral network of CIFAR (Canadian Advanced Project Institute) and find some relevant information that was enough to prove Jeff Hinton's previous status and the funds that could be obtained through various channels every year.

Only knowing the latest trend of Jeff Hinton's worth, Gu Mojie can make a targeted offer. Otherwise, directly quote a high price that a person's family can't earn for three or five years, and things will only collapse.

The second thing is that Gu Mojie contacted Fei Liluo and said that he was going to visit Tiance Law Firm and asked Fei Liluo's aunt Fei Wenli to help him introduce a lawyer who is proficient in the field of international business law to help him handle the specific negotiation matters.

Sheng Wei was very cheerful, and said that he would let Gu Mojie wait for two hours, and he would definitely send the latest information before ten o'clock in the evening.

As for Felilo, she directly asked him to go to the law firm, saying that her aunt herself was a famous expert in this area, and there was no need to find someone else.

Gu Mojie got dressed and drove straight to the law firm. (To be continued.) )