The launch of Orange Technology’s new phone, together with the developer conference, is a real bombshell.

When all the data are summarized, the numbers are staggering.

The Qingshan Lake factory had to work overtime to ensure supply.

caution!
This is not hunger marketing.

Unlike Huawei, Orange’s mobile phones are still popular among users overseas even after being banned by the entire GMS package.

This is inseparable from years of preparation for sanctions, which in turn have brought greater visibility to Orange.

Seeing that foldable screen mobile phones are so popular, friendly competitors have to start research and development immediately to prevent this huge piece of cake from being divided up.

In the same year, Orange Technology’s three flagship models sold 40 million units at home and abroad.

Quantity alone doesn’t mean anything.

But don’t forget that these are almost all high-end models, rather than a strategy of launching a large number of models like Samsung.

It can be said that after this battle, Orange Technology’s high-end strategy has truly been successful.

It has the strength to compete with Apple both at home and abroad.

Right now, the only big test left for Orange Mobile is the subsequent 5G mobile phone.

What really impresses people is the indomitable tenacity demonstrated by Orange Mobile. Even when sanctioned by the most powerful country in the universe, it can still grow against all odds.

……

……

Time quickly came to 20.

What is the hottest keyword this year?
It’s artificial intelligence.

OpenAI on the other side of the ocean has released the first generative pre-trained transformer model code-named “ChatGPT” since 2018. Although this model seemed to be a huge breakthrough at the time, its capabilities were still very limited.

In 19, OpenAI released an upgraded version of GPT, ChatGPT-2. This model has 1.5 million parameters, more than 10 times that of its predecessor. ChatGPT-2 has greatly improved its ability in text generation. It can generate more fluent and coherent text and understand the context to a certain extent.

This year, OpenAI released ChatGPT-3, a giant model with 1750 billion parameters, which surpasses ChatGPT-2 in multiple dimensions. ChatGPT-3 not only excels in text generation, but also demonstrates amazing capabilities in novel creation, poetry writing, code generation and other fields. The release of ChatGPT-3 has made the names of OpenAI and ChatGPT known worldwide.

The whole world has paid attention to this company and new term that suddenly emerged.

Many people in the country are even more panic-stricken.

“This is almost one of the four great inventions of the new era, a great creation comparable to computers.”

“Foreigners still know how to play. The creativity of 0-1 is not fake.”

“In comparison, what are our entrepreneurs doing? Artificial intelligence. If we can’t seize this opportunity, how can we talk about surpassing the national strength?”

“It’s over. It feels like there are no comparable companies in our country?”

Some slightly more rational industry insiders are also expressing their views:
“In fact, China and foreign countries are on two different tracks. We are strong in applications, such as industrial applications. The Pangu model is very powerful now, but people just can’t feel it.”

“There is also a company in China that is working on general intelligence in Hangzhou, but it is currently keeping a low profile.”

The domestic public opinion circle has long been controlled by the “Reflection Party” and the “Winning Party”.

But neither of them is a particularly normal or good phenomenon.

……

……

Inside the office.

Liang Wenfeng, who was sitting opposite Chen Pingjiang, was eager to try.

“Mr. Chen, our product is not weaker than theirs, and we can achieve the same effect with less computing power, or even surpass it. Why do you insist on preventing us from releasing products?”

This doubt is not only shared by Liang Wenfeng, but is also a widely shared consensus within his company Deepseek.

They started developing general artificial intelligence at an earlier time and have stockpiled a large number of computing power cards. The products they can currently produce are enough to beat ChatGDP, not to mention the new models they will launch in the second half of the year.

But the strange thing is that Chen Pingjiang has been suppressing them, not allowing them to release the information to the public, and requiring them to keep it confidential.

Seeing that Liang Wenfeng was somewhat impatient, Chen Pingjiang smiled leisurely, handed him a cup of tea, and then asked, “What’s the matter? Is there a lot of discussion in your company now?”

Liang Wenfeng nodded: “Yes, everyone is somewhat unconvinced. They clearly think that their products are more powerful, but they are always suppressed and not allowed to come out. As a result, their enthusiasm has been worn out.”

Chen Pingjiang nodded and expressed his understanding.

The normal entrepreneurial process of a company should be to first announce the deepseekV3 model, rely on its strength to beat ChatGDP to attract a lot of attention, and then raise funds.

Continue to develop new products and always maintain a leading edge.

It can not only inspire the people, but also suppress the arrogance of the opponents.

After all, the people over there are always talking about artificial intelligence, but they are also blocking China in return.

After thinking for a while, Chen Pingjiang asked, “Since you came to ask me this time, I will talk to you openly and honestly. But this matter is between you and me. After we leave this room, I will pretend that I have never said anything.”

Liang Wenfeng was startled and subconsciously sat up straight.

Given Chen Pingjiang’s identity, since he said this, it is obvious that this matter is of great importance.

As expected, the first sentence that followed made Liang Wenfeng break out in a cold sweat.

“In fact, your company and your project have long been on the radar of the senior leaders. The reason why we are following your lead is not only my opinion, but also the leaders’ opinion. This time, artificial intelligence is a battle for the fate of our country.”

“A battle for national destiny!?”

Liang Wenfeng was stunned. He didn’t expect to hear such a shocking word. Chen Pingjiang nodded and said, “The other party launched a new product, and we launched a more powerful one. Then they copied it and took the lead again. This is a common practice of Internet companies, but this time, I don’t plan to play it that way.”

Liang Wenfeng was startled: “Then how do you plan to play?”

Chen Pingjiang raised his lips and said, “If you want to play, let’s play big. Since they think they are far ahead of us in general artificial intelligence, let them maintain this illusion and confidence. When they continue to make additional investments, bubbles will inevitably form. When the bubbles get bigger and bigger, we will announce new products to help them burst the bubbles.”

“Of course, bursting their bubble and their self-confidence is not the most important thing! Don’t they always think that their general intelligence is awesome? That’s easy. We just beat them in the field they are most proud of, and completely open source the technology used by Deepseek so that any Internet company in the world can easily copy it. By then, there will be thousands of large models all over the world, and none of them will be inferior to the others.”

“Guess what the situation will be like then!?”

After listening to this, Liang Wenfeng thought about it carefully and was horrified. This strategic concept was so grand and astonishing.

What are the consequences of making the technological achievements developed over many years open source for free?
“That will naturally have a huge impact on companies like OpenAI, Google, and MATE. We are open source and free, but they charge for it. Internet companies around the world have already deployed large local models. Who would use their large models? If no one uses them, they will have to lower the threshold for charging or even give it away for free, and then they will inevitably be unable to make ends meet.”

Chen Pingjiang smiled and nodded: “Yes, the first step is to make the general large model very cheap, so that they can’t make ends meet. But you only saw the first point, not the second point, which is the most important means, and can be called an open conspiracy.”

Liang Wenfeng thought for a long time but couldn’t come up with an answer.

After whetting everyone’s appetite, Chen Pingjiang began to tell the story.

“The US has long been on the road of deindustrialization, and it is difficult to bring manufacturing back even through high tariffs and sanctions. In contrast, industrial manufacturing is our strength, and artificial intelligence has long been integrated into the blood of various industries. In the end, artificial intelligence will always be applied, and general artificial intelligence for language cannot represent everything.”

“We can be on par with the other party in the field where they are the strongest, and then destroy the market prospects of this field, but we can still run wild through industry, while the other party can only watch helplessly.”

Hearing this, Liang Wenfeng’s breathing became rapid.

Only now did he realize how big a game Chen Pingjiang and his superiors were playing.

It’s like playing chess!
After everyone’s chariots, horses, and cannons were matched against each other, we still had five pawns that had crossed the river, while the other side had nothing left.

“So, what your company needs to do now is to keep up with the pace of technological advancement, or even surpass it. But the most important thing is to effectively reduce the computing power cost required for general artificial large models.”

Chen Pingjiang concluded: “The other party is more powerful in hardware. They will use their chip technology to greatly increase computing power, attempting to surpass our products in multiple dimensions through computing power. But you, Deepseek, must not let them succeed. You must make a breakthrough in the algorithm and achieve the same effect at one-tenth or even one-twentieth of the other party’s price.”

“By then, no matter how good their hardware is, global investors will have to wonder whether this accumulation of physical resources is really effective or a waste. Moreover, it is not certain that we will be able to catch up with them in hardware by then.”

To be honest, it took Chen Pingjiang two years to slowly summarize and figure out this macro-strategic intention.

And the more I think about it, the more I think it’s feasible.

To this end, he communicated many times before finally making up his mind to deploy.

If it were in the past, when the first generation of deepseek came out, Chen Pingjiang would definitely be eager to launch it.

If you do that, you will gain fame, but the actual effect will be greatly reduced.

just now……

He just needs to wait, wait quietly!
Build up your strength secretly and strive to improve the quality of every line of code.

When the other party’s interstellar gate opens, close it with your own hands.

At this time, Liang Wenfeng asked again: “Mr. Chen, I get the strategic intention. There is one thing. The company’s investment is still very large. Not to mention the personnel cost, there is also the part of purchasing computing power cards. Where will our subsequent funds come from?”

When saying this, Liang Wenfeng was a little nervous.

Because Deepseek is not a wholly-owned subsidiary of Chen Pingjiang, the shareholding ratio of Liang Wenfeng and Chen Pingjiang is 51:49.

If Deepseek delays the release of new products, the funds on its books will soon run out, and it will not be able to raise funds and will have to rely on Chen Pingjiang.

But in this case, will Chen Pingjiang use his financial advantage to forcibly dilute Liang Wenfeng’s equity?

This is what Liang Wenfeng needs to consider!
Although he believed that Chen Pingjiang would never do such a clumsy thing, it was still better to ask for business reasons to avoid disagreements in the future and ruin the good situation.

In response, Chen Pingjiang just smiled and said, “I know what you are worried about. But don’t worry, for me, money is no longer what I am chasing. We can draw up an agreement, not even a bet agreement. The agreement stipulates that I will pay for all the subsequent R&D investment of Deepseek, and you only need to ensure that you work hard on R&D and always keep the technology advanced.”

“In other words, whether our strategic intention can be achieved in the end depends on whether Deepseek can shoulder the responsibility. If your product is not competitive, then everything will be in vain.”

Only then did Liang Wenfeng breathe a sigh of relief.

As long as Chen Pingjiang is willing to continue investing without having to dilute his own equity.

Technically, he is still very confident.

After all, current products have proven that deepseekV3 is far superior to ChatGDP.

In the following period, Deepseek just needs to firmly maintain its advantage.

Liang Wenfeng is no longer a novice in entrepreneurship. He certainly knows how cruel the ordinary gambling agreement is. Once the performance or task is not up to standard, the other party can take away almost everything, or even kick him out.

However, the requirements given by Chen Pingjiang were so relaxed that even if there was no need to sign a gambling agreement, or even if it could only be described as a “memorandum”, he did not dare to neglect it.

We must firmly maintain our technological advantage.

After a long while, Liang Wenfeng said, “This time when I go back, I plan to change some of the company’s employment rules.”

“Ok?”

“All those who have studied or worked abroad will be banned first, and the passports of the company’s core technical backbones and management will be taken away first. Internal personnel of the company are strictly prohibited from participating in any AI exchange conferences or related academic conferences to prevent technology outflow.”

Chen Pingjiang looked at Liang Wenfeng with appreciation and said, “Not bad. I will also deploy a security company to enter Deepseek to prevent technical espionage and unwarranted harassment. In short, the next one or two years will depend on your performance.”

He added: “Don’t worry about the money. Don’t forget what your previous company Huanfang did!”

“You mean stock trading?”

Chen Pingjiang nodded: “Once Deepseek’s new product is released and has an impact on the market, the other party’s technology stocks will inevitably fall. I don’t need to teach you the rest, right? After all, that is your main business.”

Liang Wenfeng’s eyes lit up, and then he laughed: “Indeed, when the time comes, as long as there is enough capital to enter the market to short and cooperate with the option put, it will be a piece of cake to earn billions of dollars at a time. One fight can recover all the previous investment!” (End of this chapter)