Chapter 355: Maybe Room Temperature Superconductivity Is The Norm
“Professor, I have a question.”
After Lin Ran finished speaking, a classmate sitting in a back seat raised his hand high. After Lin Ran nodded, he stood up and said.
“Go ahead.”
“Before you speak, let me introduce him to everyone first. This is Zhao Tongxue from Tsinghua University Yao Class.”
“This time, for our summer internship, besides opening it to everyone, there are also a few additional spots opened to classmates from other universities.”
Lin Ran said.
Zhao Tongxue is indeed from Tsinghua University, but it’s not that everyone from Tsinghua University must be allowed in for the internship. Rather, some influential figure put in a word, hoping to arrange for him to intern.
Basically, those with solid connections were let in.
It’s just an internship; making it convenient for others means making it convenient for oneself in the future.
Lin Ran had no elitist ideas. Whether domestic or abroad, relying on connections is inevitable; it’s decided by people.
In human society, all you can do is go with the flow.
“Professor, that’s right. What you’re doing is of course very meaningful.”
“But the problem is, are you trying to compete for discourse power in the public opinion arena?”
“This is a very dangerous thing. I think if you really want to do this, it’s best to fully communicate with relevant institutions in Yanjing and get their approval before proceeding.”
“Let me give a very simple example to help you understand this more directly.”
“Previously, Baidu had Baidu Snapshot, and everyone could use Baidu Snapshot to find some past websites that had already disappeared.”
“I believe the Baidu Snapshot function still exists, but it’s no longer open to the public.”
“A big reason is that they don’t want people to make compiled editions.”
“The same event, five years ago and five years later, both from official reports—compiled editions like that would seriously affect credibility.”
“Why the inconsistency between five years ago and now? If the past had issues, should those who made decisions back then take responsibility? If the present has issues, that’s even more terrifying.”
“This time span might not be five years; it could be ten years or even longer. And at every time point, if you make compiled editions, you’ll find that the propaganda tone on the same event keeps changing.”
“Why is that? And what does making compiled editions aim to do? Trace the past? Hold someone accountable?”
“The most interesting here is the Great Monk. The 2015 investigation report showed he had no issues; others’ accusations were all false, none true.”
“Are the issues disclosed ten years later different from ten years ago? No.”
“So why did the tone change over these ten years? Wasn’t it said back then there were none? Why now there are?”
“Making you do compiled editions is even worse: if the 2015 official investigation report on the Great Monk had issues with fabrication, then every subsequent report would be scrutinized with doubt.”
“This is the weight that Leviathan cannot bear.”
“Professor, this is my worry.” Zhao Tongxue’s expression was sincere and earnest: “Don’t try to become a coordinate in the public opinion arena.”
Lin Ran thought for a moment and then asked: “Zhao Tongxue, what I’m doing is just a large model in the humanities and social sciences field. It doesn’t involve collecting the kind of data you mentioned.”
“It’s just providing users with thinking from a professional perspective at the theoretical level.”
“It’s not a compiled edition, nor any coordinate. Current large models can do this too; I just want to go further and make a model with higher professional integration.”
Zhao Tongxue smiled wryly: “Professor, of course I know you’re well-intentioned, with no malice, just doing this from a research perspective.”
“Also to give us a suitable practice project.”
“But the problem is, being misunderstood is the destiny of the speaker. Once you make this thing and put it on the internet, it will face public opinion pressure from all sides; some will treat it as a sword.”
“You can’t control the direction this sword points.”
“Professor, you should know what the Gini Index is, right?”
“The statistics bureau occasionally publishes the Gini Index, and Southwestern University of Finance and Economics previously did Gini Index statistics too, with a huge difference between the two.”
“(Gini Index situation from official sources)”
“When Southwestern University of Finance and Economics did this statistic, I think it was just for academic consideration, a sociology project. We clearly got 0.61 or 0.62, while official is around 0.46—this data difference is so large.”
“Neutral academic research, once public, is no longer under control; it will be interpreted in all sorts of ways.”
“It could even become a sharp weapon to attack you personally, Professor, the best tool to provoke your relationship with Yanjing.”
“Professor, my suggestion is: you can do it. This model is of course very meaningful and can be made public. Just push the responsibility to us—it’s the project you gave, and the thing we made.”
“After communicating with Yanjing, put it on social media as something like an academic achievement showcase, lower external expectations, and avoid becoming a measuring stick.”
“Professor, your status is too high and your reputation too great. If done in your name or Deep Red’s, the whole thing can easily spiral out of control. If as a student work, it’s just right.”
“Students make mistakes; it’s normal.”
Lin Ran nodded: “Okay, I understand. I’ll seriously consider your suggestion.”
The project’s head also has the surname Zhao. Zhao Songxia, who successfully stayed on, went on stage after Lin Ran left and did a simple self-introduction to everyone.
At the end of the first day, back at his workstation, Zhao Songxia inwardly lamented: what kind of monsters are these people who came?
Every classmate from Lin Ran’s mathematics class is one of his direct lineages; he couldn’t really treat them as interns to boss around.
As for the other classmates not from Lin Ran’s mathematics class, there were very few, only 12 people. From that Zhao Tongxue who spoke in the morning reminding everyone, one could see they had extraordinary backgrounds; Zhao Songxia felt he had to treat them well.
Moreover, among these 12, the lower the education level, the less he dared to offend them. There was even a junior from some unknown second-tier university—god knows what big shot that was.
Lin Ran’s second task today was to discuss with the representative from Huawei.
The past half year could be called the half year of artificial intelligence’s explosive progress. GPT pulled open the shocking curtain, making the outside world realize that artificial intelligence has such power and can really be used to increase productivity.
From the 2016 blueprint to the end of 2022 implementation, six years had passed.
In the first half of the year, Wenxin Yiyan and Deep Red fired the first shot, followed by large models from ByteDance, SenseTime, Baichuan Intelligence, and other companies being opened to the public one after another.
In the entire first half of 2023, there were over 20 financing events in China directly related to large models, and the number of various large models released domestically exceeded 100.
It was also known as the Hundred-Model Battle, far crazier than the Hundred-Group Battle from the group-buying era.
Like abroad, those digging for gold currently show no sign of profit anytime soon, while those like Huawei selling computing cards—shovels—made a fortune.
Despite so many manufacturers doing large models and buying computing cards from Huawei, Apollo Technology and Deep Red remain their most important partners in this field—without question.
So this time, a Huawei executive in charge of computing card business came, along with the entire expert team.
Lin Ran said there was an important matter to discuss.
“Good afternoon, everyone. I’d like to briefly discuss my idea.”
“We all know that America keeps blocking Nvidia’s most advanced computing cards from entering domestically precisely to leverage artificial intelligence’s so-called technological singularity to completely surpass us.”
“This is a well-known open scheme.”
“Whether outsiders believe the LLM technical route can really lead to AGI, or whether unmanned factories really need no people—anyway, Europe and America media are hyping this narrative.”
“On this, I think under our current technical system, achieving a overtake on the curve is very difficult.”
“We have too many backward shortcomings. Semiconductors are a massive system, from top to bottom, needing time to grind through.”
“I’m confident we’ll succeed.”
“But if we can make the surpassing come earlier, I think it would be even better.”
“So I invited everyone to discuss whether my concept is feasible.”
“To put it this way: without a breakthrough in basic theory, there won’t be leapfrog technological development.”
“But if there are enough comprehensive technological breakthroughs, it can also lead to results similar to leapfrog development.”
“Currently, the only path for a breakthrough in the semiconductor industry is the emergence of new materials to replace past silicon-based chips.”
“Undoubtedly, achieving this in the short term is difficult.”
“Superconductors might be a good shortcut.”
An expert in the audience raised his hand and asked softly: “Professor, are you referring to the LK99 that Korea just came out with recently?”
“Do you think that’s a true room-temperature superconductor?”
Lin Ran looked puzzled: “No, I’m not familiar with what you mean by LK99.”
The expert then felt relieved. He had thought if it was LK99, with nothing set in stone, wouldn’t this be a wasted trip?
LK99 was a room-temperature superconductor claimed by a certain Korean professor. His team also published a paper, with a very simple synthesis path—simple for superconducting materials—and netizens jokingly called it a room-temperature superconductor refined like alchemy.
Plus, the material showed some traits unique to superconductors, so upon release, it caused a huge public opinion uproar.
It made many experts who had studied superconductors their whole lives question their lives: if this thing is so easy to make, what was I doing all my life?
It even sparked discussions that maybe room-temperature superconductivity is really that easy, and everyone overcomplicated it in the past.
In short, lots of debate—some jumping out to say it is room-temperature superconductivity, others saying it isn’t.
After hearing this, Lin Ran waved his hands repeatedly: “I’m not talking about that. Of course, if true room-temperature superconductivity emerges, that would be even better.”
“My thinking on this is based on the absence of room-temperature superconductivity.”
“We’ve already started the infrastructure project at the Lunar South Pole. In the shadowed region of Shackleton Crater at the Lunar South Pole, we’ve found water ice and placed various sensors.”
“Its temperature stays around 100K year-round, which is -173.1 degrees Celsius.”
“Though compared to common low-temperature superconductors like niobium-based compounds that need 9-10K, it’s still a bit high.”
“But that doesn’t mean we can’t utilize this environment.”
“Do you all understand what I mean?”
“If we elevate our perspective to the universe level, Earth’s temperatures aren’t common; ultra-low or ultra-high temperatures are the norm.”
“For those places, their room-temperature superconductivity is our ultra-high temperature superconductivity or ultra-low temperature superconductivity on Earth.”
“The cost of transporting semiconductor equipment to the Moon is far lower than the cost needed to overtake on the curve on Earth.”
“Our reusable rocket technology is fully mature. Burning One Modified routinely shuttles between Earth and Moon. Since we can do east-data west-compute, why not earth-data moon-compute?”
East-data west-compute refers to a past strategic thinking in China: placing data centers in the west, where there are abundant hydropower resources, sparse population, less power pressure, and lower temperatures.
Resulting in densely packed data centers in Guizhou Province—Apple’s cloud in Guizhou, not in other provinces.
The engineers present were already listening with blood boiling.
Linking semiconductors with aerospace was just too exhilarating.
“On Earth, we need expensive liquid nitrogen cooling systems to maintain superconductivity; the Moon is naturally a paradise for superconductivity.”
Lin Ran continued:
“We’ve done some preliminary research. Materials from the copper oxide family, like the Bi-2223 mentioned earlier—bismuth strontium calcium copper oxide in Chinese—achieve superconductivity at critical temperature Tc around 110K under normal pressure.”
“Moreover, it shows semiconductor characteristics in its parent state. Through oxygen doping, we can tune carrier density, transitioning it from insulator to superconductor.”
“This means chips can handle semiconductor logic and superconducting transmission simultaneously, reducing energy loss.”
“Besides copper-based, there’s also iron-based.”
“In the laboratory, we’ve prepared single-layer FeSe thin films on SrTiO3 substrate, achieving superconductivity steadily above 100K.”
“Its parent is semiconductor; interface effects enhance superconductivity. Imagine growing such thin films via molecular beam epitaxy in the Moon’s vacuum, integrating onto silicon-based chips—low gravity also reduces defect formation!”
“Of course, it’s not all benefits without drawbacks.”
“FeSe is a star among iron-based superconductors. Its semiconductor parent allows building hybrid systems: superconductor-semiconductor heterojunctions.”
“Using FeSe as the superconducting layer and GaAs (gallium arsenide) as the semiconductor substrate enables Josephson junctions for qubits.”
“In Earth labs, we’ve seen Tc above 100K, but Moon’s cosmic rays might interfere with Cooper pair formation. I think we may need to add radiation shielding, perhaps using boron-doped diamond as a buffer since it superconducts at low temps too, Tc around 10K, but more stable.”
“K3C60 is another option; its normal superconductivity is only 20K, but photo-induced it jumps above 100K.”
“It’s an organic semiconductor, highly flexible, suitable for the Moon’s curved terrain.”
“We can conduct tests on the Moon, combining photoexcitation to create transient superconducting circuits.”
“In short, technology can’t achieve leapfrog breakthroughs short-term, but by leveraging the existing environment, we might still accomplish great things.”
