16
Aside from arXiv, Wang Xu also created CNKI, which is essentially the Chinese version.
Management of arXiv was also handed over to the GIT management committee, and distributed mirror sites were established worldwide.
In fact, the backend support architecture for these two websites is very similar, with only the presentation layer differing.
Wang Xu would not keep such a globally-oriented website in his own hands, nor would he allow the United States to have full control over it; instead, he made it operate like the United Nations.
In short, this meant that basically no one could completely sever the connection between arXiv and the world.
Furthermore, arXiv even opened up an external access Automation interface, allowing users to pay for executing download tasks similar to web crawling.
Because it is resource-intensive regarding hardware and bandwidth, this service requires a fee.
CNKI, however, was more complicated because it truly required negotiations with major copyright institutions.
Domestic ones were manageable, but international ones were quite troublesome, so it had to be run at least semi-commercially; otherwise, the costs might have been truly astronomical.
The establishment of these two pieces of infrastructure was actually in preparation for the future era of artificial intelligence.
Wang Xu had not experienced a full-fledged era of artificial intelligence in his previous life.
In this life, however, he was using various means to accelerate its arrival, which made him feel somewhat apprehensive.
In his previous life, China's backwardness during the era of artificial intelligence had been plain to see.
Of course, it might just have been a difference in strategic approaches, but most of the so-called domestic artificial intelligence was essentially just open-source large models from elsewhere.
Wang Xu couldn't claim to know everything, but many were indeed like that.
Wang Xu did not release the most popular models from his future; instead, he released relatively early content, such as convolutional neural network models, which were currently the primary focus for papers from Wang Xu's laboratory.
Ultimately, this technology was better suited for the field of image recognition, forming the foundation for products such as license plate recognition, handwriting recognition, and facial recognition.
By laying the theoretical groundwork now and allowing a few years for promotion and development, it would gradually emerge in practical applications.
Of course, there were also recurrent neural network models better suited for human language, which were primarily used for semantic analysis, such as analyzing the context of text.
These two types of neural networks would be the hottest topics during the early 21st century, and Wang Xu's laboratory was now slowly and steadily publishing papers in this field.
Producing one or two breakthrough papers and a few extension papers each year was considered a return to a normal state for the laboratory.
Now, it was considered a fairly productive laboratory in the field of artificial intelligence and had gained some renown in the industry.
However, it was still all theoretical.
Since China's hardware, software, and network equipment were still in their initial stages, Wang Xu consciously suppressed the output in this area; there was no real need for it, as even if it were developed, China would not be able to use it.
He did, however, provide some to the authorities, letting them decide for themselves whether to utilize it.
After all, both neural network learning and training required massive data resources, and the necessary computational power was virtually limitless.
Wang Xu even reminded the authorities to attach importance to data collection, as these things would become the new gold mines of the future.
Of course, there was no need to collect data just for the sake of it; rather, it should be an integral part of the process of digitization and informatization.
In fact, according to global technological development, from the eighties into the future, it could be roughly divided into five stages: digitization, informatization, Automation, unmanned operation, and intelligence.
By the way, this "digitization" was not the same concept as the industrial digitization advocated in later generations.
That referred to the comprehensive digitization of the IoT era, encompassing everything from factories and equipment to processing procedures and products.
The digitization currently being pursued was only aimed at products and design processes; there were not even strict requirements for the production process.
It was much like the "Toss the Drawing Board" project proposed by the state a few years ago, which was an attempt at industrial design digitization for this era.
In fact, there was another plan simultaneously that had not been widely publicized: the "Toss the Letterhead" project, which marked the beginning of China's move towards office digitization.
People in later generations would call these two projects the "Double Toss Projects," but because few people used them, they were gradually forgotten.
The "Toss the Drawing Board" project had actually started in '91, but progress had not been rapid.
Initially, it was only promoted in key enterprises and large manufacturing firms.
However, due to too many constraints, the project's progress was never satisfactory.
In his previous life, it wasn't until around '97 that the pace gradually accelerated.
There were many reasons for this, but the core issues were nothing more than three: a lack of money, a lack of personnel, and a lack of indigenous software.
The nineties was a decade of rapid, booming development for the domestic software industry.
These software companies essentially developed by copying the functions of similar overseas software one-to-one.
However, most of the algorithms and code were indeed created by themselves, because they simply couldn't copy the original things.
For many things, they truly didn't understand how the originals were implemented, so they had to use decompilation to examine the assembly code, derive the algorithms, and finally implement them on their own.
In his previous life, Wang Xu had worked at a company developing CAD software.
Their testers would compare their own product against AutoCAD; whatever the original could draw, they had to be able to draw the same, otherwise, it was considered incorrect.
There was certainly no creativity involved, but it was through this process that China produced its first batch of CAD software products.
In the early nineties, relying on the progress of the "Toss the Drawing Board" project, they still managed to sell some.
However, after '95 or '96, when CD media became widely popular, entire new versions of AutoCAD could be taken on a single disc, even including many pre-drawn resource libraries.
From then on, Chinese CAD plummeted and gradually faded from the public eye.
AutoCAD could almost be said to dominate the entire landscape.
Therefore, Wang Xu dared to say that these companies, by allowing piracy to become rampant in China, must have had deep-seated considerations.
They did not rely on the Chinese market for survival, and the Chinese market at that time could not support any large software company.
Yet this indulgence, or perhaps even intentional action, could naturally eliminate almost all small competitors, extinguishing even the possibility of them budding.
Perhaps some would say they didn't necessarily think so far ahead or so deeply, and that this was just a conspiracy theory.
Wang Xu had no evidence, but the factual outcome was exactly that.
Wang Xu's open-source plan, however, was meant to reshuffle the entire software ecosystem from another level.
By using free, open-source products to gather the power of the entire world, and then using these products to form countless domestic and foreign commercial companies, these companies would, in turn, nourish the open-source community, forming an entirely new software ecosystem.
Since these things were highly relevant to the film and television industry, Wang Xu naturally placed special importance on their development.
Shuimu itself had a company making CAD products, and now, with the open-source underlying layer, they had fully migrated to a brand-new architectural system.
Early CAD was, in essence, just for drawing.
The word "CAD" itself is an abbreviation for Computer-Aided Design, but it didn't truly offer much assistance for design itself from a product design perspective.
It did, however, offer a huge improvement in reducing drawing time and difficulty; for instance, the various annotations that caused the most headaches could be completed almost with a single click.
In particular, it could speed up drawing modifications.
Electronic files could be modified directly on the original, and then re-plotted, whereas physical drawings offered almost no such possibility, which saved a vast amount of time on repetitive drafting.
This was the original intention of the so-called "Toss the Drawing Board" project: to liberate engineering and technical personnel from cumbersome drafting, allowing them to truly devote their thoughts and efforts to designing the products themselves.
The 3D product design that Wang Xu brought was conceptually completely different from traditional design methods.
It presented products directly through 3D modeling, allowing them to appear intuitively in the engineer's field of view first, enabling engineers to make 3D modifications intuitively based on functionality.
It could even perform some animated simulations of working conditions, such as checking whether gear rotation caused interference or if a shaft conflicted with the casing.
Furthermore, it could define materials to automatically generate manufacturing processes, tolerance requirements, and various other complex operations.
Only when it was time to generate drawings for manufacturing would it automatically create views from various angles, and it could also freely generate complex technical drawings like cross-sections and assembly diagrams.
This was true digital design.
Even in the West, only the top aerospace design fields had adopted such a design model, such as France's Dassault; their digital design software was the best in the industry, at least back in the nineties.
In fact, China had never slowed its pace in this regard.
The "New Leopard" recently developed at Xi'an Aircraft was built using fully digital design—at least in his previous life—and the platform they used was CATIA, yes, a product of Dassault.
However, in this life, it was said—just hearsay, mind you—that they had chosen Wang Xu's open-source product.
Although this product was not yet mature enough, it was more advanced in its basic functions and clearly had stronger future scalability.
Therefore, they teamed up with many universities in Beijing, Shanghai, and Xian, including Shuimu, to jointly conduct research and development on many foundational software items, all to build a tool platform for the digital "Flying Leopard."
From time to time, Wang Xu would receive a thick collection of questions, which he would answer based on his own understanding.
If he couldn't answer something, he would tell them straight, but he did not personally participate in the project itself.
One had to understand that aircraft design was very different from traditional CAD.
Designing an aircraft, especially a modern third-generation fighter-bomber, required the use of dozens of different specialized software programs.
Most of these software were developed by companies in the aviation field, which then charged exorbitant fees to iterate and create better products.
Similarly, after a period of market validation, these software became increasingly mature and eventually won the entire market.
This was, of course, not a major issue in the West.
At that time, these software could also be sold to China.
On one hand, their own software was not yet fully mature; on the other hand, they might have thought that even if China managed to design it, they wouldn't be able to build it.
Furthermore, once they were used to using their software, China's security agencies might not be able to maintain as much secrecy; after all, it was far easier to plant something in software than in machinery.
But China had stumbled its way forward, and the philosophy of "buying is better than making" was not that prevalent in those units.
While it wasn't non-existent, most military industry workers were genuinely afraid of the various unexpected glitches that kept popping up.
Therefore, in this life, they chose an immature, and one could even say incomplete, open-source product, built their own core tools, and used the money they would have spent on overseas software to develop it themselves.
This might even slow down the development speed of their fighter jets, but they still chose this path.
Perhaps it was due to a line like this: "In the past, I had no choice; this time, I choose to do it myself."
This was a massive project, and the help Wang Xu could provide was limited; after all, he hadn't even seen many of the software products involved.
But all things share the same underlying principles, and this highlighted the importance of open-source foundational libraries.
Some of the basic libraries that no industrial software could do without were available in the open-source community.
Things like 2D and 3D graphics processing, image processing, finite element analysis, various mechanical calculations, various fluid calculations, and even many biological and chemical simulation calculations—basic libraries for all of these now existed in the open-source community.
Moreover, because they adopted a unified communication standard, calling and connecting these libraries was very convenient.
They could be integrated from the bottom up, without needing to be connected from the top level as was the case with foreign finished products, where integration was extremely difficult and could only rely on finished data files for communication.
In the unified design network that Wang Xu conceived for them, these underlying libraries were further encapsulated and called, connecting to each other to form larger functional libraries, which then served as the basis for directly forming a networked work platform.
This was a design concept from nearly twenty years in the future.
Wang Xu had planned everything—distributed computing, cloud storage, unified work platforms, and external display dashboards—and had provided design advice.
In this way, they could make full use of emerging network technologies to make coordination between departments easier; this, in fact, already surpassed all existing industrial software products.
Wang Xu even planned an Automation manufacturing module, from CAD to CAM, which could save a vast amount of work in coding manufacturing processes in the future.
Of course, those Automation processing devices were not only expensive, but most could not be purchased, and even if they were, their usage would be restricted.
Naturally, it was even less likely that they would open up the underlying control rules to you; you would only be able to use the programming language they provided to write the Automation manufacturing processes.
In the future, once they had their own Automation Machining Center, they could connect things like four-axis, five-axis, or even six-, seven-, eight-, or nine-axis equipment to directly control production.
Just imagine: unmanned workshops.
In fact, this was also an innovation in the philosophy of military industry workers.
Using a fully 3D design concept to build new aircraft was a brand-new attempt; even advanced Western companies had not fully realized this.
The vast majority of parts still relied on engineers to draw 2D CAD drawings directly; this process relied on the engineers' spatial imagination, not the capabilities of the software itself.
However, Wang Xu's support for 3D here could be said to be ahead of its time.
Not only did it have underlying 3D libraries, but it also had a complete operating interface.
This was something he had prepared for his own filmmaking; without an interface, how could computer laymen possibly use it?
(It's hard to say if there will be more; I can't be sure.
I'm heading out in a bit and don't know when I'll be back.)
