Algorithm system

Continuing StormWhat surprised Lin Yuan even more was yet to come. His move to open-source all of Tesla Delivery’s algorithms was truly groundbreaking. It meant he personally gave up enormous commercial interests and, at the same time, thoroughly offended all foreign capital.

After all, Tesla is a foreign company. Although Tesla Delivery was built by Lin Yuan from the ground up, and Musk had promised him that Lin Yuan would have complete control over Tesla Delivery.

However, neither Musk nor Lin Yuan anticipated Tesla Delivery would achieve its current scale and influence. For the foreign capital behind Tesla, it naturally viewed Tesla Delivery as its own.

Therefore, Lin Yuan’s sudden open-sourcing became an inexplicable major international commercial dispute. From the looks of it, it was highly likely to involve state relations, as foreign capital worried about the intervention of Dragon Country authorities behind it.

But Lin Yuan was actually very naive. He simply wanted to break the entry barriers of the food delivery industry, and as a result, he inadvertently became involved in an international dispute.

After Zhang Yan’s retelling, Lin Yuan finally understood the core of this international dispute. Many people were concerned that the commercial rules between Dragon and Eagle Countries were being broken. Although there had been constant disputes in the trade sector in recent years, unlike the foolish polar bear in the North, the conflicts between Dragon and Eagle Countries followed rules.

The form of conflict between Dragon and Eagle Countries was a trade war. What is a trade war? It’s a war in the trade sector fought when neither side wants to completely overturn the table. Even if such disputes seem intense, they still reveal that neither side wants to completely tear things apart for now.

In contrast, on the polar bear’s side, there was no such thing as a trade war. Eagle Country and the polar bear directly confiscated each other’s assets. Their disputes were without rules, like a dagger revealed at the end of a scroll.

It was precisely because of this that Lin Yuan’s action could easily be interpreted by those with ulterior motives as a precursor to intervening in trade disputes, which naturally made many foreign investors tremble with fear.

It was only then that Lin Yuan understood why there was such a massive effort to protect him, while also bestowing upon him such a great honorary title.

After a thorough investigation, it became completely clear that the technology used by Tesla Delivery had no connection whatsoever with Tesla, or even with foreign capital. In other words, Tesla Delivery was essentially just operating under the Tesla name, with all actual technical work led by Lin Yuan.

However, the most troublesome part was that Lin Yuan had long ago transferred the intellectual property of Tesla Delivery’s core

AI Manager

algorithm to Musk. In essence, Lin Yuan open-sourced an algorithm whose legal ownership did not belong to him.

This kind of operation is a complete breach of confidentiality in business, but Lin Yuan truly didn’t overthink it. For him, transferring the algorithm’s ownership to Musk was out of his gentleman’s promise. Both he and Musk knew that the algorithm was actually entirely created by Lin Yuan alone.

Therefore, Lin Yuan had no psychological burden about open-sourcing. He had even discussed it with Musk beforehand, and the latter was quite supportive of him.

Because this algorithm is now nominally owned by OpenAI, and OpenAI, due to its complex equity structure, is becoming increasingly divergent, to the point of being jokingly called

CloseAI

because it’s not open at all.

Lin Yuan could sense a hint of Musk’s helplessness, because this was increasingly contrary to his original intention of founding companies like Tesla and SpaceX, but he was powerless to do anything about it.

So, Musk was also very happy to see Lin Yuan boldly open-source this algorithm, and he even paved the legal way for Lin Yuan’s open-sourcing.

Musk had Tesla’s legal team prepare ample documentation to argue that the

AI Manager

algorithm actually utilized and referenced a large amount of open-source software. Given the license contamination of open-source software—users of open-source must also open-source—it was legally justifiable to open-source this algorithm.

In Musk’s original words,

OpenAI has already spent 20 billion dollars, and I don’t know how much more money it will take to see a profit. Everything here is too inefficient, Lin. You know, although the 0 to 1 flash of inspiration is one in a million, we don’t have the ability to go from 1 to 100 here. This is truly frustrating.

I’m increasingly finding that no matter how well designed, it seems we can’t solve the production problems here.

So I’m fed up. I originally wanted to use your Tesla Delivery’s success story to convince these people here, but they are completely engrossed in pure software products.

They still prefer light assets and playing with concepts. What the future holds, I don’t know.

Our financial strategy has dragged everything down. When huge profits can be extracted by relying on financial products that do not generate value, the threshold is raised, and the top earnings in other industries must also reach the level of the financial industry, otherwise why wouldn’t the smartest people engage in finance?

Lin Yuan deeply understood Musk’s helplessness. The smartest minds on The Other Shore had all plunged into the financial industry. What was more terrifying was that after the financial industry’s profit margins spread to other industries, would the executives in the AI field also start comparing their salaries to those in finance?

If the income gap between the two industries is too large, people can’t help but ask a question—why bother with general AI? Why not just develop specific quantitative trading AI software around financial transactions?

However, these specialized AI tools for the financial sector are clearly not enough to tell a story or support a future trillion-dollar industrial chain.

So, in the end, they had no choice but to accompany the crown prince in his studies, and other high-end industries also followed suit by raising labor costs. This led to the situation where OpenAI had already spent 20 billion dollars without seeing any hope of profitability.

If Tesla is already being outcompeted by newcomers in the electric vehicle sector, it at least retains the title of a pioneer. In the AI field, however, OpenAI, this unblooming flower bud, gives Musk a sense of crisis that it could wither at any time.

Because Eagle Country’s efficiency is simply too low, and large AI models are an emerging field that doesn’t require a massive industrial chain.

Building electric cars requires a large number of supporting companies in the upstream and downstream supply chains. These companies cannot all be established first before car production begins; this also involves a host of issues such as cost problems brought about by economies of scale.

However, the AI field currently does not have a massive supply chain system; at most, it’s GPU computing cards plus a large number of algorithm engineers. Although NVIDIA is far ahead in GPU computing cards, there has never been a precedent in human industrial history where a competitor was completely stifled by relying on a single type of equipment.

To give a simple example, the total market value of companies in various segments of the electric vehicle supply chain combined is definitely much, much larger than that of the giant NVIDIA.

Scaling and cost control are far more difficult to solve than one or two technical problems, which is counterintuitive. People often think how brilliant the pearls on the industrial crown are, but manufacturing a lighter at the lowest cost is itself a very impressive art.

When this advantage in scaling and cost-effectiveness is reflected in all aspects of manufacturing, its cumulative advantage far exceeds one or two high-end technologies.

Especially for the AI field, this area’s updates and changes are simply too fast, and the variables are too great.

Even if there’s a two or three-year gap between Dragon and Eagle Countries in the AI field, everyone knows that even if Eagle Country were given thirty years, it wouldn’t be able to rebuild its manufacturing system.

A single sharp sword cannot defeat a well-organized legion; this is reality.

This is also where Musk’s helplessness lies.

In fact, the moment Tesla rose relying on Dragon Country’s production capacity, Musk already understood in his heart.

The storm stirred by Lin Yuan continues.