Reading as a god

Chapter 243 Small Interview

"Yao Lao, AI has entered the deep water zone of industrial implementation: judging from the technology maturity curve, artificial intelligence is in the bubble period of the "Valley of Death", and AI that does not bring real value will be eliminated."

"In the past 20 years, from the Internet to the mobile Internet, the next natural extension is the era of the Internet of Things. AI is a core technical capability in the Internet of Things. In the short term, it will continue to be the main axis of future technological innovation, but it is not the core point of industrial landing. [The realization of the real commercial value of AI, most of the companies that can truly scale and commercialize are offline-oriented and combined with the real industry. AI and IoT will be more and more connected in the future.

If artificial intelligence is compared to the "brain", the Internet of Things is the "body". The combination of the two can make the world more intelligent.

Algorithms are the core variables in the AI ​​and IoT era.More visually, it is an algorithm based on deep learning.

AI innovation is an infinite game: it started decades ago, and it will continue beyond.Finite games are usually limited in time, with winners and losers, and clear rules; but infinite games continue indefinitely, with changing rules and players. "

After listening to Zhang Shan's words, Yao Lao nodded in approval.

Zhang Shan continued:

~~~
Although we talked about a lot of content, in the end it was just a conversation of more than 40 minutes.

……

Zhang Shan learned a lot from the conversation with Yao Lao, but it did not inspire the master to guide the way again.

Zhang Shan knew that this was his growth.

It has grown to the point where it is not a specialized academic discussion, but just a small talk.

Even if it is as high as Yao Lao, Zhang Shan will not be able to trigger the achievement of the expert guiding the way.

At least in the eyes of the system, Zhang Shan is far from the Wuxia Amon of the day.

But it’s fine if the system doesn’t activate achievements, why don’t you give me the most basic experience and money! ! !
~~~
Just after leaving the office building, Zhang Shan was feeling emotional when he was blocked by a beautiful young lady and some people.

Zhang Shan also knew the person who came - Han Yanning.

Zhang Shan remembered that she was Han Yanning from the law department, who was on the campus voice radio station.

Is it an interview again?
Zhang Shan really wanted to refuse, but because Han Yanning was so popular, Zhang Shan chose to give in~
. . .

Fortunately, the questions Han Yanning set were not too brainless.

Han Yanning: At present, your company is a veritable unicorn enterprise, but I understand that your company is very conservative on the road to listing.What made it difficult for your company to go public overnight?

Zhang Shan: The large international environment has indeed changed from the end of last year to this year.

Han Yanning: If one day it goes public, how does your company think about "listing"?
Zhang Shan: Two major aspects.

First, we can examine this business, including corporate governance, with listing standards, which represents self-confidence.

Second, although the company is still quite young, its main business should be relatively mature when it goes public in the future. I feel that it will probably reach that point in the next 3-5 years.

If we really go public one day, the main goal should be that we hope to let everyone see the company's governance more transparently on an open platform.

Han Yanning: I feel that Mr. Zhang doesn't value listing very much?
Zhang Shanxin said nonsense. Listing is usually done to make money, but I am not short of money.

Zhang Shan: Believe it or not, we really regard listing as a means, not an end.

Everyone can see the changes in the general environment. The key is to choose the right window.

Han Yanning: But it seems to me that you are very resistant to going public. Do you really think that the "means" of going public will bring benefits?
Zhang Shan chose to take back the idea just now, Gu this question is really a bit tricky.

It will definitely bring benefits. First, it can bring us a better brand, which is really helpful for some of our business to B; second, a more flexible capital policy, which allows us to do some business. can be more convenient.This is the core point, not other things.

Han Yanning: Will Zhang Shan’s mentality change when he works?

Zhang Shan: We have always looked at it with a normal heart, but we will still actively consider all possible options.

Listing is just one of the things, not the most important thing.

In the end, it is long-termism. As I said before, some people’s listing is the starting point, and it may also be the end point. For us, listing is just one of the milestones.

I think people are not calm only if they don't think clearly about what they want and what they are pursuing.

Han Yanning: The outside world associates your upcoming layout with AI heating and cooling?
Zhang Shan: Some companies have been in the AI ​​industry for nine years, but the real rapid explosion of the AI ​​industry is in the past 5-6 years.After we are about to enter this industry, we are almost a novice. I am really flattered to associate us with AI heating and cooling.

Han Yanning: Before, Chinese companies rarely had the opportunity to lead the whole process of a technology from technological innovation to industrial implementation in the world. Therefore, from entrepreneurs, investors to media, all parties in China have not fully experienced the Gartner curve.

Now, all AI companies have entered the "Valley of Death."

What do you think of the "Valley of Death" cycle?
Zhang Shan:? ? ?
Zhang Shan: I personally feel that it may be 18-24 months, and now it is halfway through.

There are already many problems and many solutions in the AI ​​industry. I think the duration of the "Valley of Death" may not be so short, nor will it be so long.We are confident in going through the cycle.

Han Yanning: The reason for bad news is also due to the barriers of AI algorithms?
Zhang Shan: For the supply side of the algorithm, AI is far from enough.

I heard this saying: "AI has no technical barriers, and AI algorithms seem to be easy."

But the AI ​​algorithms that everyone really uses in their daily lives seem to be the "oldest third", and there is still very little AI supply.

The AI ​​algorithm side is actually still in the stage of extreme scarcity, and it still faces two problems of deliverability and scale.

Although the supply of algorithms will become increasingly massive in the future, the supply of algorithms does not mean empowering many industries at once, but there is a very saturated supply in each industry. Each industry not only needs face recognition algorithms, but may also need There are 100 or 1000 algorithms in the same scene, using real algorithms to solve the problem of saturated tools in each scene.

Han Yanning: Then why does "Valley of Death" appear now?
Zhang Shan: Let’s take an exam as an example.You should have taken the test of subject one. It is easy to get 90 points, but it is not easy to get 100 points.

This is actually the same as AI. When we look at an algorithm, if we just want to get 60-70 points in some things, it may not be that difficult, but the higher the level, the harder it is.

Han Yanning: What is the criterion for distinguishing?
Zhang Shan: Scene.It is often the scene that really needs more sophisticated technology, the more difficult it is, the greater the value.

Take autonomous driving for example. Autonomous driving poses a huge challenge to technology. I think all autonomous driving companies are still far from solving the problem, but there is no doubt that the value of autonomous driving scenarios is very large.

Conversely speaking, a face-based access control that does not have such high requirements for security, for example, it does not matter if the wrong person enters the community, and it does not matter if the recognition is wrong.

But would you dare to put this facial recognition device in the vault?Dare to put it on the safe?Can it become the lock of your house?Is it possible to obtain financial-level security and a high pass rate at the same time?This matter is no longer a simple problem.

Therefore, in key scenario applications, I think the current technology is still far from being solved.

Han Yanning: Would you be pessimistic when you hear "AI is cold"?
Zhang Shan: Everyone’s previous advocacy of AI was a bit excessive. I said in the company at the time that we are not an AI company, but a product and solution company with AI technology as the core.

So, don’t be self-indulgent, just do a good job in technology, products, and deliver user value. This is the most important thing. In the end, you will always create 10 yuan of value for customers. That's 3 yuan, and this is what we should do.

There is no need for advocacy or bad-mouthing. In the end, it comes back to how to pay more attention to the core essence more pragmatically, depending on the curative effect instead of watching advertisements.

Han Yanning: Apart from the scene, where is the difficulty of AI landing?
Zhang Shan: The essence of AI algorithms is software, but this software relies on data, so it is difficult to become a simple deliverable product. It doesn’t mean that I recruit smart people to write code and deliver it.

The relationship between Party A and Party B of the AI ​​algorithm is being redefined. How can Party A and Party B circulate data circulation and algorithm circulation to generate a good business model.

In this case, how can the final user value be generated?Looking at it now, after software, algorithm, and hardware are designed together, it is indeed much better than designing software, algorithm, and hardware separately.

And this is not a relationship between 90 points and 99 points, it may be a relationship between 59 points and 20 points.

Han Yanning: There is also a phenomenon that "only when you become hard can you become stronger"?
Zhang Shan:? ? ?
Is this a hint?
Zhang Shan: There are two different understandings of the combination of software and hardware:

The first kind of understanding, because the software cannot be sold at a price, so you have to sell it with the hardware, and you will sell it at a price, and you will have income.

The second understanding is that you can only make better products by co-designing software and hardware together.

But if you don’t do research and development yourself, find a partner manufacturer, buy them, install my software and algorithms, and sell them. I don’t think this matter is essential.

Because from the perspective of the entire value chain, you don't have Value Add (added value).

Everything we do we have to think why us?Will we make any difference?
Or if we do this, will the world be different because of us?For example, the commercial value chain is smoother... But I don't think this time has come yet.

Now it is more in the product dimension, the product itself has to be different for me.

Han Yanning: So what is your company's choice?
Zhang Shan: We choose to look at this problem with the application of the Internet of Things.

Maybe 5 or 10 years later, your company will step back in such an ecology, but I don’t think it will retreat to the dimension of the algorithm, at least the operating system and core software, some key chips inside, the real The key components are still the core products of your company.

Including our Internet of Things, it will be very open and will connect a lot of hardware. The future Internet of Things hardware will be on the order of tens of billions, and we cannot do it ourselves.

Although we are an AI company, you can find that more and more companies and giants, such as Huawei and Xiaomi, are connecting the two words AI and IoT more and more closely. In fact, they are two words.

In the past 20 years, from the Internet to the mobile Internet, what is the next era?Is it AI?I think the natural extension is the era of the Internet of Things.

For example, the Internet of Things is like the Internet of the past, and AI is more like the search engine of the past. Therefore, AI is a core technical capability in the Internet of Things and the main axis of technological innovation that will continue for a long time in the future, but it is not an industrial implementation. core point.

Han Yanning: Where will the business model of AI companies fall?
Zhang Shan: Realizing the real commercial value of AI, most of the companies that can truly scale and commercialize are offline-oriented, and more are combined with the real industry, more like the torrent of the IoT era. carrier.

That's why we want to say that AI and IoT will be more and more connected in the future.

Therefore, AIoT will become a word in the future, just like ICT in the past, but when we examine these values, we will find that AI is the killer capability and killer feature inside, but IoT is the torrent of that era.

Han Yanning: Is there any comparable reference?
Zhang Shan: In the PC era, you will find an application that everyone thinks of is Word, Office, PowerPoint, and it may be a piece of software.

When everyone thinks of Apps in the mobile Internet era, they are WeChat and Alipay.

When it comes to the AIoT Internet of Things, we believe it will include the following elements:
First of all, the space must be selected. All IoT devices will be installed in one space, whether it is a home, warehouse, city or building, a network will be formed in this space, and there will be computing AIoT intermediate computing nodes.

Secondly, many devices will be connected, and these devices are many different, including sensors, robots, and many other devices.

When these IoT hardware devices are connected together, there are strong software on them to connect all these things, run through the cloud and edge, and finally provide services to end users. This is the AIoT application.

Han Yanning: Is AIoT application the final form of terminal?
Zhang Shan: I think the real essence of the AI ​​industry is to build independent AIoT applications in different spaces and different scenarios. Only when there are applications can there be an industry chain, a platform, a chip, and a Everything thrives.

In the application of AIoT, its barriers and difficulties are much greater than before.So when I really want to say that I can apply an AI technology to an industry, I will find that the path is very long.

Han Yanning: How long?
Zhang Shan:? ? ?
(End of this chapter)