What's wrong with me being a rich man?

Chapter 459 Empowerment
The disappearance of Weishi was expected.

With external competition and internal constraints, from the perspective of the Weishi team, it was indeed a strategic abandonment. However, from the perspective of Tencent as a whole, as the frequency of short video usage in WeChat Moments increases, it is only natural to focus resources.

There is no right or wrong, only perspective. Despite touting daily active users during the Spring Festival, once the marketing campaign ended, without solid content support, Weishi's daily active users continued to decline, and with limited resources, it ultimately failed to survive even the first half of the year.

Yu Xing wasn't sure how long the original Weishi platform would last, but he felt it was inevitable sooner or later.

He sighed slightly, hoping to provide a warm haven for the frustrated core members of Weishi, but he did not comment on the matter. Tencent must also feel that the merger of the previously publicized team was not so glorious, and did not disclose the change to the outside world at all.

However, Yu Xing soon noticed Zhong Zhiling's insincere comments on his WeChat Moments.

"What a pity. Tencent responded quickly. I originally thought that Alibaba's WeChat had a chance to break through the homogeneous competition by using short videos that are more user-friendly. I didn't expect Tencent to make such a decisive decision. What a pity."

After meeting, Yu Xing asked his junior brother, "Why are you pretending when you have nothing better to do? Mus will be exposed by next year at the latest, and people will say I'm hypocritical then."

With Mus exposed, Silicon Carbon Data is entering the domestic market, and the evaluation of Tencent's short videos, which began on a program last year, will inevitably complete its closed loop.

Zhong Zhiling just laughed: "Brother Xing, I was just saying it casually on my WeChat Moments, I didn't mention you at all."

Yu Xing shook his head: "What's the difference between you telling me and me? It all comes down to me."

Zhong Zhiling thought about it and realized it was true.

He immediately became righteous and said, "It's on your head, so be it. If you've got the guts, Mr. Liu, let him come and make you cry at Silicon Carbide!"

Yu Xing sighed, "I've noticed that you all's skin has really gotten thicker. Dr. Yu Kai is also getting quite skilled."

Zhong Zhiling didn't answer this time, but just looked at his senior brother and smiled.

Yu Xing was a beat slow to realize what his junior brother meant: birds of a feather flock together...

After a silent rebuttal, Zhong Zhiling asked about the recent commotion at Silicon Carbon Data Company: "Brother Xing, how's the situation with those scientists going?"

Carbon Silicon Data Company is, after all, one of the shareholders in the preparation. Although it was at the request of his senior, as the CEO, he also wanted to know a little about the progress.

“Not bad, it’s rebounded from the bottom,” Yu Xing said. “Hillhouse Capital and Five Resources are both willing to invest. They are also very optimistic about the future of electric vehicles and want to actively participate in the next round of funding in Lingang. They can put up $1500 million for the foundation, we have $1000 million, and the city will see if we can raise $500 million, which will be $3000 million. Red Falcon has been raising funds recently and may also contribute some.”

Zhong Zhiling, anticipating the final outcome, asked curiously, "Brother Xing, if we only raise half of the target amount, will the foundation continue?"

He knew Yu Kai's goal was $1 million, and given the current situation, offering $5000 million was already quite good.

Yu Xing pondered for a moment and said, "It might work, but nothing is perfect. We all just make do."

He actually thought that such a large-scale operation would only cost around 5000 million. Even if the Red Falcon could contribute 1000 million, these scientists probably wouldn't be able to raise 1000 million US dollars, and other institutions would have to make up the difference.

If it can get running smoothly this year, the funds in Lingang will change every year, and things will definitely get better and better.

In late April, the excitement in both circles in Lingang was coming to an end. Investors were happy to learn about cutting-edge knowledge and were willing to meet with Mr. Yu, but they still had to wait and see before making any money. However, a serious disagreement arose at the end of the "Machine Learning and General Artificial Intelligence Conference," which led Yu Kai to bring over the financial backer, Yu Xing.

Whether the LDF Foundation will operate in an open-source or closed-source manner in the future is an important question.

Yu Xing didn't meet with these scientists much because he genuinely couldn't understand their discussions, but the last question... that was definitely something to talk about.

"I support open source."

"When Dr. Yu first approached me about this, he said he wanted to join me in this great undertaking. I didn't take it seriously. What great undertaking? He just didn't want to work at Baidu anymore and wanted to start his own business with people."

"As a result, your arrival surprised me. I didn't expect that a group of people from Microsoft, Facebook, and MIT would be willing to return to China to promote the development of machine learning. Later, after listening to Dr. Yu talk about his past work at the NEC lab in Silicon Valley, I felt that... this spirit of promoting technological development from beginning to end is the root of our ability to come together."

"Machine learning definitely has a future. Microsoft is doing it, Google is doing it, Facebook is doing it, and that's doing the right thing!"

"However, it's not enough to just do the 'right thing'; it's also crucial to know how to do things correctly."

"The best example of open source is in smartphones. Android has grown bigger because it is willing to open source. The same is true for the DLF Foundation. If you want to attract more support, you can't be stingy with your own ideas, and it's far from the time to be stingy with your own ideas."

"We want to use open source to attract the global power of machine learning, and also to cultivate more power in this way."

"Lingang will firmly support the DLF. We have already contributed money this year and will continue to do so next year!"

Yu Xing began by stating his position and concluded by making a commitment.

As a member of the "Morning Enlightenment" committee and a major financial backer, his statement carried considerable weight, ultimately leading to DLF's decision to operate in an open-source manner.

Of course, DLF is classified as a non-profit organization, but in order to maintain its operation, it will also have a commercial entity. The technological achievements of the latter cannot be open source.

Yu Xing doesn't need to know what kind of changes the world's top machine learning involves; he only knows that the more people use machine learning tools, the better, like the DMLC community.

With the DLF finalizing the last point, the meeting came to a close, and the participants gradually returned to their respective regions and positions.

On April 25, Yu Xing and others who were attending the Shanghai Machine Learning and General Artificial Intelligence Conference for the first time took a group photo, all hoping that this field could be developed and nurtured in China.

However, the committee members stayed behind for the time being, and Yu Kai and Chen Tianqi, as representatives, accepted interviews with Shanghai media, praising the significance of the meeting and taking the opportunity to secure 3000 million RMB in investment support from Shanghai State-owned Investment Corporation.

This amount of money is not much compared to Yu Kai's fundraising goal, but it is still a significant support for the city.

After seeing off his academic colleagues, Yu Kai expressed mixed feelings about the trip to Lingang for the exchange.

The good news is that Horizon Robotics also secured $1000 million in angel investment from three institutions, allowing him to start his own business.

The worrying thing is that DLF has raised less than half of the funds, and the next step may really depend on Mr. Yu's judgment.

"Mr. Yu, thank you so much. If machine learning in China can make great progress, you will be the one to make the greatest contribution." Yu Kai ate in the canteen of the Silicon Carbon Group and thanked Yu Xing repeatedly. "Horizon Robotics also has to thank you. Without your emphasis on Hillhouse Capital, they would not have been able to make a decision so quickly."

Although there are supposedly free gifts, there are still considerations before deciding to invest money.

After discussing their startup plans with Yu Kai, Hillhouse Capital, 5G, and Yida Capital, the three investment firms that invested in Horizon Robotics, also sought advice from Yu Xing on this hardware-side project that claims to focus on intelligent driving mobility solutions. Yu Xing didn't repeat what Yu Kai had already presented; instead, he discussed Horizon Robotics' future from the perspective of Silicon Carbon Group's needs.

"The chips we are currently using are from Israel's Mobileye. If Horizon Robotics succeeds, we will definitely switch to Horizon Robotics in the future."

"Not only us, but all domestic manufacturers that will be involved in new energy vehicles in the future will have this need. When we communicated with Mobileye, we hoped that they could make some improvements and optimizations to the technology for domestic road conditions, such as road cones, but they refused any localization modifications."

“These international suppliers all have similar problems. They generally believe that the products they develop for the international market are the best in China and there are no substitutes, so they let us use them or not.”

"It's not that I'm trying to act impulsively, but I firmly believe that companies that can't provide good localization services will definitely struggle to offer sufficient competitiveness in the future."

When questioned by the three organizations, Yu Xing was very resolute.

It was after this meeting that they finalized the angel round for Horizon Robotics.

Yu Kai hadn't heard President Yu mention this matter, but he learned about it from Hillhouse Capital, and he couldn't help but feel grateful.

Yu Xing readily accepted the thanks: "The greatest gratitude is that Horizon Robotics can do well. It's actually a win-win situation. We need better suppliers to provide better products. Even if they are already international Tier 1 suppliers, their complacency is an opportunity for innovators like you."

"The Silicon Carbon Group will use stronger products to create even stronger products."

The head of the silicon carbon group didn't seem too excited when he said this, but Yu Kai, who had already entered the entrepreneurial mindset, was much more excited.

In his excitement, he crossed the line: "Mr. Yu! I believe in you! If you think it's worth trying cryptocurrency with the funds raised by DLF, then go all in!"

“…” Yu Xing believes that people still need to have a bottom line. “Dr. Yu, Alibaba is very reliable, let’s go with Alibaba. DLF is registered in Hong Kong, and it’s convenient to use the funds. Hmm, the city’s investment may not be able to be used like this. Let’s think about it again later.”

He planned to let Xiaoying take a look at the process after she finished her work.

Liu Wanying now has extensive experience in financial operations. In her own words, even if she doesn't short sell, she's a master at money laundering.

She has been busy with compliance procedures at Peak Capital this month, planning to use up a batch of shell companies from offshore firms all at once.

European financial markets are not as conservative as Japan's, but they are not as open as the United States' either. The target this time, Renault, has a 19.73% stake held by the French government, a 15% stake held by Nissan, and about 50% held by long-term investors, such as BNP Paribas' funds, which hold 2% of the shares. Therefore, short-selling operations that go over the peak need to be more cautious.

Yu Xing had already read Shanfeng's short-selling report on Renault and was just waiting for her final move.

On April 28, in the president's office of the Silicon Carbon Group, the latest documents of the day were being reviewed. Yu Xingshun was also taking a look at the implied volatility of Reynolds options when Vice President Cui Zhiyu knocked on the door and came in, expressing a question he didn't understand.

Cui Zhiyu, who came from Great Wall Motors, is responsible for many of the company's daily operations, including coordinating the work of teams working on vehicle systems, software, and other areas.

He received a protest from the voice team yesterday, who felt it was absurd to overturn all the results of past work. So, after learning about the situation, he came to petition today.

"Mr. Yu, are you aware that the car's voice system needs to be completely overhauled?" Cui Zhiyu asked bluntly.

Yu Xing closed the webpage on his computer and casually replied, "I know, I heard from next door that they think the model we're using is outdated, you know, that one..."

He thought for a moment, then looked up and asked, "Is it called Gaussian mixture model?"

“Yes, it’s a hybrid architecture of Gaussian mixture model and hidden Markov model,” Cui Zhiyu said seriously. “This is the one that everyone is using, and the company has been working on it for almost half a year. How can we just tear it down like that?”

He listened attentively to the voice team's complaints and opinions yesterday, and also took note of the core issues.

The core problem is that the core is gone...

Yu Xing pondered for a moment: "These are the opinions of the scientists who came to Hong Kong for a meeting earlier, and also from Dr. Zhu from MIT. He and his team are helping with it, saying they're switching to a more intelligent architecture, called..."

He reached into the files and flipped through them, finding the ideas for a small assignment from next door. He read out the core architecture: "It's called a Deep Neural Network (DNN), which is a Deep Belief Network (DBN) and a Feedforward Neural Network, capable of reconstructing acoustic models."

Cui Zhiyu said with dissatisfaction, "President Yu, this is too hasty! Although they are experts, they haven't thoroughly understood our stuff, and they've messed up the work like this. I don't agree with it!"

Yu Xing put down the documents and said, "It wasn't easy to get them to help; it cost us $1000 million."

Upon hearing this, Cui Zhiyu was taken aback: "1000 million US dollars? Just for this job? This..."

He was heartbroken: "This money shouldn't have been spent!"

Cui Zhiyu was practically ready to yell at President Yu for being a spendthrift!

Yu Xing calmly said, "It wasn't Silicon Carbon Group that provided the money; it was Silicon Carbon Data from next door."

Cui Zhiyu calmed down and thought for a moment, then said, "On the other hand, some money should be spent."

A thought flashed through his mind: the experts he'd hired for $1000 million, this so-called DNN architecture, this so-called reconstructed acoustic model—perhaps it wasn't what he imagined…

Cui Zhiyu respected the efforts of his neighboring unit and decided to withdraw his petition and observe the situation for the time being.

However, he did not expect the observation period to be so short.

Just three days later, Ren Yixuan, the head of the speech team, came to visit again, full of admiration for the effect of the latest acoustic model. The DNN directly learns the probability mapping from speech features to phonemes through multi-layer nonlinear transformations, replacing the probability density estimation of GMM. Combined with the Hidden Markov Model, it shows the potential to approach the level of professional human transcription.

Ren Yixuan reflected, "President Cui, I didn't expect their model to be so good. The speed at which they completed it... was too fast! And it was too good! We couldn't accept the speed at first..."

Cui Zhiyu felt it was expected and said sincerely, "As long as it works, it doesn't matter how much money you get or how much they get... This is a dimensional reduction attack from top scientific research, so it's not your fault. Follow the direction of this new model and strive to produce results as soon as possible. Anyway, the bonus will be paid to you."

Ren Yixuan breathed a sigh of relief. Those people only took a week, which seemed to make his team look too useless.

Fortunately... the other side also had a reference point. It is said that Microsoft's technical path in this area is similar, thus allowing them to be outmaneuvered by a combined attack from top global experts and international giants...

Well done!

(End of this chapter)