Chapter 1099: Does Artificial Intelligence Really Have to Surpass Human Intelligence?

Alpha Dog is certainly not a dog, but an artificial intelligence robot called "AlphaGo" in English.

Since the advent of electronic science, human beings have had a dream, that is, to develop a robot that can match human wisdom or even surpass human intelligence.

In the early years, there was an artificial intelligence called Deep Blue, which was famous for a while in the big chess championship.

However, even so, Deep Blue at that time could only defeat the chess champion. And when Shenlan is put on Go, Shenlan can't do anything even if he faces a more professional player. Even if the company later developed a "deeper blue", it did not make any achievements in Go.

As a result, for so many years, humans have thought that artificial intelligence is just like that, and he will never be able to surpass humans.

For some scientists, however, they believe that AI may not be able to surpass humans right now. However, in some areas, AI can definitely get a big lead over its opponents.

Alpha Dog is an intelligent robot developed by Gu Ge Company for Go confrontation.

Originally, Gu Ge Company developed this product just to conduct a small-scale Go battle with humans, and then constantly improve the product.

I don't know what wind they smoked, maybe it was influenced by Huaguo's side, and Brother Gu's company suddenly became high-profile.

Not long after, Gu Ge publicly announced to the global media that they had developed an artificial intelligence robot called AlphaGo. In addition, the robot was also put into the online platform to compete with humans in Go.

At the beginning, the global media was skeptical about the development of this artificial intelligence robot by Brother Gu.

They believe that although artificial intelligence is superior in some aspects, it is absolutely impossible to compare with human intelligence.

Not to mention, products like Go that best embody human intelligence.

This is certainly not a casual denial that has not been studied.

You must know that the circumference and horizontality are composed of a total of 19 horizontal and vertical lines interlaced and form 361 points. Each point can appear in three situations: under the sun, under the white, or empty without swinging. Then, at 361 intersections, there is the possibility of 3 to the 361st power.

What is 3 to the power of 361?

Belch...... There is some short-circuit in the author's mind, and it is impossible to calculate how much it is anyway.

In short, this is a very big, very big, very big number, so big that even a supercomputer would be difficult to calculate such a number.

However, do you think that's how Go has changed?

If it's just 3 to the 361st power, that's okay to say.

After all, the development of science and technology for decades, science and technology are accumulating with each passing day, and his development is also extremely terrifying.

No matter how big the 361st power of 3 is, he is a fixed number, and sooner or later, artificial intelligence can surpass this number.

When the AI exceeds this number of words, it means that the AI can beat any Go player.

However, in Go, he is not just 3 to the 361st power.

Because in Go, he still has "robbery".

What is "robbery"?

It's complicated, but to put it simply, it's eating seeds.

When the white son surrounds the black seed, the white child can eat the black child, that is, the other person's child is removed.

When the sunspot completely surrounds the white seed, the black child can eat the white seed.

This is called "robbery".

In this way, the cycle repeats, and the robbery can continue to occur.

When you think that the whole board is about the same, and the changes are about the same, because the opponent can "eat the child", he can vacate a large space. After vacating it, it can be thrown in white and black to compete in this area, and then there are countless changes.

In other words, the variation of Go looks like the theoretical number is 3 to the 361st power.

But in fact, if I really want to say anything, the variations of Go are not 3 to the 361st power, but endless.

How to play endlessly?

Therefore, it is precisely this truth that is understood.

IBM, which developed Deep Blue, gave up this research on Go intelligence.

Unexpectedly, IBM gave up, but another technology giant, Brother Gu, came up.

And it seems that they are still determined to win.

"What a joke, there are endless variations on Go, how can your robot fight humans?"

"I guess Alpha Dog can't even fight a person who plays 3rd dan of Go. ”

"Paragraph 3 should be fine, paragraph 5 is not possible. ”

"As for the sixth dan, the seventh dan, they are not opponents of the same level at all. ”

"Not to mention the nine paragraphs. ”

"I don't know what Brother Gu's company is using to launch this AlphaGo, and it's impossible to really win the human race, hype?"

There has been a lot of discussion in the media.

But to their surprise, AlphaGo is not hype.

In just three days, AlphaGo defeated more than 1,000 players on the Go platform, and jumped to the 9th dan of Go in an instant.

"I grass, I played more than 1,000 games in three days, how can this be played, doesn't he want to eat and drink, doesn't he sleep, is he there 24 hours a day to find his opponent to play chess?"

"Brother, see clearly, people are not people, people are artificial intelligence robots. Robot, he wants a fart to eat and drink, and not to sleep. ”

"Okay, I see. Played 1,235 rounds in three days, won all, and did not lose a single defeat. Nima, don't be so scary. ”

"Even if he is the best chess player in the world, he may not be able to achieve such a record. ”

The media was instantly shocked.

Contrary to what they suspected, AlphaGo didn't know how to play Go, nor was it incapable of fighting humans.

AlphaGo not only knows how to play Go, but is also a master of Go, and even a master of Go.

"Is artificial intelligence really going to surpass human intelligence?"

"Although I have something unacceptable, it seems that it is. ”

"No, I don't agree, I have the ability to find those Go masters in reality to play chess. ”

"Yes, there are some rookies on the Internet, even if it is a total win, this does not represent the wisdom of human beings. ”

Some media outlets, as well as some friends who couldn't accept it, kept shaking their heads.

The online is virtual, what is it to beat the players online, and have the ability to PK with those professional Go players.

In fact, anyone who knows the online Go platform knows.

Although the online Go platform is a platform for Go enthusiasts, in fact, many professional players also play Go online.

Not only are they professional Go players, but last year, several world-class Go masters with 9 dan of Go secretly opened vests on the online platform.

AlphaGo was able to defeat more than 1,000 Go players in three days, and there may be many professional players among them.

It's one thing to just know, and it's another to accept it.

In any case, the online platform is not a formal battle platform.

The real masters are the top professional players in the real world.

Only when AlphaGo defeats those professional masters in reality will they recognize AlphaGo's status.

This is, of course, self-soothing.

However, the company seems to be very aware of the thoughts of the media public.

Yes.

You say that online Pk doesn't count, that's good.

Let's PK with those masters in reality.

The next day, Brother Gu issued an announcement to the world: it will issue a challenge to all 9 sections of the world.