As a mage, I just want to pursue the truth

Chapter 362 M Language and the First Occupation Killed

The Lion City appears to be an independent country, which is still attached to American to some extent.

In the official website of American, it is roughly written like this:

The two sides have built a broad and enduring relationship based on shared economic interests, defense and security cooperation, and enduring people-to-people ties.

In recent years, American has actually been strengthening its diplomatic relations with the Lion City.

Among them, a fact sheet on strengthening the strategic partnership with the Lion City was released in 2021, which was repeatedly emphasized during Vice President American's trip to the Lion City.

From 2021 to 2031, the relationship between the two parties has been strengthening.

The Li family changed hands and sold the Lion City to Zheng Li, which was unexpected by American. When they knew, the Li family had already completed the intention negotiation with Zheng Li.

The intention is over, and it's too late to stop it now.

And for American, Lion City is not that important compared to the potential future benefits from Zheng Li.

Therefore, they remain on the sidelines of this transaction.

And if the Li family can make a deal with Zheng Li, can't American?

Under the manipulation of American's real family in power, Zheng Li obtained all of American's hands in Lion City with three potions that prolong his life for ten years.

At the same time, the previous agreement on leasing the Lion City military base was abolished. Since then, American's interests and organizations in the Lion City have fully retreated.

The Lion City councilors who thought they could pull a tiger's skin behind their backs, that is, American, lost their temper in an instant.

American has withdrawn his troops honestly, so why bother.

As for contacting the armed forces in Lion City for a coup, this is pure wishful thinking.

The image of Merlin in the outside world is not an entity at all. At the same time, he took over the artificial intelligence of the Lion City. How can your armed forces kill the artificial intelligence?

Furthermore, this is the result of the decision of the whole people. You have no legal basis, and there is no real physical goal at all.

In addition, the armed forces in Lion City are not stupid to bet on their future and even their lives for the benefit of politicians?
The remaining more than [-] civil servants in Lion City were divided and transformed by Zheng Li.

Zheng Li used the name of the Lion City to newly establish several state-owned enterprises.

These firms will recruit from a broad pool of public servants.

Those who are honest and obedient participate in the selection, and those who are dishonest and obedient, just make trouble, Zheng Li expressed an attitude that it is impossible to compromise.

At the same time, the benefits of the unemployed will be greatly improved, and money will be paid directly according to the minimum guarantee.

After the combination of punches, all the unemployed civil servants became obedient.

Everyone knew that Merlin would not do any secret operations. Since it was a selection test, these civil servants who were selected from the best naturally felt that they had a chance.

And for them, the state-owned company established by Merlin has a bright future, no matter in terms of treatment or technology demonstrated by Merlin.

In between, Lion City released a series of artificial intelligence frameworks, in which it reads:

"The artificial intelligence framework announced below follows the open source spirit that has been popular since the rise of Bluestar Internet, and discloses some AI frameworks that are better than Bluestar's existing AI frameworks in some aspects.

All individuals and businesses can use these AI frameworks for free. "

In Zheng Li's view, Bluestar's productivity growth rate is not fast enough, and the progress in the integration of artificial intelligence and other industries is still not as good as expected.

After fully displaying his muscles in the field of artificial intelligence in the government artificial intelligence in Lion City, he hopes to further promote the progress of Bluestar's artificial intelligence and industrial integration by making the AI ​​framework public.

The process can be completed by him, or he can bring some inspiration to Blue Star's artificial intelligence field and let them complete it under his own guidance.

The further implementation of artificial intelligence, combined with bio-photovoltaic technology, will reduce the employed population of Blue Star by 80.00%, or even more.

A large amount of productivity is released, whether it is to engage in more technical work, or to enrich the virtual world ecology, it is better than doing some meaningless work now.

Zheng Li's time in Blue Star is not short. Although Zheng Li has been at the top of society for more than ten years, he has a certain understanding of Blue Star's social ecology through the Internet.

In his opinion, many of Blue Star's jobs are meaningless, they exist purely for the purpose of torturing people, and they don't even produce any output.

The manufacturing and service industries have produced output at any rate.

Even if it is productive work, Zheng Li also feels that it is meaningless. The intelligent life that was born with great difficulty in the vast universe should not spend its life in this kind of boring mechanical labor.

Intelligent life needs to find the meaning of life by itself.

Instead of setting numerous checkpoints designed by society, even if you stumble through all the checkpoints, you will still feel like your life is wasted.

For Zheng Li, the Lion City is just a test field, which will eventually be extended to the entire Blue Star.

Zheng Li's public AI framework under the domain name magic.com includes a new programming language and open source packages around this programming language, that is, tools around this programming language.

The reason why python has been so popular in recent years is not unrelated to the fact that python has a large number of useful toolkits.

These encapsulated packages greatly reduce the difficulty of programming in python language.

As soon as the M language released by Merlin was announced, it quickly triggered an earthquake in the entire Internet industry.

"Too humble, not just being good in some ways, being good in everything.

Although I haven't figured it out yet, I always feel that this thing is much more advanced than Bluestar's current mainstream artificial intelligence framework. "

The main thing is that this thing is too professional, and things that are too professional are destined to be insulated from the mainstream group.

Even professional algorithm engineers need a period of time to study.

Most of the comments from netizens are concerned about whether they will lose their jobs:
"I feel that Merlin is not a magician, but a programmer, trying every means to promote the large-scale implementation of artificial intelligence.

First, use artificial intelligence to replace human labor in the Lion City, and then release open source AI programming languages ​​and frameworks.

I seriously doubt that the use of magic by a magician is somewhat similar to programming? "

"Merlin accepted the supernatural powers, I finally found a job, Nima won't optimize me again, I don't want to graduate from Weibo.

This is the dream job that I had a hard time finding. "

The person who posted the following Weibo is a Master Huang.

Such occupations do exist, and they pay well.

But for him, belonging is a prophecy.

After Merlin released the AI ​​language and supporting frameworks and toolkits, these big Internet companies asked the following algorithm engineers to put down all their work:
"Everyone's only job during this time is to learn how to use the M language, the M framework and the supporting series of packages.

Then combine our work to see where the M language can be used for rewriting. "

Among them, the first to be used is AI Huang.

AI pornography has always been the core appeal of content security. Basically, from pig factories and goose factories to Weibo and Byte, they are all researching this thing.

The classics are easy to get started and difficult to master.

The early identification of yellow was basically a manual review, which was a labor-intensive job.

Later, the number of people online increased, and the content also increased, and the cost of manual review became higher and higher. Therefore, the use of AI+manual methods to identify pornography has become the mainstream.

The AI+manual method is generally to first filter out most of the images that must be normal and must have problems through the machine, and then hand over the rest to manual review, which can greatly reduce labor costs, and the better the machine recognition effect, the lower the cost of manual review. lower.

AI pornographic identification is actually a relatively broad concept, which can be realized through a rule system, such as setting a blacklist database based on MD5, user IP and other information, and directly blocking based on rules.

Most of them will still use the algorithm model, that is, use the algorithm model to judge whether an image contains sq information, which is essentially image recognition.

Image recognition is currently performing even better than humans on some tasks.

The most common image recognition algorithm is the image classification algorithm, from AlexNet to VGG, from ResNet to DenseNet.

The current image classification algorithm can more accurately distinguish the 1000 types of data in ImageNet. Jianhuang itself also classifies the input image, so it is natural to use the image classification algorithm.

Moreover, the target detection algorithm can be used to detect the dew point in the sq image, which is also a relatively reliable means.

In addition, there are features and logic based on the business level, such as whether there are people, the area of ​​the skin, etc., which are used to assist judgment, and are indeed effective in some cases.

The difficulty of AI pornography mainly lies in soft sq without dew point, sq with small features, non-universal sq, cartoon sq and so on.

This is the difficulty of identifying pornography with AI, and it is even more difficult to identify pornography with video and audio.

And for these big Internet companies, even if they can achieve a 90.00% interception rate, the remaining [-]% of the content will be nothing.

Taking Weibo as an example, the data generated every day is in T.

Even with dozens of Ts and a [-]% interception error rate, it's enough to make the whole thing miserable.

And more importantly, Huaguo's content review industry is not limited to pornography, OCR review, a group of Internet giants, has been doing it for a long time.

Click till.

By the same token, the interception rate of 90.00% is unacceptable to them.

So until 2031, it is still the way of AI plus artificial pornography.

It’s just that in 2021, a Weibo-sized company may need thousands of Huang Jian masters, but in 2031 it will only need hundreds of people.

Four digits become three digits.

As for WeChat, every picture and video you send on WeChat will be OCRed once, and then thrown to the back-end review interface.

Use the weight algorithm to judge whether you have violated the rules. After triggering a specific rule, it will determine the weight you prefer. After the weight exceeds the threshold, you will be focused on.

There will be a human to review you.

Of course, this kind of review is not limited to major domestic manufacturers. Fb, ins, youtube, Google and twitter all have a large number of manual review teams.

Their review team is based in the Philippines.

In 2018, PBS released a documentary about this incident.

For the Internet giants at home and abroad, they can achieve close to 98% in text recognition, but in the image classification test, they can only achieve about [-]%.

Moreover, there are very high requirements for computing power, and it is not used in the actual production environment at all.

This is the result of ImageNet's annual image classification test competition. Image and video recognition in actual operation is much more difficult than the ImageNet competition.

After Zheng Li disclosed the AI ​​algorithm, he used the model written in the M language.

From deployment to use, it has broken through the cognition of these Internet companies. An AI model that can achieve a 99.9% recognition success rate for content requires almost the same computing power as before.

That is to say, the scale of AI pornography with hundreds of people in Internet giants has now been compressed by 90.00%.

AI Huang Huang is only the first application of the M language.

The AI ​​technology it represents has greatly improved its advantages in computing power and the difficulty of implementing the algorithm itself.

It allows easy construction of most otherwise complex architectures and can be widely used in industrial fields.

A large number of easy-to-compose modular components, writing your own layer types, computing graph abstraction, data and model parallelism, dynamic type stability, etc. are all perfectly adapted.

After being used by algorithm engineers, M language-related projects on Github became popular in the community.

In the foreign programmer community, the active big cows are complaining, and some big cows who rarely speak are also bubbling:
"Merlin's technology is far beyond our imagination, if he developed the M language himself, then his attainments in the field of AI will surpass all of us.

It's really rare to see an AI programming language without any flaws.

For example, Caffe can transplant Matlab's fast convolutional network implementation to C and C++, which is suitable for feedforward network and image processing, and can train the model without any code.

However, caffe cannot be used in recurrent networks, and its scalability is extremely poor.

The TensorFlow framework launched by Google can generate calculation graphs and then perform automatic differentiation, eliminating the need to manually code when trying new neural network arrangements.

But it runs very slowly and is very prone to errors in large software projects.

Basically, the AI ​​frameworks launched by major Internet giants in recent years will have more or less problems.

Of course this is unavoidable, nothing is perfect.

Just like no matter how many people use Java, and no matter how long they have been active, they can still find shortcomings.

But the M language seems to have no shortcomings in the field of artificial intelligence.

It seems that after countless tests and optimizations, it has reached a kind of perfection.

This makes me seriously doubt that the artificial intelligence that Merlin masters is strong artificial intelligence.

He can realize almost all the fantasies of the blue star people about artificial intelligence.

Including self programming.

It doesn't seem so uncommon for one piece of code to compile another piece of code.

After all, the current artificial intelligence model can already perform simple programming by itself. "

(End of this chapter)