"PHC is not just a product, it is a new field, a new blue ocean ecology, and its ability is enough to breed the technology giants of the next decade, and in the next ten years, one unicorn company after another, and even super unicorn companies will be born here."

"We, Coastline Company, will be committed to building and enriching this closed-loop ecological chain ...... based on PHC."

At the press conference, when Ye Hua talked about the story of "closed-loop ecological chain", all those who wanted to enter PHC suddenly sobered up, and then their mood became a love-hate combination.

Sure enough!

When anyone in the mixed technology industry, especially those in the mixed Internet technology circle, heard the "closed-loop ecological chain", they all immediately thought of the closed-loop ecosystem created by Jobs for Apple.

This means that Shoreline will also want to build a PHC-Store-like app store, which will also mean that as long as the content is launched on the PHC platform, the other party will get a commission.

It's the same as Apple.,It's the rhythm of lying on countless third-party developers to make money.,It can be said that the third party loves and hates this closed-loop ecosystem.,By the way, I also complained about a wave of Qiao gang leaders.,He is the leader.,If it weren't for the deceased's respect.,I really want to scold the Qiao gang leader together.。

However, scolding is scolding, and at the same time scolding, companies that want to build a closed-loop ecology grab a lot of money, because everyone knows that establishing a closed-loop ecology can build a solid moat and make a lot of money.

Ye Hua's resolution is certainly scolded, but as long as you can make money, it's not a big problem, Apple has made a good start, and the third party is willing to accept it, provided that you can make money.

Ye Hua did not emphasize the story of "ecology", it was enough to mention it in a few words, because this is for a third party, and these people all understand the Tao and do not need to be repeated.

And Ye Hua knows that the key lies in users and sales, you can't sell PHC, the story of "ecology" is all over the place, and those third parties don't bird you.

But PHC is sold, forming a user scale, which means that you have a huge infrastructure, and if you don't talk about it, those third parties will naturally flock to it, after all, the world is all for profit.

Therefore, the core is still users and sales, and the final consumer is the premise of everything.

Ye Hua always divided the weight issue very clearly, paused for a while, walked a few steps on the stage and continued: "If you compare it to a life, PHC is the backbone, and the holographic display is the face image, but a complete life seems to lack an inner soul ......"

As soon as these words came out, they also aroused the curiosity of many people, and people also heard the implication, which is a hint that there is a heavyweight product?

Suddenly, Ye Hua said: "Powerful AI is the key to giving PHC its soul. ”

Many people in the industry were a little surprised, the coastline company is actually developing artificial intelligence? However, it was soon relieved that the current technology companies were embarrassed to say that they were a technology company if they did not have anything to do with artificial intelligence.

It's no surprise that a company like Coastline is developing artificial intelligence.

But the surprise is that they were able to launch a full-fledged AI application so quickly. You know, the people present know very well that Coastline Company's press conference today is to launch a series of products, and this technology development ability is too terrifying!

But more, what everyone is looking forward to is how Ye Hua's achievements in the field of AI are, the cowhide has now begun to blow, but in the end, whether it can surprise the world like holographic technology is the most concerned.

"Before announcing Shoreline's AI product, and before evaluating it, we should perhaps understand what kind of development the global AI field is at present."

At this moment, Ye Hua's voice resounded at the press conference, and he paced calmly on the stage, and said eloquently:

"Maybe through various information channels, we heard that artificial intelligence is about to surpass people, and the topic of artificial intelligence in today's era is hotter than ever, and even says that it will replace the work of countless people, and many people are concerned about the development of artificial intelligence, but they are not because they feel how excited they are about the development of science and technology, and the real starting point is to start with an anxious mentality."

"I believe that everyone is suddenly so concerned about the field of artificial intelligence, and more concerned about whether their jobs will be robbed by artificial intelligence in the foreseeable future."

The melon-eating netizens who followed the live broadcast of the press conference and the fans on the scene nodded silently in their hearts, this is indeed the focus of many people's concerns, after all, jobs are very important.

Ye Hua continued:

"So how far has artificial intelligence developed in today's world? It seems that it has become extremely fast, and AI can already do most of the work, etc., but if you ask the experts in the industry, the answer is the opposite, they are concerned about the fact that AI is still doing too little, not too much. ”

"As for the philosophical issues that discuss the replacement of humans by artificial intelligence, the elimination of human beings, and so on, I can say with certainty that none of these people are really experts in the study of artificial intelligence."

As soon as these words came out, the director playfully gave the live broadcast camera to Elon Musk who was in the front row of the seat area, and a close-up shot of about three seconds stayed on Musk, who stared at the stage calmly, and his expression couldn't see what the big guy's inner fluctuations were at this time.

Some technology product enthusiasts were watching the live broadcast, and when they saw this close-up, they thought of what Ye Hua had just said on the stage, and they understood it in an instant.

Because Elon Musk is one of the celebrities who has voiced concerns about artificial intelligence.

At this time, give such a shot and broadcast Li Shizhen's skin.

The barrage is also a variety of skins.

Silicon Valley Iron Man also wants to save face, okay?

However, Elon · Musk is indeed not an expert in AI research, he is good at playing rocket recycling technology and electric vehicles, and AI is really not his specialty.

At the same time, Ye Hua on the stage said:

"One thing I have to complain about is that some media and even some mainstream media have now exaggerated the propaganda of artificial intelligence to the extent that those of us who are engaged in AI are vomiting blood."

There was a burst of low-pitched laughter at the scene.

"For example, I see that the person who is often mentioned in the media is the late Mr. Hawking, and there is no doubt that Mr. Hawking is a great scholar, so what does he do? He is a Ph.D. in physics, an authoritative expert in the fields of black holes and cosmology, but he has never been engaged in artificial intelligence, and Mr. Hawking has suffered from ALS for many years, so he can't do scientific research in the first place. ”

"As a result, the media regarded him as an expert in the field of artificial intelligence, and asked him to come out and talk about the dangers of artificial intelligence, which is a very funny thing in itself, and it is inevitably suspected of hyping up hot spots, in my opinion, it is rubbing the popularity of AI."

So why should the opinions of some non-specialists be taken so important, even as classics and truth? You should look at those senior experts who really study artificial intelligence, then you will find that they have a kind of caution and fear, including I also think that the current media talks too much about artificial intelligence. ”

"The development of artificial intelligence can be traced back to the 50s of the last century, which is a very tortuous history of development, and it is also very helpless, because everyone is in a hurry, and it will not take long to disperse, and every time there are high hopes for it, and then a large number of people enter this field, including a large amount of capital, and finally find that there is no breakthrough and immediately run away."

"In recent years, people have flocked to it again, and now any company feels that as long as it is related to artificial intelligence, it will rise tall, and its market value will double or multiple, which is very dreamy and incredible."

"So what's the state of AI right now? Most people study artificial intelligence, I think it should be said that it is an application, all day long the media say that artificial intelligence does this and that, plays chess and wins, games and wins, these are actually a relatively simple technical route in the eyes of industry insiders. ”

"To put it simply, it's about collecting big data and labeling data. For example, if you are engaged in image recognition technology, you can train with a large amount of data, and then do a job equivalent to fitting, a multivariate function, you don't know what the structure of this function looks like, but it doesn't matter, you have big data fitting, which is actually just the idea of a neural network. ”

"Finally, if you fit the result and give a good output to an input, you can identify what the figure is, and the recognition ability is even higher than that of a person."

"Of course, this is good, but the premise is that you have to have data support, and it is supported by massive big data, which is actually a relatively low-level technical route for artificial intelligence research, and building a large-scale data center is very, very energy-intensive."

"And now there is a more advanced route, that is, the kind of route of Alpha-GO, but it is still inseparable from big data, but the difference is that this data is not manually labeled, but let the AI play with the AI itself, and then let it automatically generate data, which is why Alpha-GO became so strong later."

"It started out learning human chess games, and then it became Alpha-Zero when it leveled up again. Why is it called Alpha Zero? Because it doesn't need to learn human chess games, the only thing he inputs to it from scratch is the rules of Go, and then plays against himself within the framework of this rule, and he can quickly accumulate data and quickly find the best way to play Go, so as to surpass humans. ”

"So this is a relatively advanced technical route, for research and development capabilities also have a lot of requirements, especially for computing power and algorithm requirements, the Alpha-GO team's biggest contribution is to prove that this road can be taken, so it is a big sign of progress in the development of artificial intelligence, and this directly leads to the fact that many industries have now refreshed the technology, to do things that could not be done before, such as image recognition technology, the application of face recognition has brought great commercial value."

"But the route still seems to me too narrow and far from good. In my opinion, all the artificial intelligence on the market today is not intelligent, and I know that what I say will definitely cause controversy, so it depends on how to define intelligence. ”

"I mean, the only artificial intelligence application on the market today is data statistics and analysis, you give it big data, and then he learns from it, so as to get a set of big data based on it It does better than people, and the so-called big data deep learning is what it means."

But is this really how we humans learn a knowledge? I don't think it's at all, for example, parrots and crows, as everyone knows, these are two kinds of birds, about the same size. Parrots are smart, right, but I would say that parrots are much more mentally retarded than crows. ”

"Why? It seems that the parrot looks very smart, and it learns after you say a few words in front of it and repeat it a few times, and the parrot learns to imitate human speech. But if you try to talk to the parrot, you will find that it doesn't know what it's talking about, because it can't understand what it's saying, it's just a mechanical repetition. ”

"Similarly, many so-called chatbots, chatbots of artificial intelligence applications that everyone usually comes into contact with, there are also many videos and screenshots of conversations with chatbots on the Internet, which are amazing at first glance, such as Microsoft Cortana, Xiaomi's Xiaoai...... Well, I'm not specifically targeting any AI product, I mean all the AI on the market is the same (spicy chicken). ”

It caused a burst of laughter again, and the audience subconsciously helped to add the word "spicy chicken" in their hearts, so Ye Hua dared to say it, and all the AI on the market was denied at once, and the media on the scene were most happy to see such a thing happen.

This is what the media loves to do.

And Ye Hua continued:

"Why? First of all, has it passed the Turing test? As for the brainstorming on the Internet to say that they deliberately did not pass the Turing test, it is pure nonsense, the reason is very simple, the upper limit is there, you can't say that a three-year-old child understands calculus, he can't understand it, just say that he is deliberate, okay, this brain hole rumor has been broken, so is it self-aware? Obviously, no! ”

"I remember that the Saudis granted citizenship to Sophia, an artificial intelligence developed by the British, which looked shocking and caused a heated discussion in the media at the time. But in fact, you really can't talk to it much, and if you talk too much, you will find that it is actually just talking back and forth, just a few words back and forth, and you may find that it doesn't actually know what it is talking about, just like parroting. ”

"Then why are crows clever? A real-life example, I'm talking about this crow is a real-life crow, it was filmed. ”

At this time, a huge screen was projected at the press conference, playing a record of the old age, and the scum picture was in quality.

"The crow in the picture, it gets a tall nut and wants to eat the fruit inside, but it can't peck the husk, what should I do?"

"So it made an observation and found that the high nut landed on the road, and the wheels of the car came out as soon as it was pressed, OK, the crow made this observation, and at the same time it realized that such cars often can't eat the fruit on the road, because it's too dangerous."

"Then it made a second observation and found that there was a traffic light on the road, and the car would not go when the light was red, so it was safe to eat the fruit."

"At the same time, the crow made a corollary that what was observed at this traffic light intersection would also pass at another traffic light."

"Look, how clever a crow is this, and one might see this example and say that it is showing true intelligence. But the paradigm of the crow doing things is exactly the opposite of the situation we usually understand about artificial intelligence, artificial intelligence uses big data analysis to do small tasks, and the crow uses small data to do big tasks, which is completely opposite. ”

"What data did the crow get? It has only observed a handful of examples, because it only has one life, and it can't run over and test it on its own, otherwise it would have died N times long ago. ”

"Therefore, real artificial intelligence can extract the internal rules through very small and limited data, and finally evolve a very reliable execution paradigm that can be deduced immediately."

"Why is this possible? Because it has a causal relationship, and traditional AI is just a statistic and correlational data analysis. ”

"Another chestnut, the rooster crows and the sun rises, the two are a correlation relationship, but it is not a causal relationship, it is not that the sun rises because the rooster crows, in fact, we all know that the sun rises every day as usual whether it crows or not."

"So there's a core issue here, which is that causal logic is not the same as correlated logic, and that's important."

"Otherwise, if you go to image recognition, the root of recognition is its function, for example, people can see at a glance that this house is a kitchen, a study, a bedroom, etc., because the first thing we think of is what the function of this place is for, so that we can immediately see many features in the image. For example, people can easily judge a modern study room and an ancient study room, although it looks very different in appearance, but if you use the current computer or artificial intelligence to identify, you will not think that these are two identical things, it can't be recognized, but people can accurately judge at a glance, this is the study, because its function is the same. ”

"That's because people are able to analyze the causal relationship behind it."

"The study of artificial intelligence can not ignore this point, so if you want to do artificial intelligence, I think you should consider from two points, one is the laws of physics, that is, the laws of nature in the physical world, such as how an object can be stable and what kind of support is needed; The second is human motivation, which is used to drive people to think, in my own words, i.e., ...... Self-emotional logical analysis! ”

"And the AI research and development of our coastline company has taken such a route and has made revolutionary progress......"

In other words, Ye Hua is implying that he has really created artificial intelligence in the true sense? As soon as these words came out, it was tantamount to a depth charge being detonated, and all the insiders at the scene were shocked.

……