Crossover: 2014

Chapter 257 A Future Worth Expecting

In the absence of abuse, weak artificial intelligence is never the real threat to human beings.

In the final analysis, only strong artificial intelligence can truly threaten human beings.

In appearance, strong artificial intelligence can be similar to humans (share a set of life rules), or it can be very different from humans (form a new set of life rules).

In terms of thought, strong artificial intelligence can share a set of thinking models and moral principles with humans, and can also have a unique reasoning method of its own system, becoming a type of "machine with a soul".

Generally speaking, strong artificial intelligence has the ability of self-awareness, self-learning, and self-determination.

And this is undoubtedly very risky.

Although it is easy to be risky, think about it from another angle.

Would God create a rock that he cannot lift?
In the same way, why do humans create an uncontrollable strong artificial intelligence?

We must know that we only need to restrict and monitor the underlying value orientation of strong artificial intelligence through corresponding rules and memories

A so-called strong artificial intelligence rebellion is out of the question.

Don't think that it is troublesome to design rules for strong artificial intelligence.

In fact, design related to rules is almost full of human life.

Take the charger as an example. An unremarkable charger may have several charging protocols for constraints.

It is not impossible to limit the scope of AI's actions by designing explicit rules.

Speaking of which, people's worries about strong artificial intelligence are more or less unfounded.

Although strong artificial intelligence is the ultimate goal of artificial intelligence development.

But how can strong artificial intelligence be so easy to achieve?

The technical principle that strong artificial intelligence wants to achieve is not complicated, and the corresponding technical indicators are easy to explain.

It is nothing more than self-awareness, self-learning, and self-determination.

Or more generally speaking, it is to let the machine learn from people.

Learn the way people know the world, and then try to use machine learning to perform logical bionics to reproduce the way people know the world.

Everyone understands the truth, but there are various difficulties in implementing it.

In the previous life, the artificial intelligence preached very well.

Anyway, until the eve of Lin Hui's rebirth, people didn't tinker with strong artificial intelligence at that time.

Having said that, since strong artificial intelligence is risky and difficult to achieve.

Why pursue strong artificial intelligence?
The first is that strong artificial intelligence will greatly reduce training costs compared to weak artificial intelligence.

In the previous life, the application of artificial intelligence generally relied on big data for training.

It sounds very tall with the word "big data".

But in fact, it is very hard to operate, and in many cases with the help of big data, it is actually a last resort.

We all know that artificial intelligence is the pursuit of machines that can deal with problems like humans.

But you must know that people learn things from small samples.

Take the small field of image classification as an example. Humans do not need big data at all to classify images.

Accurate classification can be achieved with only a few samples.

When two or three-year-old children begin to understand the world.

Parents do not show their children thousands of photos in order to let their children know animals, but only show their children some animal atlases.

These atlases may still have dozens of pages, and the sample size is actually very small.

To be more specific, if you want children to know what kind of animal is a cat.

At most, show him a few pictures of cats, and tell him what characteristics cats have, and how they differ from other animals like dogs or pigs.

Soon, the child can identify the dog very quickly and accurately.

These are small sample training.

But the artificial intelligence in the previous life wanted to achieve the same function.

However, a large amount of data is required for training.

That is to say, in the usual sense, it depends on big data to feed.

Also take the example just given as an example, the same is to recognize animals, and the neural network is required to learn to achieve the level of children's recognition of animals.

Not to mention the neural network application that is very stretched in time and space.

It is based on the more advanced deep residual neural network in the previous life.

If you want to achieve the distinction between human children and animal images.

It also requires tens of thousands of pictures to carry out relatively sufficient training, in order to obtain more accurate results in practical applications.

The reason why neural network learning "slowly learns" and "learns hard" is that the deep residual neural network does not yet have the ability to think logically, associate and reason.

With the help of strong artificial intelligence, there is no need to worry about this aspect at all.

Not only is traditional neural network learning and training easier.

Even with the help of strong artificial intelligence, it may also help breakthroughs in research in basic disciplines.

These are the reasons why people pursue strong artificial neural networks.

In addition, there are many reasons why researchers pursue strong artificial networks.

It is worth mentioning that.

Some researchers believe that biological humans are inferior.

Humans yearn for a more perfect self.

In many cases, the desire for strong artificial intelligence or even super artificial intelligence is also the pursuit of a more perfect self.

This is not only the limitation of this single neural network, but also the limitation of the weak artificial neural network.

Ideal is full, the reality is very skinny.

People in the previous life did not come up with strong artificial intelligence, let alone the current life where neural network learning is slightly lagging behind in all aspects.

The current scientific research situation is not to say that it is the realization of strong artificial intelligence, but the fact is that even weak artificial intelligence can understand it.

The neural network learning algorithms developed by scientific researchers still lack the advanced functions of human beings such as association, memory, learning, and reasoning.

It is difficult to rely on such an algorithm to develop weak artificial intelligence.

I really don’t know why people still worry about the risks brought by artificial intelligence ethics and machine ethics under such circumstances.

In fact, even if Lin Hui uses the existing information advantages to make a fuss, he can only make a fuss about weak artificial intelligence.

And it is also at a loss for strong artificial intelligence.

Under such circumstances, Lin Hui felt that instead of worrying about these risks and other trivial matters, people nowadays might as well engage in concrete research.

Taking a step back, do you hesitate even if there is a risk?

Is there anything in the world that is completely risk-free?

Lin Hui thinks that artificial intelligence has appropriate risks, but it will help the correct development of artificial intelligence.

To give such an example, the content of the Temple Run game is very similar to most parkour games, overcoming obstacles and traps, and constantly rushing forward.

In the mini game, however, the user controlled character to escaped in an ancient temple in the rainforest, and was chased by a group of monkey-like demon guards in the temple.

The character runs forward automatically, and the player needs to control him to avoid various dangers encountered on the way to escape.

All kinds of dangers along the way make people focus more on parkour itself.

In the same way, when a certain technology has the possibility of incurring certain risks.

Appropriate known risks may instead assist research.

Because researchers may instead consciously stay away from the corresponding risks.

So as to focus more on algorithmic logic and application scenarios.

……

Perhaps this statement is not serious enough.

But Lin Hui always firmly believes that the problems in development must be solved by development.

Development cannot be rejected just because there may be problems in development.

The same is true of developments in artificial intelligence.

Even though artificial intelligence may cause a certain level of controversy.

But Lin Hui also believes that many researchers also hold the same views as him.

At least Lin Hui knew that Eve Carly agreed with him.

It is naturally good to have peers on the way forward in the future.

To some extent, Lin Hui also needs someone to go with him.

In the previous life, under the influence of the media, the hype about artificial intelligence can be said to be swarming.

There are also various claims about artificial intelligence, but they are messy.

But these propaganda generally focus on the promotion of artificial intelligence replacing humans.

But did not understand the true value of artificial intelligence.

The real value of artificial intelligence lies in the transformation of labor and production relations.

This is the reason for its real impact.

This is also the reason why the national level paid so much attention to artificial intelligence in previous lives.

What is the usual form of labor and production relations in the past?

Either reform or revolution.

As something that affects labor and production relations.

The connotation derived from artificial intelligence has an impact no less than a revolution!
In the previous life, artificial intelligence was once regarded as a strategic technology leading a new round of technological revolution and industrial transformation at that time.

In the previous life, artificial intelligence was officially considered to have a "head goose" effect with a strong spillover effect.

The corresponding document directly pointed out that accelerating the development of a new generation of artificial intelligence is an important strategic tool for us to win the initiative in global technological competition.

The previous official statement is as follows:
"Humanity has experienced three rounds of technological revolutions: the steam age, the electrical age, and the information age. Now, the rapid development of a new generation of artificial intelligence technology is setting off the fourth round of technological revolution and pushing human society into an intelligent age.

Think about the significance of the steam engine to the steam age, the significance of electric power technology and internal combustion engine to the electrical age, and the significance of computer and network technology to the information age. We can probably understand the significance of the new generation of artificial intelligence technology to the intelligent age. "

It is not difficult to see that the previous life paid great attention to artificial intelligence technology.

It's no wonder that artificial intelligence was highly valued in previous lives.

In previous lives, artificial intelligence has shown a very strong level in many fields.

In air combat, artificial intelligence can defeat the colonel of the Air Force. This happened in the previous life in 2016. The artificial intelligence system Alpha developed by the University of Cincinnati in the United States drove the third-generation F-15 against Jean, a colonel of the U.S. Air Force/Army with more than 20 years of flying experience. The F-22, the result is that Alpha wins;
Medical care, artificial intelligence can defeat senior doctors, and this also happened in 2016 - IBM's artificial intelligence "Watson" learned a large number of medical papers, and when human doctors were helpless on a patient, it judged within 10 minutes The patient suffered from rare leukemia, and a treatment plan was given;

In Go, artificial intelligence can defeat the world champion. This happened in 2017 in the previous life-Go artificial intelligence AlphaGo defeated the world champion kEJIE 3-0;

Scientific research, artificial intelligence can defeat scientists and accurately predict the advanced structure of proteins. This happened on December 2018, 12 in the previous life. The artificial intelligence AlphaFold launched by DeepMind defeated human experts from various countries in the global protein structure prediction competition CASP. won the championship.

Because of some disturbances in time and space, these things that happened in the previous life may not be able to happen as promised in this life.

But these things in the previous life fully illustrate the great potential of artificial intelligence.

Artificial intelligence is reproduced in this life, because of the characteristics of deep learning and cross-border integration of artificial intelligence.

Also in this life, artificial intelligence is still likely to have a significant and far-reaching impact on economic development, social progress, and international political structure.

In short, the coming new wave of artificial intelligence will not be technological innovation.

To some extent, it is not too much to call the coming new wave of artificial intelligence an alternative revolution.

Such an alternative revolution obviously cannot be pushed by Lin Hui alone.

It's just the words of Lin Hui himself.

Perhaps Lin Hui can continue to carry technology.

However, it is not enough for Lin Hui alone to implement the technology.

If the technology is implemented, it still needs a lot of like-minded people.

Even some technical transfers in the later stage unless Lin Hui has sufficient academic status.

Otherwise, even the handling may need some help to enhance the persuasiveness of the technology itself.

After all, many times when things at the technical level are not done, someone will know the corresponding meaning.

It is also a kind of helplessness to be high and low.

Although these are things for the future.

But Lin Hui believes that this day will eventually come, and even the future is already on the way.

That is a future worth looking forward to.

As for what Eve Carly said.

To be honest, whether it is Eve Carly's strong interest in Lin Hui's supplementary content in the paper or his expectations and worries about the future of artificial intelligence, these are not very beyond Lin Hui's expectations, but Eve Carly's interest in Lin Hui was a little surprised by the guesses made by Lin Hui about the use of her patents.

According to Eve Carly's guess, the reason why Lin Hui acquired her patent "A New Method for Text Judgment, Screening and Comparison" is to use the framework of automatic text summarization (covering content representation, weight calculation, content selection and content organization under the four-part task) to make a fuss about content presentation.

Uh, I have to say that Eve Carly is really a smart person.

Unexpectedly, he was able to understand why Lin Hui made such an acquisition so quickly.

Lin Hui thought it would take a long time for the pursuers behind him to comprehend.

I didn't expect to be partly understood by Eve Carly so quickly.

However, even though Eve Carly guessed it right, she was only partially right, not completely right.

Why did Lin Hui give such an evaluation?
(End of this chapter)