Chapter 2731 Black Box

Zhou Buqi and Larry Page had a good personal relationship, and the two men worked together to promote it. Soon the artificial intelligence teams of both parties began business discussions and cooperation in some specific directions.

One big thing, one small thing.

The big thing is the joint establishment of a project called "AI-Bio", which mainly focuses on many aspects of biological science, including protein structure prediction, disease diagnosis and treatment, drug discovery, etc.

Using artificial intelligence to combine big propositions of this level has actually always existed, twenty or thirty years ago.

But that was simply not possible at that time.

It’s different now. The new “black box” neural network-based machine learning model adopted by the Go software Rokid-go has found the basis for the development of such a grand proposition.

As the father of Rokid-go, Shen Xiangyang simply explained the logic of this "black box model" during the cooperation and exchange between the two parties. It’s not just about taking care of laymen like Zhou Buqi. In fact, even many Google executives with technical backgrounds may not understand what this means if they are not engaged in the field of artificial intelligence.

The concept of artificial intelligence has a history of 50 or 60 years, and related products have been around for 20 or 30 years.

However, previous artificial intelligence products were all guided and formulated by engineers with "explicit rules".

For example, the famous artificial intelligence product "Deep Blue" that defeated chess. The technical logic behind it is actually very simple. It is that engineers implanted the relevant rules of Go into the chess-playing software.

Software understands the rules and relies on the huge computing power of computers to defeat humans.

However, when it comes to the field of Go, this idea no longer works.

Go is far more complex than chess.

Just relying on computing power, with the computing power that humans currently have, it is simply impossible to calculate every change in Go. This requires something deeper than computing power.

Among human players, it’s called thinking and logic.

It would be too difficult for a computer to have "Logic". It would be impossible to allow a computer to obtain this level of computational thinking through the implantation of human rules.

Rokid-go uses a black box operation.

Just put the Rokid-go program in a black box without adding any rules to it. What engineers do is to feed data to Rokid-go in the black box.

As for what happened in the black box, no one knows.

Then, a miracle happened.

Rokid-go in the black box has formed its own understanding of Go through independent learning through a large amount of human Go record data.

I can play Go.

As for how Rokid-go understands Go, no one knows. All I know is that Rokid-go used many moves that humans couldn't understand, and then easily defeated humans at a crushing level.

It is also because of this black box model that there has been huge controversy in the entire artificial intelligence industry.

Among them is Musk.

They believe that the "black box" direction of artificial intelligence technology is very dangerous, because humans have no idea what is happening in the black box and cannot understand what artificial intelligence is doing.

All in all, artificial intelligence is out of human control and may form a "black box conspiracy" in the future, posing a threat to mankind.

However, this type of argument is relatively niche.

At present, the entire industry is very excited about Rokid-go's new ideas and new directions, because it provides humans with a new perspective on discovering the world.

In the past, it was humans who discovered the world.

Now we can rely on artificial intelligence and use the black box model to help humans discover, explore, and understand the world.

Just need to feed the data.

For example, what medicine can treat cancer?

Humanity has been exploring for decades, but has never been able to overcome this difficulty. Humanity's understanding of tumors and drugs is not enough. If relevant data is fed to artificial intelligence, perhaps artificial intelligence can give the best drug to eliminate malignant tumors.

This is a large field that affects all mankind.

Whether you make money or not is secondary.

Even if Google and Ziweixing International are competitors, they are willing to cooperate in this field and promise to open source relevant research and development results and share them with all mankind.

This is a big deal. As for small things, they have certain commercial attributes.

Google will join forces with Ziweixing International to launch an artificial intelligence "World Cup" for developers around the world every year, just like the Football World Cup.

Every year, scientists from both sides will discuss, research, and select three tasks in the field of artificial intelligence to publish.

Contestants have 6-8 months to complete the task.

Whoever comes up with the best solution and the best effect will be the final winner. Google and Ziweixing International will each provide US$10 million as prize money for such events.

The tasks released here will definitely not be like the Mathematical Olympiad competition, where you set questions to test others.

But some directional exploration.

Artificial intelligence is like a boundless world, but this world is dark and no one can see clearly. Occasionally, a few top scientists have explored some areas and published some directional papers.

These papers are like faint beacons in the distant dark world of artificial intelligence.

The light is very weak.

Not sure if that direction is a dead end.

Who is going to explore this lighthouse?

This kind of exploratory projects are all groundbreaking. It doesn't mean who has the best technology to create a successful project. What is needed is a "thinking warrior" to charge forward.

If we talk about the AI ​​experts with the strongest technical capabilities, they must be in large companies such as Google, Microsoft, and Ziweixing International, but they are not necessarily "warriors."

Many times, salary, qualifications, age, position and industry status are all baggage.

This type of warriors are mostly college students, master's students, doctoral students in some universities, or small teams or companies, who can move forward without worries without performance pressure or KPIs.

This is the significance of Zhou Buqi and Larry Page joining forces to create this artificial intelligence "World Cup".

To overcome the shortcomings of large companies.

The technical level of the "warriors" may not be very good, but they are more innovative and can find better angles and entry points to find the way to the lighthouse ahead.

With the road, top engineers from technology giants such as Google, Purple Star International, Microsoft, Amazon, Facebook, and Oracle will be able to put their talents to use.

Then you can send out elite soldiers and generals to rush forward.

It's like leading an army into a war.

If you can't charge with a large army when you come up, you must first send exploration horses to determine the direction; then send vanguards to find out the true situation. When clear dynamics are grasped, the army will exert its strength and the main force will rush forward.

The domestic science and technology industry is engaged in "following the trend" projects, which is too easy. They are all doing scientific research in the "light".

But in the current field of artificial intelligence, Ziweixing International is leading the development of the industry. There is no trend ahead to follow, and we have to find hope in the "darkness".

This is much more complicated.

Throwing money alone is not enough, nor is talent alone. You have to mobilize global industry resources. Only Silicon Valley giants have the ability to mobilize global resources.

What kind of technology company in another country wants to mobilize global resources... They won't pay attention to you at all, so it is more practical to just follow the trend quietly.

…

I went to Google in the morning and went home for lunch.

Prepare to go to PayPal in the afternoon.

Ning Lu and Yan Qiqi served as personal assistants all morning and followed the whole process. Yan Qiqi was deeply shocked. Unexpectedly, Boss Zhou really cared about the affairs of all mankind, so she said with admiration: "Senior, do you want to pass the exam?" Is technology going to change the course of humanity?”

Zhou Buqi laughed.

Yan Qiqi tilted her head and said: "I think that Musk who is going to Mars is so funny. I think he is far behind you. Your ambition is the greatest."

Rao Shizhou Buqi was thick-skinned, but he couldn't bear it. He simply said: "Only top projects can have top talents. Only with top talents can you participate in top competition."

"ah?"

Yan Qiqi was confused.

Zhou Buqi hummed: "Do you think these companies in Silicon Valley are all for technological development? To change the world? Musk is a psychopath, that's why he is special. 99.99% of Silicon Valley companies are just to make money, unlike domestic companies. Those companies are no different.”

Yan Qiqi said weakly: "Is there... a difference? These companies in Silicon Valley obviously invest more in technology."

Zhou Buqi said: "That is determined by the general environment. It does not mean that they are superior to others. Again, only top talents are qualified to participate in top competition. If you do not invest large amounts of money in technological research and development, you will not be able to attract people. If you bring in top talent, it will cause damage to the business side. At the end of the day, it’s still business.”

Yan Qiqi frowned slightly, "What about at home?"

Zhou Buqi said: "It's different in China. It's... more complicated in China. Even if you launch top-level projects in China, you won't be able to attract world-class top talents. Even those top talents who go out from China are not willing to do so." Go back. If domestic technology companies are like these in Silicon Valley, whoever does not engage in scientific research will not be able to recruit people and will die. Do you think they can do it? It’s hard for a clever woman to make a meal without rice. There is no need to promote these technology companies in Silicon Valley. We cannot belittle domestic technology companies. This is all determined by the general environment. If we really let them go, everyone’s environment would be the same...it’s not necessarily clear who is stronger and who is weaker!”

Yan Qiqi suddenly realized: "So cooperation with Google is actually for long-term business interests. It is not to make money with technology, but to fool everyone into joining Ziweixing International."

Zhou Buqi coughed and said frankly: "I can't say that. If I can really change the world and benefit all mankind, then of course it would be the best. I can't let the whole world point at my spine and say that I They just make money, pick up girls and enjoy life, and they don’t know how to make long-term efforts for the future development of all mankind.”

Yan Qiqi laughed heartlessly, "If you pick up more girls and have more children, you will also contribute to human reproduction."

Zhou Buqi was in awe, "High! Very high awareness!"