Xueba's computing system

Chapter 264 Looking for a Partner 2
Lin Yuan even went out of his way to find the video about OpenAI that the boss of Dudu had posted.

Its boss is from Silicon Valley, and he was considered a technical elite who returned from overseas. He was also an engineer. But technology is like martial arts, if you don't practice it, you will lose your skills.

So, that guy said something extremely stupid - it is impossible for Longguo to have a local large language model comparable to OpenAI.

Lin Yuan scoffed at this.

However, according to the Dudu family's usual character, it is normal for the other party to make such remarks, because Dudu has never fought a tough battle.

If Google hadn't committed suicide, if it hadn't been for the protection of its own country, if it had been fair competition, just looking at Baidu's positive ranking strategy, Lin Yuan could only think of "haha".

How can we expect a company that doesn't have the gene to fight tough battles to be strong?

However, this kind of view on AI large language models is not uncommon internationally. Because OpenAI invested 200 billion yuan, it scared latecomers.

In the eyes of latecomers, it is almost impossible to compete with OpenAI, because AI requires continuous investment, and the 200 billion is only the initial stage investment. What is more frightening is that the AI field belongs to the field of global supply and demand, and the top AI talents are mobile around the world, so it is impossible to reduce R&D investment by using exchange rate differences.

This is just like electronic components and energy sources such as oil, where the bulk transaction prices are common around the world.

You pay high-end AI talents according to your own currency, but they just turn around and run to the United States. Isn't it nice for them to earn US dollars?

Of course, these are past experiences, and Lin Yuan does not deny that these experiences are wrong.

However, the field of AI is far from mature, which means that the field of AI is still in a turbulent period of development.

This thing is not like the internal combustion engine, which has been toyed with for hundreds of years and all the technical paths have been explored. In a field where the technical path is already extremely clear, the large amount of R&D costs incurred in the early stage will block the way for latecomers.

This is not only based on the recognition of customers in the market, nor is it based on the maturity of the product. The most important thing is that various patent barriers have been set up on various technical paths, making it difficult for latecomers to take a step forward. Then latecomers naturally have no way to go.

Therefore, before the rise of new energy vehicles, domestic cars were unable to enter the mid-to-high-end market, and even the price above 15 yuan had already begun to weaken.

However, the field of AI is a fucking wasteland with no established path.

OpenAI seems to have spent 200 billion, but it has not created a feasible breakthrough technology path. In essence, it is still running wildly on the traditional AI training path, relying only on the power of inference.

Then, relying on his sensitivity to technology, even if he did not have the system, Lin Yuan would come to a conclusion - black swan events are most likely to occur in such a barren field.

Just like the great voyages of the past, on the vast and foggy sea, who knows when an island will faintly emerge.

Benchmarking AI's technical implementation, what if someone comes up with a new algorithm or new idea? After all, AI training relies too much on the performance of algorithms and models.

As for the degree of effort corresponding to computing power, anyone who has received education knows that studying hard does not necessarily lead to good grades.

Therefore, Lin Yuan even felt that OpenAI's route of madly piling up computing power might be a conspiracy by the other side. After all, the other side can fully play their own advantages in this way.

The world's best GPU computing power card company is on the other side of the strait, and so is the largest capital.

If we take the technology + capital intensive route, there is really no place in the world that can out-play the other side.

Lin Yuan plans to find a company with backbone and dreams to cooperate with. A company like Dudu that has already fallen on its knees cannot be saved.

Soon, Lin Yuan contacted the Chrysanthemum Factory.

He had high hopes for the Chrysanthemum Factory because it was an ambitious enterprise. But he soon discovered that being too ambitious was not a good thing.

Because the Chrysanthemum Factory solemnly declares that they have their own rhythm.

Juhua Factory has its own understanding of AI large language models. Juhua Factory does not agree with spending a lot of money on general large language models, but instead builds an AI ecosystem around its own products.

Simply put, the Chrysanthemum Factory now has a wide range of businesses, from telecommunications equipment to cloud computing, from smart devices to car manufacturing, and to the lower-level operating system and chip design. The Chrysanthemum Factory is more concerned about the construction of the entire ecosystem.

As for the attitude towards AI big language models, they are keeping up with the pace of industry development and trying to use big language models to serve their own products, for example, in their own cars.

Chrysanthemum Factory has no interest in making big news with a big language model. Their views are similar to Lin Yuan's. There are still too many uncertainties in the field of AI. When facing these uncertainties, the best way is to keep your feet on the ground.

The so-called down-to-earth approach means building AI around applications.

Lin Yuan agreed with this. If he did not have the computing power system, he would definitely do the same.

After all, ordinary companies cannot make storytelling their core business. For companies like the Chrysanthemum Factory that have been subjected to suppression and sanctions, being down-to-earth is a matter of life and death. Limited R&D funds cannot be invested in areas that will not see cash in the short to medium term.

But Lin Yuan also thought of another thing. The Chrysanthemum Factory has been driven to death by the other side in the fields of mobile phones, chips and communications. If it makes any major achievements in the field of AI, it will not attract enough firepower. By then, I am afraid that the Eagle Sauce will do its best to deal with the Chrysanthemum Factory.

Therefore, for strategic considerations, Juhua Factory will definitely focus on making breakthroughs in the current key area - the chip industry chain.

What's more, its car-making sector is just emerging.

Given these two considerations, naturally, I have no interest in AI large language models.

However, the Chrysanthemum Factory was willing to give Lin Yuan the title of Honorary Technical Consultant, and the other party also emphasized that it was an honorary title, not an employment relationship.

This is certainly not because the Chrysanthemum Factory is stingy, but out of courtesy.

With Lin Yuan's current reputation, the Chrysanthemum Factory thought it had no suitable position for him. But Lin Yuan had come to discuss cooperation, so for the sake of appearance, they at least showed their friendship to Lin Yuan.

Lin Yuan naturally declined the title of honorary technical consultant. He respected the decision of the Chrysanthemum Factory and had no plans to communicate further.

So he crossed another one off his list.

There are only a few companies left in the country with a certain scale and strength.

Coarse grain factory? Lin Yuan was thinking about it.

But he had little confidence in the coarse grain factory because the supply chain of the coarse grain factory was not completely in his hands.

As the saying goes: Every generation of Snapdragon produces a new generation of coarse grains.

The supply of core components is controlled by the enterprises on the other side of the strait, not to mention the extremely chaotic equity structure behind the coarse grain factory. No matter how powerful Lei Zi is, he can't twist his arm against his thigh. Lin Yuan is now hated by the capitalists on the other side of the strait, so he won't deliver himself to the door.

It is very likely that the investors behind OpenAI, Tesla and the Grain Factory overlap, so what did Lin Yuan achieve by going to the Grain Factory? It was like going to tell his enemy: I want to make a big news to mess with you.

Of course Lin Yuan is not that stupid.

(End of this chapter)