I went out for about 20 days this time. I originally wanted to go to Wuhan to see the situation of Yadai Company in person. After all, their achievements during this period have been very impressive. It is necessary for me to go and see it in person to express my feelings. recognition of their efforts.
It's just that after this circle, I have given myself so many tasks, and I don't have time to go to Wuhan. In addition, I am about to start a major operation, so I have to calm down.
So he returned to Kyoto early. As soon as he came back, he went to the Qingnong Biological Laboratory to check the situation here and the completion status of the tasks received by each company.
Finding that it would take half a month to get all the components in place, he took advantage of this time to quickly finish the matters at hand.
So starting from the next day, Zhao Yi basically stayed at home and worked honestly. The first thing was to write an intelligent programming tool.
The Lenovo function that was commonly used in the previous life is indispensable. In addition, the integrity of the program needs to be automatically detected, and the detection of BUGs also needs to be automated.
In fact, programmers actually do not spend much time writing code. Instead, they spend much more time finding problems and fixing bugs.
After solving these two problems, the efficiency has at least doubled. The remaining problem is how to deal with programming. Zhao Yi plans to use some of the technologies and algorithms he used in primary artificial intelligence.
In other words, in the future, programmers only need to be responsible for the algorithm and architecture parts. For the programming part, the embedded intelligent program module written by him will automatically complete it.
Although it cannot reach the level of a junior artificial intelligence program, because of the programming process of a junior artificial intelligence program, Zhao Yi basically does not have to do anything himself.
However, the intelligent program module he produced cannot yet reach the level of fully automatic programming. It belongs to a semi-automatic programming mode. However, the quality of the automatically generated code is also very high. Programmers only need to make slight corrections to make it usable.
This alone has improved the efficiency of programming tools by more than 10 times. However, the efficiency has improved, but the requirements for computers have also become higher.
Because this level of intelligent program, if it wants to be fast and at the same time improve its accuracy, it must provide a large amount of computing power support. It is very difficult to drive it with the current level of PC.
However, running on a server can also cause a lot of trouble, especially now that we are more and more accustomed to using notebooks to write programs. This change is quite uncomfortable for programmers.
Therefore, it is necessary to customize a high-performance desktop computer and notebook computer from Polaris Company. It is also very difficult to bring ordinary logic chips, and the computing power provided by ordinary GPU may not be up to standard.
Fortunately, Polaris Company has developed an intelligent supercomputer before. The chip used is a smart chip, which can effectively reduce the consumption of computing power by intelligent programs.
Therefore, if they want to use this programming tool, each of his companies needs to customize intelligent personal computers from Polaris Company. Otherwise, it will be very laborious to carry and even more frustrating to use.
After spending two days to get the smart programming tool ready, Zhao Yi worked non-stop to improve the new cross-platform language developed by Fanxing Software Company.
One is that this new cross-platform language needs to be further improved, and the other is that this cross-platform language needs to be improved so that it can adapt to more usage scenarios, and it also needs to have a certain degree of intelligence.
Having an intelligent language, not to mention other things, at least saves a lot of effort in post-maintenance. Many problems can be easily detected, which can greatly reduce operating costs.
Many programs will not cause any problems during the writing process, but once they are officially deployed in the environment, various inexplicable problems will appear, which will take operation and maintenance personnel a lot of time to check the problems.
Therefore, Zhao Yi added a certain amount of intelligence to this new cross-platform language. In addition to further reducing programming costs, the main purpose is to reduce the workload of post-operation and maintenance personnel and reduce operation and maintenance costs.
Of course, Zhao Yi also hopes to further improve the inherent low operating efficiency of cross-platform languages. The underlying language used by Fanxing Software was C language.
This determines that this new cross-platform language cannot be higher in efficiency than C language. After all, the bottom layer has passed a lot of encapsulation, which reduces efficiency.
In order to improve efficiency, Zhao Yi directly abandoned the virtual machine written in C language, and instead used a large number of machine language and assembly language, so that the efficiency of the cross-platform language built on it can be greatly improved.
But if you do this, it will not be something ordinary people can do if they want to update virtual machines in the future. After all, this technology is too low-level. There are only a few people who know it, and those who are proficient in it are even rarer.
Therefore, Zhao Yi also specially wrote a toolkit for writing virtual machines. Using this toolkit to write virtual machines is much simpler, and there is no need to directly write machine code and assembly.
He can use the toolkit language to write code that ordinary programmers can understand, and then these codes are compiled into low-level machine language and assembly language through the compiler that comes with the toolkit.
However, various chips will appear in the future, and the instruction set of each chip may be different. If adaptability cannot be achieved, the corresponding virtual machine will need to be rewritten for each instruction set chip.
In order to solve this problem, Zhao Yi added an intelligent detection program to the virtual machine toolkit, which can directly extract the instruction set from the chip, and analyze the purpose of each instruction in the instruction set through self-test mode.
With this smart module, new chips will appear in the future, and virtual machines that adapt to them can be quickly written through this toolkit, making this cross-platform language more vital.
After finishing all these things, Zhao Yi personally wrote a program to test the execution efficiency of this new cross-platform language. It was found that the execution efficiency was much higher than that of C language.
This is the benefit brought by the extensive use of machine language at the bottom level, that is, the proportion of assembly language used is not very large. The reason why not all machine language is used is because some instructions are simpler to use assembly, and the performance loss is not much.
In addition to improving the performance of the virtual machine, it is also necessary to optimize the compiler of the cross-platform language itself so that the instructions in the generated executable file are closer to the machine language, which will better complement the optimized virtual machine.
Just like this, the overall execution efficiency of this cross-platform language has more than doubled compared to the C language, and it can be used as a mainstream development language in the future.
To be honest, although the execution efficiency has been improved so much, it is still limited by the operating system itself, which makes Zhao Yi feel regretful.
After finishing working on the cross-platform language, you need to start writing some related components based on this language, such as big data analysis components, artificial intelligence components, etc.
Since he was already familiar with these things, the algorithm was written very quickly, and the architecture design was not difficult at all. It only took three days to write dozens of components.
Then there was the task of building a technical communication platform, which was even less difficult. I completed the work in less than a day, and then uploaded the components I wrote.
Later, Zhao began to write a more efficient intelligent game engine, and did not plan to use other languages. He directly used this cross-platform language to write this intelligent game engine.
Games written using this smart game engine will also have cross-platform properties in the future. This has many benefits. In the future, there is no need to write different codes for consoles and PCs. All that is needed is to differentiate in operating habits. .
Due to the need to increase the development efficiency of smart game engines by 50 times, Zhao Yi reviewed the previous versions and decided to tear them down and rebuild them all, just in time to use this new language.
Therefore, the time it took Zhao Yi was extended instantly, and it took him a full week to complete this intelligent game engine.
Zhao Yi only made a sample himself to test indicators such as development efficiency and operating efficiency. This process could not be programmed using primary artificial intelligence, so it took him a day.
Looking at the game program in front of him, Zhao Yi felt that if he used the previous smart game engine to create it, it might take two months of work to complete. From this perspective, his requirements were met.
More importantly, programs written in new languages are very readable and do not require understanding of complex underlying logic. Novices who have studied for a year can get started writing more complex programs.
Coupled with the large number of components he has written before, the development work is no longer so demanding, which brings huge benefits to various companies that are currently severely short of manpower.
Of course, if you want to learn how to build a wheel, the requirements are relatively high. You need to learn a lot of underlying computer logic, otherwise you will not be able to build a wheel at all.
However, these deeper things do not require too many people to know. This work can be left to more talented people. Ordinary programmers do not need to understand it so clearly.
As for the need for ordinary programmers to optimize programs written in this new language, Zhao Yi thinks that they are overthinking, because he has already achieved this to the extreme. If ordinary programmers want to do more, it will be counterproductive.
As long as ordinary programmers write programs according to the specifications of the new language, and with the assistance of intelligent modules, there will basically not be any big problems, and there will be no loss in performance.
However, if you want to become a master in the future, just being able to program will not be enough, because most of the programming work will be undertaken by intelligent modules. On the contrary, the requirements for algorithms and architecture are relatively high. If you want to get a higher salary in the future, you must go here. efforts.
In other words, in the future, programming will no longer be a physical job, but a real mental activity. The level of programmers will no longer be reflected in programming, but in their understanding and innovation of algorithms and architectures.
This will expand the scope of recruiting programmers. In the future, background in the software industry will no longer be particularly emphasized. Students majoring in science majors such as mathematics and physics will have a much lower threshold for entering the software industry.
Therefore, Zhao Yi needs to adjust the software professional courses in universities and increase the number of courses related to algorithms and architecture. He also needs to increase training in logical thinking skills and add more theoretical teaching.
When machines can replace human labor, emphasizing hands-on ability will actually seem backward. On the contrary, theories that are not valued highly will play an increasingly important role in the future.
He took a look at the software-related major courses at Tianwen University, and then began to modify the syllabus and content of related courses. He dictated it, and the junior artificial intelligence helped him enter it, and the efficiency was slightly improved.
After finishing these tasks, he directly sent these new courses and content to Tianwen University, and also gave a copy to Tianwen University of Information Technology, which had just been established not long ago.
Then he drafted a document and submitted it to the Ministry of Education. As for whether other schools would follow suit and change, he could not interfere. He could only do what he could do. Anyway, he said everything he needed to say.
In his heart, he hopes that schools across the country will make changes in the curriculum of software-related majors, because the major domestic software employers are his companies.
Most of the software major graduates from these schools will enter his companies. If they do not make changes, they will still need a period of adaptation to their companies.
I believe that all major universities understand this. Zhao Yi's suggestions still have some weight. No matter what, they should make some changes based on his suggestions, and even directly use the relevant syllabus and tutorials of Tianwen University.
Because these schools will understand that if they do not modify it according to Zhao Yi's opinions, the students they train will definitely be relatively weak in competitiveness, which will hinder the students' careers.
The success of a student's career seems to have little to do with the school in the short term. In fact, this is not the case. If everyone knows that the employment situation of students graduating from this school is not ideal, the industry influence is very weak.
So as long as individuals are unwilling to apply for these majors if conditions permit, then the influence of their majors will become weaker and weaker, and they will eventually face the risk of being withdrawn.
In addition to the more intelligent programming tool he wrote himself, Zhao Yi also upgraded the programming tools he currently uses so that existing tools can program this cross-platform language.
This upgraded programming tool will be an open version and will also be free for major schools to use so that they can teach new cross-platform languages.
As for why the latest intelligent programming tools are not open to the public, the first one is that they do not want to reduce social employment. This is the direct reason. Zhao Yi can prevent his own company from laying off employees, but it is hard to say for other companies.
The second is to reduce the competitiveness of countries such as Europe and the United States in this area. Our foundation in this area is relatively weak. When we rely on such efficient tools to catch up, there is no need to capitalize on the enemy.
The third is that in school teaching, more attention needs to be paid to basic teaching. If such a highly intelligent programming language is used, students' foundation will be unstable, which will be detrimental to their future development.
After finishing all this, Zhao Yi began to help the Building Block Software Company write the intelligent core of a smart city brain, which is not difficult.
Because the intelligent core he prepared for the Building Block Software Company was a subsystem of primary artificial intelligence. Since it was for internal use, he did not pay much attention to this aspect and was not worried about the possibility of leaks.
Moreover, the subsystem of primary artificial intelligence is used as the core of intelligence to process data faster and the operation is more stable. More importantly, it can automatically repair some problems and has strong self-healing properties.
After finishing the smart core, Zhao Yi sent it directly to the Building Block Software Company. After all, they had been stuck on it for a long time. If you give it to them earlier, they can save some time.
After that, Zhao Yi began to use the products sent by Qianxun Technology Company one by one, and analyzed their system architecture and algorithms to see how to improve them.
The first is the portal website. Zhao Yi plans to use two sets of display modes. The first set is the traditional category mode, and the second set is the intelligent recommendation mode.
The reason for preparing two sets is that in the PC era, the traditional portal model is more adaptable to users, but this model also has serious shortcomings.
For example, it is not possible to push the latest news to users in a timely manner, nor is it possible to recommend similar news information based on the user's preferences, nor is it possible to make users stay longer.
As for the smart mode, the biggest advantage is that it does not require users to specifically search for news, and it has a more friendly experience for users who are not very purposeful.
Moreover, a large number of news recommendations that meet users' preferences can make users spend a lot of time on news sites without knowing it, and sometimes they can't stop reading. This has a great promotion effect on the website's advertising revenue.
At the same time, Zhao Yi also added self-media functions to the portal. Although it is currently not suitable to integrate the news portal and the short video platform, it also laid the foundation for the future integration of the two.
As for their short video platform, Zhao Yi completely follows the intelligent recommendation model. There are not so many categories of directories. The page adopts a waterfall design. Users only need to drag down with the mouse to refresh the videos recommended by the system.
When writing this, Zhao Yi remembered a photo sharing website that had been popular for a period of time in his previous life, which made him think he could try it out, so he easily wrote a photo sharing website using a model similar to that of the short video platform.
After doing this, Zhao Yi opened the music platform of Qianxun Technology Company. It has all the basic functions. Most of the songs can be listened to online and cannot be downloaded. If you want to download, you have to spend money.
Because these songs all have formal copyrights, whether they buy out the copyright or choose to share the copyright, music companies and platforms have invested a lot of money, and it is impossible to use them completely free of charge.
However, after Zhao took a look, he found that most of the songs were from Europe, America, Japan, and Hong Kong and Bay Area. There were far fewer songs from the mainland, and previous revolutionary songs accounted for a huge part.
There is also pop music, but it has been brought about by the boom of TV dramas and movies in recent years, and there are not many styles. This allows Zhao Yi to truly feel what the current situation of the mainland music market is. (End of chapter)