Chapter 354: Galaxy Of Shining Large Models
“Isn’t this very normal? For us, 400 million Hong Kong Dollars is an astronomical figure, possibly one we couldn’t earn in a lifetime.
But for Brother Ran, this is just a little extra from a transaction with Young John Morgan. Young John Morgan doesn’t lack this bit of money, and similarly, Brother Ran doesn’t lack it either.”
When they returned to their cramped apartment that evening—a tiny, narrow space where even turning over was difficult—after Li Yiqing and Xu Xian vented about recent experiences, Xu Xian’s reaction was unusually calm.
“That’s right, Ran Shen is the world’s top in mathematics, the top in aerospace engineering management and technology research and development, and the top in artificial intelligence.
A single identity might have someone to compare with him, but with these three identities stacked together at the same time, there’s no one in the world who can compare with him.
Moreover, in each one, he can absolutely rank in the top five. In pure mathematics, there might be more than five with greater contributions than him, but among those in their prime, Ran Shen can absolutely rank in the top five.
For such a globally unique talent, an old office building in Central is indeed just a negligible little gift.
I can understand this; there’s really nothing unbelievable about it. It’s just that I’ve always known Ran Shen is awesome, and on various wealth lists, he has tens of billions of US Dollars in assets. Bloomberg, Forbes, and others even directly give Ran Shen several question marks, meaning he’s not included in the evaluation.
Whether it’s Apollo Technology or Deep Red, neither has listed. The primary market is also not very public and transparent, and the entire fundraising process hasn’t been publicly disclosed, so they use question marks.
These institutions all say the most conservative estimate won’t be less than tens of billions of US Dollars.
But after all, this is just valuation, a number. No matter how big this number is, I lack a real sense of it.
Right now, in the hot and humid summer of Hong Kong Island, I’ve been visiting owners one by one, negotiating in turn. I’ve been to that office building more than once—the old-style decoration, the outdated elevators, the aroma of roasted meat outside, the dust from cars speeding away—these are all things that just happened recently.
They remain in my mind, not yet faded. Our entire team put in massive effort, just to give Young John Morgan a small surprise during communication with Ran Shen.
This is the mighty power of wealth, and also the most real portrayal of wisdom transforming into influence in reality.
Do you know this feeling?
Especially now that I’m in my cramped apartment in Hong Kong, while Ran Shen is the same age as us.”
Li Yiqing’s tone was full of emotion.
Xu Xian’s voice came from the other end of the telephone: “Brother Qing, of course, I certainly understand what this feeling is.
I never told you that the major PhD graduation problem my mentor assigned me, Brother Ran just glanced at it and directly held a Zoom meeting on the spot to remotely solve it for me.
I suspect he just read the problem and knew what to do.
I think Brother Ran has integrated mathematics into his instincts. After witnessing it, I believe the mathematics community’s past statement that Poincaré was the last mathematician to take all of mathematics as his research field is wrong.
At least, Brother Ran is such a person. I don’t know who the last one is, but Poincaré definitely isn’t.
So what I want to say is, this isn’t much more exaggerated than the pure wealth gap.”
Xu Xian thought to himself, you’ve experienced the shock that I experienced long ago:
“So it’s no big deal. If you always compare with perverts, I think your mindset will twist.
Brother Qing, your bachelor’s, master’s, and PhD from Yenching University are already very excellent.”
Li Yiqing explained: “No no no, you misunderstood what I wanted to express.
I’m not seeking comfort, nor am I frustrated. I just want to vent and share my experiences with you.
Additionally, after this experience, I don’t want to do finance anymore.
Although I can’t become a cloud-dwelling figure like Ran Shen, who can change countless people’s fates with a breath, like you told me, your university roommate came to you just to stay at Deep Red.
But I also want to strive once.
Doing finance boils down to the service industry—either you have resources yourself or become a broker of resources.
Otherwise, even if you’re extraordinarily talented and incredibly successful in trading, earning billions through stock trading on your own, such a life is still bland, facing nothing but numbers, like the great Gatsby, wrapped up in consumerism.
Whereas a life after successful entrepreneurship is truly interesting enough.
If I can’t become a dragon that swallows clouds and spits fog, at least I have to strive in that direction.
I plan to first join a promising company, then jump out to start my own business.
Only like this will life be interesting enough; even if I fail, I’ve tried.
In a place like Hong Kong where 100 square meters counts as a luxury apartment, it’s really not suitable for me.”
Xu Xian heard ambition in the other’s words and laughed: “Then I wish you success in advance.”
Xu Xian fully understood the other’s mindset. Just like after Lin Ran solved his major problem, he only lay flat for half a year, then went to his mentor for another major problem to work on.
“General Manager Lin, this is our financing proposal. Please review it.”
Song Nanping knocked on the door and walked into Lin Ran’s office.
Lin Ran took the document titled “Apollo Technology Financing Proposal” and asked: “Has Pony reviewed it?”
Pony is the President, and Song Nanping is half a level below as Vice President.
He replied: “Yes, this version has been revised based on Pony’s proposal.”
“Apollo Technology is a private enterprise registered in China, with main businesses covering high technology manufacturing, aerospace equipment manufacturing, infrastructure construction, and other fields.”
Because they pulled 30 billion US Dollars in wool from the Morgan family—10 billion for Saturn V, 20 billion for Burning One Modified.
The latter is more expensive than the former because the key engine technology of Saturn V isn’t transferred at all, while Burning One Modified is a complete package transfer, including engine technology.
In the future, whenever General Aerospace needs moon landings, they’ll have to place orders with Apollo Technology to buy Saturn V engines.
Honestly, the entire shell and such, buying the whole package from China, is much cheaper than making it themselves.
Although Young John Morgan still has passion now and wants to manufacture on American soil, so he’ll manufacture on American soil.
But Lin Ran thinks that sooner or later, the other side won’t be able to stand it and will choose the buy-over-make development path, completely surrendering, directly having Saturn V shipped from China to American ports.
So Apollo Technology isn’t short of money at this point.
With Young John Morgan around, for things like Nvidia’s latest AI computing cards, they have plenty of ways—ship to Lion City first, then from Lion City to Hong Kong, and finally to the Mainland.
But these US Dollars are all in Hong Kong accounts, so Apollo Technology ultimately decided to raise 200 billion RMB from the market.
China’s current situation is quite awkward.
Yanjing has money, residents have money, but the money can’t be spent.
Residents dare not spend, bank deposits grow year after year.
Yanjing has money too and tries every means to stimulate consumption, but consumption can be stimulated through consumption vouchers and shopping subsidies. However, employment positions and industrial upgrades ultimately need enterprises to handle.
The money doesn’t know how to flow into suitable enterprises and can only circulate idly in the banking system.
Now, Apollo Technology, a promising aerospace enterprise, has finally emerged. Just the spin-off Deep Red alone makes Yanjing willing to invest 200 billion.
So this financing has Yanjing eagerly wanting to stuff 200 billion into Apollo Technology.
“Wasn’t it supposed to hold a financing presentation meeting, then negotiate one by one?
I see here that the targeted issuance parties are already determined.”
In this report, the targeted investors are led by an investment fund under China Development Bank, with China International Capital Corporation and China Investment Corporation following, allocating 5 percentage points of the financing quota to Pony.
“That’s right. Yanjing’s meaning is that since it won’t raise from private capital anyway, it’s all financing between state-owned enterprises, avoiding putting you in a difficult position.
Otherwise, local state-owned investment institutions from your hometown like Yuexiu Capital and Shenzhen Capital Group from Guangdong Province would bother you through various connections.
Shanghai local institutions like my former unit Shanghai Sci-Tech Investment would also try every means, each showing their talents to visit.
To avoid such situations, Yanjing simply designated these three as state-owned capital representatives.” Song Nanping explained.
Lin Ran smiled wryly; he knew this was very likely: “Indeed, thank you for thinking of this.”
The financing proposal Song Nanping brought was written very standardly, and it even adjusted the voting rights allocation issue they thought Lin Ran cared most about.
That is, A/B shares with same shares different rights model—Lin Ran’s B shares, one share’s voting rights equal to 10 A shares’ voting rights.
This is also how domestic internet giants ensure the founder team, despite shares dropping below 20 percent, still firmly grasp company control.
Not only do these three state-owned institutions hold A shares, but Shanghai Sci-Tech Investment will also transfer its shares to these three state-owned institutions, converting them to A shares upon transfer.
In the proposal’s final result, Lin Ran holds over 70% voting rights in Apollo Technology, ensuring absolute control over the enterprise.
After reading it, Lin Ran nodded: “Just proceed with this; I don’t have many opinions.”
Yanjing thought Lin Ran cared, but Lin Ran didn’t care at all.
Because whether control is in his hands or not has little to do with regulations; it relates to Yanjing’s will.
If Yanjing is willing, even with only 10% voting rights, enterprise control is in his hands. If Yanjing is unwilling, even with 90% voting rights, he can’t decide anything.
Like Jack Ma saying the wrong thing at the Bund in Shanghai, immediately vanishing from the Chinese Internet, with control over Alibaba and Ant Group plummeting—only professional managers seen, no voice from Jack Ma.
And this is still a private enterprise; no need to say more about a company like Apollo Technology.
Lin Ran is very clear: as long as he’s there, company control will always be firmly in his hands. After he’s gone, if the heirs lack the ability, it’s a disaster rather than a blessing.
After a brief chat with Song Nanping, he got up and headed to the meeting room. Summer vacation had started, and the classmates from Shanghai Jiaotong University’s Lin Ran Mathematics Class had finished final exams; everyone was in place, ready to begin summer internships.
Yes, without exception, all 60 classmates applied to intern at Deep Red.
Currently, Deep Red and Apollo Technology share office space. After all, Deep Red only has over 200 employees left, not enough to fill even one floor, so no need for a separate office area.
The outside world went completely crazy upon learning that Lin Ran Mathematics Class classmates could directly intern at Deep Red with no threshold—all could go.
This year’s gaokao even had a provincial top scorer publicly declaring they would apply to Shanghai Jiaotong University.
Shanghai Jiaotong University even nabbed two IMO gold medals this year—ones that used to only go to Yenching University.
This made local rival Fudan sour to death; Tongji isn’t sour—Tongji can go to Apollo Technology, highly matching, and Tongji never aimed for top 3 anyway.
“Hello everyone, I see you’re all very excited about interning at Deep Red.
I want to tell you all that your internship at Deep Red won’t be like at other internet institutions, doing miscellaneous tasks. I’ll assign you a complete project, assisted by Deep Red staff—the ten Deep Red employees you see here.”
Lin Ran pointed to the ten employees sitting on the right side of the meeting room, who looked noticeably more mature, but not by much.
AI engineers are all very young.
“With them leading you, do a large model purely based in the liberal arts field.
It needs to cover content including simulating sociology scenarios, analyzing psychology experiment data, political policy evaluation, economic model prediction, etc. Restrict it to the liberal arts field, with sufficient professionalism, the model reaching human expert level on specific tasks.
I’ll name this model Shining Stars.”
Huh? This was a topic no one had thought of.
Such demand exists abroad, but in China, damn, it’s too scarce. A large model limited to the liberal arts field—what use could it have?
Everyone really couldn’t think of it.
For Lin Ran, this is what he needs. Named Shining Stars, symbolizing that when placed on the far side of the moon in the future, it can illuminate humanity’s stars.
So, when America tried to single-handedly break away from the Bretton Woods Agreement in 1971, and human economists didn’t know what it meant, Shining Stars could guide humanity.
When the Soviet Union is in distress, struggling economically with high oil prices in the past, Shining Stars could tell them exactly what mistake they made.
When it comes to China, China has no problems.
Anyway, it’s Shining Stars.
“I can briefly tell everyone what your work might cover, and from what angle I’d approach if I did it.
The most important is actually data—focus on high-quality, field-specific text, like philosophy ethics texts, metaphysical works, sociology theory research and field investigation reports, political science policy documents, data texts, and such.
To find these, you need to know where to find paper indexes, where to find open resources. If not enough, purchase from field journals to collect data.
After collecting open data, if needed, contact the comprehensive department to acquire usage rights for complete databases from field journals.
I expect to ensure at least 50 billion tokens for professional depth.
After determining data, clean it—remove noise, duplicates, and biased content.
Some texts may be very popular with tons of citations or downloads, but actually valueless—this needs your judgment too.
This is the data stage work. In model development and training stage, you can directly use Deep Red for training—no need to develop from scratch; have staff assist you in training based on existing models.
If training results aren’t satisfactory and model adjustments are needed, you can try yourselves first. For anything you don’t understand, email me, and I’ll arrange time to provide answers for everyone.
My goal is to hope you open source a sufficiently neutral, objective, professional model. When outsiders use it, throw any social topic at it, and it gives sufficiently professional answers.
If good enough, I’ll consider selling it to Weibo as an independent, objective third-party opinion on Weibo social topics.
So, everyone, your work is very meaningful.”
Lin Ran inwardly thought: everyone, it’s not very meaningful—it’s extremely meaningful. It might change the course of the entire era, change the entire world.
As for why not do it himself?
Because Lin Ran will stay in this spacetime for a long time; he doesn’t want to spend special time on this. Assigning it to students, under Deep Red’s existing resources, as a practice topic is just right.
It’s not to be finished this summer; it’ll be an ongoing project. If this batch doesn’t finish, the next batch of students continues. Shanghai Jiaotong University’s freshmen get better year by year, and Lin Ran Mathematics Class threshold rises yearly.
He’ll stay here about seven years—plenty of time.
“Through this project, you’ll contact almost all work needed for artificial intelligence large models. I believe this will be quite helpful for you now and in the future.”
Finally, after Shining Stars launched, its impact in the 1960 spacetime aside for now, on Weibo, it directly exploded the cesspool—tons of bloggers accusing this large model of not being objective, not neutral, not some third-party, but purely Lin Ran’s personal will.
Even classmates involved came out to post explanations, only to be collectively attacked.
This was something Lin Ran and the present classmates never anticipated.
