In November 2013, Meng Fanqi's life became rare and peaceful.
For more than half a year since his rebirth, he has been busy preparing initial papers, experiments, and the just-concluded IMAGENET competition.
These are the core assets that will enable him to negotiate terms with technology giants when he goes to Australia to attend a conference in December.
To be fair, even for a born-again person, it is not easy to do these things well alone.
As for Don Juan, it is just a useless pendant, and its effect is better than nothing.
The better and more solid the preliminary work is, the higher the starting point for entering the industry will be. This will have a huge impact on future roads.
Now, with the announcement of the results of the competition and the publication of papers on generative adversarial networks, the hard work of the past six months has come to fruition.
When Meng Fanqi participated in the competition and published his paper on arxiv, he deliberately did not leave his email address because he wanted to take a good leisurely month.
DreamNet’s victory and the announcement of the Generative Adversarial Network, the impact of spiral superposition is indeed beyond Meng Fanqi’s imagination. This made him extremely grateful for his decision not to leave an email address.
Meng Fanqi's deliberately titillating way of publishing the paper did play a certain role, but the original dissemination speed and effect were far from as good as they are now.
A real catalyst was Bengio’s promotion on social media.
"Ian came to me early today and said that he had thought of a brilliant idea. I chatted with him for a long time and agreed that this was a very genius idea. Unfortunately, this work has just been completed by someone else. And it can It is said that they have done a perfect job. They have released a complete paper and experimental results (paper link). The results are very amazing, but at the same time it also makes me a little worried. While AI capabilities are improving rapidly, how to distinguish the authenticity of information becomes more and more difficult. It has become an issue worthy of attention."
In fact, Bengio did not deliberately intend to publicize it. The issue he is most concerned about in this matter is still AI ethics and the issue of distinguishing true from false. He hopes that AI technology can benefit mankind, but at the same time it needs to be controlled to a certain extent and not cause a lot of trouble to mankind.
But as the three founders of the genre, the papers named by him will naturally attract the attention of a large number of cutting-edge scholars, as well as a large number of related media waiting to be used as material.
And these scholars who have discovered the value and advancement of this paper, and the good media will continue to recommend and promote each other.
Bengio’s tweet has greatly accelerated the process of generative confrontation technology attracting academic attention.
But after the first round of popularity subsided, everyone looked at each other in confusion, wondering who the author Meng was.
Now the only person who knows Meng Fanqi's contact information is Alex, who has joined the Google team.
Alex had already written an email to ask about Meng Fanqi earlier. Unfortunately, Meng Fanqi was Chinese and was in China. If he didn't take the initiative to contact others by email to ask about something, he really didn't see much of it or pay much attention to it, so he would Never saw it.
Alex didn't pay much attention at first. There were still thirty or forty days before Meng went to the Australian venue in early December and appeared in front of the world.
His purpose was just to invite people for Google and just select a time before the conference to have an interview. This matter won't take long.
But the huge response caused by the generative adversarial network made him suddenly nervous.
First of all, this means that Meng's level and achievements are still far beyond their imagination. As someone who took a similar path last year, Alex thought he was the person least likely to underestimate the value of the opponent, but he never expected that he would underestimate him after all.
Secondly, a very fatal point is that although Meng seems to be deliberately hiding his contact information. But the professor Fu who collaborated with him this time is a well-known figure in mathematics in a Chinese elementary school.
Alex searched for Dean Fu's information, and his sense of crisis suddenly rose.
Through this collaborator, it is likely that others have now obtained the contact details as well.
Even Chinese technology giants like Baidu can already go to this university to interview him.
Alex recalled the fierce scene of Baidu and Google bidding for his team last year, and suddenly felt a little dazed.
The offensive and defensive dynamics are different.
Thinking of this, he didn't want to wait until Meng Fanqi responded to his last email to confirm whether he was the person he was looking for.
Don't think about the reason why the other party didn't reply to your email.
He found Hinton, his mentor, and Jeff, his main talker. We quickly finalized the general scope of some specific conditions and directly prepared a letter of intent.
In this way, once the other party responds, he can directly send the detailed conditions to Meng Fanqi's mailbox.
It contains full Google sincerity, from the researcher's position, nature, working methods, specific salary and benefits, to the settlement of visa and green card, residence, moving expenses, equity, signature fees and other miscellaneous matters.
Everything was carefully considered and multiple alternatives were provided for him to choose from.
Finally, an invitation.
The letter stated that Google hopes to have an interview as soon as possible, either in Mountain View, Silicon Valley, California, or Google people can go to Yanjing, or in Sydney, Australia, which is the final exchange venue of IMAGENET this year.
Later in the evening, Meng Fanqi finally remembered to check his mailbox and felt very surprised.
Alex responded very proactively and solved his problem.
As a result, when he received the letter, I had not checked my mailbox for almost a week. Meng Fanqi blushed, even he felt very embarrassed.
After reading the specific time of the email, Meng Fanqi probably guessed what was going on.
Maybe it was because I was impatient to wait, or maybe it was because the paper I published caused quite a stir.
The second email Alex sent was no longer in his own name, but on behalf of Google, expressing his interest and invitation.
Meng Fanqi replied briefly and formally got to know each other. Not long after that, Alex responded directly to an email with an attachment.
This email gave myself a preliminary letter of intent. This is not a very formal letter of intent, more like an introduction or even a menu.
He does not just have to passively choose whether to accept it or not. It provides many options for him to consider first and then finalize it during the interview.
This was the most nervous and exciting moment for Meng Fanqi since his rebirth. Everything he did before was in exchange for larger and larger numbers in the letters of intent from these technology giants.
When he received this email, Meng Fanqi understood that the most risky stage had passed. He would obtain far more resources than he did now, and the road ahead would only become smoother and wider.
The moment he clicked on the file, he had a vague feeling that his time and his life were starting to accelerate from this moment on.
The familiar environment, life and rhythm of life are no longer there. A brand new door opens at this moment, emitting extremely dazzling golden light.
