In early January, Vision Vision Technology Co., Ltd.’s first product, the AI camera, was officially launched and put into use.
Time assassins, fake smile emoticons and the like quickly became popular. It belongs to tap water and the heat is not low.
Meng Fanqi did not invest any publicity funds in this matter. He only made a deal with Robin Li in exchange for the recommended resources of 91 Mobile Assistant, which is currently in full swing, and Robin Li’s subsequent connections in government projects.
As a mobile application platform with tens of billions of downloads at that time, the support of 91 made the start of AI cameras a smooth journey.
The price Meng Fanqi paid for this was a promise to confirm a financing with Baidu within this month.
Robin Li doesn't care how many shares he wants to control in Vision Vision Technology, or how much profit he wants to make from it.
From his perspective, it is just a matter of holding equity in Meng Fanqi's company to build a relationship between the two parties. In his opinion, this is likely to be of great benefit to Baidu's AI technology in the future.
For this reason, he would rather exchange Baidu's shares for the shares of Meng Fanqi's start-up company.
After looking at Meng Fanqi's app, Li Yanhong found that the so-called camera is just a carrier that provides users with the option of taking photos without changing the app.
The real point is not this, but this:
This function can rejuvenate the people in the photo and make them grow old overnight.
This feature makes the person in the photo cry with joy and burst into laughter.
The AI algorithm hidden behind the App shocked the knowledgeable Li Yanhong.
At this time, it was actually only more than two months since Meng Fanqi released the generative adversarial algorithm.
During this period, Baidu implemented the YOLO real-time detection algorithm and cooperated with the government. The progress of this matter has been quite smooth, it can be said to be rapid.
But Meng Fanqi directly transformed the generative model from a groundbreaking idea into a powerful and controllable editing tool.
Edit the specific properties of any face directly in any image.
Age and expression are only very preliminary attempts, just the simplest among many possibilities. It can be said to be a drop in the bucket.
If Meng Fanqi has found a way to modify certain attributes of images through artificial intelligence algorithms, then there are too many things he can do.
Intelligent coloring of images, changing the clothing and posture of characters, making up for missing parts, and adjusting the overall style of the picture.
Or even generated directly from scratch.
There are so many possibilities, and it doesn't take long to do it if you're willing to put in the effort.
Since meeting Meng Fanqi, Robin Li has made progress in the field of artificial intelligence, but at the same time he has felt deeply powerless.
Baidu invests billions in artificial intelligence every year, and its achievements are not at the same level as Meng Fanqi's progress in the past few months. He can only catch up and study Meng Fanqi's papers and codes every time.
Just a few days ago, while the government was successfully advancing the Skynet project, the government had already proposed the technologies it might need in the next step.
That is the face recognition system based on a huge population base.
In the government's plan, there is a mid- to long-term demand, or vision.
That is, we hope to be able to directly associate the observed people with the residents’ identity information through simple images/videos.
For example, in places like customs, facial recognition + big data can allow police officers to intercept key personnel while allowing ordinary tourists to pass customs smoothly without any feeling.
Or in a department such as the public security department, facial photos of some key figures can be retrieved relatively quickly from the database.
Robin Li has just started thinking about this matter, and Meng Fanqi's new company has already made this camera app that obviously has some show-off qualities.
When it comes to human faces, editing various attributes is easy and effortless.
In less than a week after the app was released, it had over 10 million downloads.
Although it relies heavily on the 91 Mobile Assistant platform and promotional resources, judging from user retention and active behavior, the quality of the App itself is quite excellent.
The number of users that exceeds 10 million can only be said to be tepid. To become a phenomenon-level app, it needs tens of millions, or even over 100 million users to be called it.
But even if it is not very popular, the initial server must have been insufficient for a long time.
Fortunately, as Gao Wenhe said, Google settled part of the revenue at the end of 2013, and Meng Fanqi quickly followed up with a batch of equipment.
Even so, Meng Fanqi had to greatly reduce the free quota and start a paid usage model.
After all, making an app is not charity. With those graphics cards running all the time, the cost is not low.
Meng Fanqi's prices are not high and very conscientious.
Little things add up, and gradually I have an income of millions.
Not long after joining the entrepreneurial team, there were products and revenue. Xia Yan and several other students who were still studying were very happy.
Although this amount of money is not enough for the boss to pay back the original equipment, let alone the purchase cost of the venue, this is just the beginning! The better days are yet to come!
Wang Kai, who was still going through the handover procedures before resigning, burst into tears. Meng Fanqi clearly said it would be two months later, so why did he act so quickly?
If this continues, he won't even be able to drink hot soup.
After the number of App users exceeded 10 million, the growth rate slowed down a lot. This situation was worse than expected.
But originally it wasn’t the main course.
In mid-January, Google passed an internal review and released Meng Fanqi’s FaceGAN face generation technology paper and the [None of These People Exist] website.
The repercussions caused by this incident far exceeded Meng Fanqi's expectations. Within 24 hours, the situation was completely out of Meng Fanqi's control, and even Google's control.
More than half a year after Meng Fanqi's rebirth, he has released a number of world-class technological breakthroughs, but the popularity in the United States has remained within the circle and cannot be called a social event.
But this time, for some reason, well-known media in the United States responded to this technology within 24 hours.
These include the Wall Street Journal, New York Times, Forbes, CNN and many other big-name and well-known media.
Meng Fanqi quickly browsed these media reports, feeling a huge wave of emotions in his heart.
"Why are generative technologies dangerous?" 》
"Generative technologies will cause huge disruptions to society, and we are totally unprepared. 》
"What is generative technology and why is it dangerous?" 》
"Generative Technology: Security Crisis"
"I'm super? Why is the reaction so intense???"
In Meng Fanqi's view, the current face generation technology is far inferior to the visual technology that can be used on the whole body, such as face replacement or Fake series.
After all, he underestimated how advanced technology could shock today's human beings.
As the first author, he has now been @exploded on Twitter.
Journalists, scholars, common people, and ignorant incited people.
At this moment, Meng Fanqi is at the center of the storm of public opinion and the focus of discussion.
The professional field of AI/artificial intelligence, under Meng Fanqi's repeated promotion, has finally led to qualitative changes from quantitative changes, and has become a social topic.
