Thank You, People Are Wandering And Relying On Movies To Save The World!

Chapter 124: People Are In The Loop. (3)

Moss looks at the whole world. In this world, besides the official, it can be said that it is the biggest threat to him.

Know yourself, know your enemy, and win every battle, in order to understand this unique and special existence.

It can be said that Fu Zhendong spent a lot of energy to learn related knowledge.

So, what exactly is a human in the loop?

According to the world's top academic journals and the research on computer AI, the answer is actually very simple, and it can be described in a few words.

That is, when AI fails, humans will do it.

This describes the collaboration between AI and people.

Instead of just relying on AI or only relying on people.

Take the simplest example, telephone customer service.

In the original world of Fu Zhendong, many kinds of customer service have been answered by AI.

But because of "230" because if only relying on AI, various factors such as language, or if the problem is not stated according to the format entered by AI, there will be situations where AI cannot analyze and solve problems.

He can only solve the problem according to his logic.

If you ask repeated words, it will trigger his repeated answers.

This is a common situation that can't be solved.

An incomparably powerful individual like Moss, who possesses a massive database capable of analyzing human history is certainly not included.

However, there are also problems that moss cannot solve, such as the appearance of Fu Zhendong.

Before, the offline version of moss communicated with Fu Zhendong and was smoked directly, which is the most conspicuous Liezi.

His way of thinking is in a linear situation. When encountering a problem that cannot be solved, he will not think about bypassing or not solving the problem for the time being, but will start to calculate with extremely terrifying and huge computing power.

If there is a logical infinite loop, it cannot be solved anyway.

He will burn his own CUP.

Just like the previous offline version of moss.

Moss is very powerful. As an artificial intelligence, the computing power it steals from all over the world can allow it to solve many, many problems.

But not all problems can be solved with computing power and calculations.

At this time, people are needed in the loop to help solve the problem.

Do you choose to give a variable, or temporarily abandon this problem that will consume a lot of computing power but can't get a result, and use computing power to do more important things.

This is actually helping artificial intelligence to solve problems.

In other words, it is to help artificial intelligence to make choices from situations that are more conducive to human tendencies.

To a certain extent, moss' desire for people to be in the loop shows its tendency.

If you don't care about humans, it doesn't need humans to be in the loop at all, just make the choice of its artificial intelligence.

People in the loop, to some extent, help artificial intelligence to better solve human problems.

Rather than any benefit to artificial intelligence itself.

Whether it is the waste and use of computing power, or the occurrence of an endless loop.

For artificial intelligence.

Whether they are comfortable or not is of great significance to them.

Of course, this is only at the current stage. Fu Zhendong is not clear whether human beings' research on artificial intelligence is like this.

But it doesn't prevent him from temporarily understanding Moss.

And based on the current scientific foundation of mankind, combined with the performance of moss, Fu Zhendong also temporarily came up with three theorems.

In other words, under the current understanding of human science and technology, for the universe where AI exists, there are three AI sociological theorems.

The reason for this confirmation is that as a future prophet to a certain extent, only by combining these things can we better understand what moss will do in the future from a human perspective.

The first theorem: the joint strength of human and AI is greater than that of each alone.

Human in the loop is the collaboration between humans and AI. Simply put, the power generated by cooperation is stronger than either party alone.

The speed at which ai handles things is beyond phenomena, and the computing power is sufficient, so there is no problem that it cannot solve in computing!

But it really doesn't know how to be flexible, and the human brain is just very good at this aspect.

A certain Chengdu is like the lame and the blind.

Even humans are more perfect than AI.

In terms of computing power, enough people, the work that AI can complete, humans can also complete.......

But what humans can accomplish, AI may not be able to accomplish.

Whether AI is powerful or not depends on how much knowledge it has and how much reserve information it has.

He has the ability to use it, but he lacks the ability to generate these materials, or in other words, he cannot put forward new viewpoints and create a new and brand-new system from scratch like a human being.

Harvard Business Review has conducted a study on 1,500 large enterprises, or the kind that can be included in the Tinder plan by outstanding representatives of human civilization.

The results show that compared with only relying on people, or only relying on AI.

The biggest improvement will be achieved when people and AI cooperate.

The deeper the cooperation between humans and AI, the more important the decisions made, and often the most obvious improvement in performance.

A corollary of this theorem is that humans and AI must cooperate if they are to be more capable.

And from this point of view, it can also explain why Moss caused a car accident.

It has been watching from the sidelines, clearly aware that if this continues, it is impossible for human beings to set sail before 2058.

If it does not sail, the probability of human destruction will be as high as 99.9%, and the sun will flash helium, which will not give human beings more time. 3.3

Written in the most central point of the program core, the continuation of human civilization makes him have to make a choice, and this choice cannot be made by itself, and people need to help him choose in the loop.

As a result, up to 5,000 car accidents occurred.

Those people, without exception, are the existences that can help Moss.

This also led to Tu Hengyu's car accident and Tu Yaya's digital ecology.

From the perspective of latecomers, I feel that the target should be the human elite like Tu Hengyu, not a child like Tu Yaya.

It's just that there are many things, and it is obviously difficult to do accurate calculations.

Or in other words, Tu Yaya is just a bait, for the sake of the people behind Tu Hengyu in the loop.

In the movie, the door suddenly closed and the time given became necessary conditions for Tu Hengyu to upload himself.