What can AI Internet bring?

Recently, because of entrepreneurship and Dr. Chen Xiaoliang (founder of Acoustic Technology, acoustics, speech recognition experts), the future is discussed more, and I agree with Dr. Chen's two key judgments on future trends:

1. The AI ​​Internet is coming after the mobile Internet, but this is a slow process.

2. There will be a callback in the AI ​​field in 2~3 years (similar to the Internet in 2000).

Describe in common language that the future is bright, but the road is tortuous. This article mainly wants to explore what will happen on the road of twists and turns, and what will happen at the end of the road.

Nowadays, every day in our daily life, there are traces of the mobile Internet: today's headlines, Weibo, WeChat, starting point reading, Jingdong, Meituan take-out, and Didi taxis are all around us. Mobile phones have become an extension of the organ, and people's lives now revolve around mobile phones. The AI ​​Internet will be further refactored to the pattern of everyone's life.

In the morning, an assistant like Echo will wake us up from our sleep. When we are still in bed, it will combine a set of interesting news according to our preferences. After getting up, the assistant helped the scheduled breakfast to be delivered, and because the day is home, so I can have a leisurely breakfast.

Then I will bring an AR glasses to start my day's work. When I need to communicate with my colleagues, that colleague will be quickly projected to my side and share the same perspective for discussion. After working for a while, the Assistant will remind you that you need to do the activity and recommend today's lunch according to your daily habits. You choose one of the recommendations and continue the day's work. In the evening, when you need to go out, an unmanned car will pick you up at the designated place according to your previous appointment.

After returning home at night, the assistants suggested that the price of the home you are paying attention to can be changed. At this time, you need to bring AR glasses and feel the situation of the target house. Remote office is more and more convenient because of AI Internet, so you may consider bidding farewell to the high-priced metropolis and returning to your hometown. . . . . . .

Under the AI ​​Internet, intelligent assistants are everywhere, automation is everywhere, and augmented reality is everywhere, which is different from mobile internet.

Previously, people tried to use IoT (Internet of Everything) to describe the next era. The imaginary picture is as follows:

However, the statement of the Internet of Everything does not grasp the key points of change. Compared with the way of connection and even the generation of data, it is obviously more important ways of data consumption (scene and human-computer interaction) and processing methods (such as cloud).

The reason why this time in 2016 is exciting is that every point from the assistant to the AR glasses to the automatic driving shows a little bit of light, so people have poured so much enthusiasm on the AI, but the so-called bubble is not at all From the irrational and unimagined space of the matter itself, it comes from the misjudgment on the scale. To grasp this scale, it is necessary to return to the fundamental technological drive of the new wave.

All aspects of technological progress are indeed suggesting that our AI Internet will come.

Obviously, only intelligent assistants who are fully aware of changes in the environment can become truly intelligent. Only true perception of the environment can be truly automated. Suppose we say to the assistant: Please help me put a song. If there is no data at this time, the only thing the assistant can do is to randomly select one. The end result is that you may like Andy Lau but the deer is out. This is not a smart and unintelligent problem. It is the same result that you let a stranger recommend you a song when you call. At this time, in order for the assistant to be really smart, it needs to have eyes, ears and memories. Assuming that it has memory, you will find out from your historical data that you are a middle-aged uncle, listening to songs is also nostalgic, then you will not push the deer to you, but if there is no eyes and ears, it is still possible When you want to put a romantic song, come to the sad atmosphere of grief. In order to solve this problem, you need to perceive the ability of the environment to use the eyes (such as the camera) and the ears (such as the microphone array), and you can identify you through their assistants. Is it a person lying in bed or with a girlfriend, in a coffee shop or alone in the street.

The positive signal on the sensor point is that the accuracy and type of the sensor is indeed increasing (see the figure below), and the size is indeed shrinking. However, the problem is that the front-end computing power is insufficient, the development of the battery is generally forceful, and the data is split, one The data for different dimensions of people is in the hands of different companies.

In all the variables, this is the most certain. Whether it is 5G or WiGig is steadily advancing. People expect 5G to reduce the delay to 5 times and the bandwidth to 10 times to 100 times. However, the landing of 5G involves operators, so landing will be relatively slow, and it is predicted that the first 5G network will appear in 2019. As a short-range wireless WiGig is ahead of the game, the first to announce the 802.11ad standard, which will increase the WIFI transmission rate to 1GB. High-rate, low-latency networks paved the way for the deployment of VR/AR.

The only obstacle to the signal being basically positive in bandwidth and latency is time. It is obviously unrealistic to expect 5G to spread out within 3 years.

The third cornerstone of the AI ​​Internet is deep learning, data and computing power. Among them, the growth of computing power is the fastest and the most deterministic. The training speed is increased by 60 times and the Inference speed is increased by about 16 times. Data acquisition is subject to sensor deployment, but further breakthroughs in algorithms are highly uncertain, and scientists are likely to be uncertain about NLU and unsupervised learning in a decade, which is the biggest risk of delaying the AI ​​Internet.

The commercial signal proves from another side that many big companies have made the same judgment.

The three recent acquisitions related to this are Qualcomm's acquisition of NXP, Samsung's acquisition of Harman, and ADI's acquisition of solid-state laser beam steering technology Vescent Photonics. If these companies do not prejudge the future, they will not happen.

Among them, Qualcomm's acquisition is the most representative. The following picture is the core summary of the acquisition (it should be WSJ, can't find the source...):

When you look at it, you will find that the only reasonable explanation for the largest semiconductor acquisition in history is that the AI ​​Internet will come, and this change begins with the car. The other two mergers and acquisitions are also true under this logic.

A deep understanding of the degree of development of this technical element and a deep understanding of the average person's linear expectations of business performance can better understand Dr. Chen Xiaoliang's judgment on future trends.

Who dominates the AI ​​Internet is the next BAT.

But this place has a very critical point that is covered up by inertial thinking. That is, there will be no Windows or Android in the future, but only a software and hardware product. So the future BAT is likely to be born first among hardware product manufacturers.

Let's imagine a situation where Oculus Rift, Sony and HTC Vive eventually became extremely popular VR devices and accounted for 60% of the market, with the remaining market share being divided by other small vendors. So at this time, will the three vendors be willing to let the operating system out or let the app store out, make a Microsoft or Android and become a pure device manufacturer? No one will do this, which is equivalent to farming in the spring, but in the fall, the harvest is given to others for free. Every family will use every means to ensure that they can lead the ecology and get the most profit. One step is to say that there will be no more Microsoft, no Android, no Lenovo and Dell, and some apples, big or small.

The inertia of history has given many people an illusion that they always think that the past model will be repeated, but in fact the past commercial reality has deeply educated all participants, no one wants to be a simple shell, a sales company. . Hardware product developers have realized that the core of doing products is not the quantity of sales but the brand and ecology. The amount of sales that are not sticky is only short-term value.

This has led to the deliberation of many existing Internet companies' hardware products. The core profit model of the Internet era is advertising and games. The feature is that the front end does not make money and the back end is realized. This has led Internet companies to be less interested in earning front-end profits. Compare the following two styles:

One is to fight from high to low. In the early stage of the new category, it does not pursue low prices and sales, but polishes the pursuit of reputation. The ideal result is the brand.

One is to play from low to high, and to make cheap products in the early stage of the new category, expecting to accumulate quantities and form valuable entrances. Internet companies' back-end monetization ideas lead them to choose this method when they do hardware, because the low price will be more, and all market data does prove that the low price will be more.

However, when you are doing a new type of technology, you will lose from low to high. We can compare things with Rokid (represented from high to low) to see things more clearly.

Cheap smart speakers made from low to high can indeed sell more, such as 300,000 units. From the high to the low, Rokid will obviously sell less than a few thousand or tens of thousands. But this time the market did not start, so even if there is not much, there is no possibility of quickly occupying the market. The cheaper products are subject to product experience restrictions (you can't make cheap and good things on new products) and will be quickly abandoned by users. People no longer remember such a product. Rokid has not sold a few, but Rokid has accumulated its brand value, and even has represented this category. If the smart speaker really starts, it can quickly launch mid-range products to occupy market share.

It's still hard to guess who is the winner of the AI ​​Internet era, but if the two points mentioned above (successful product players will control the ecology rather than open up the operating system, app store, etc., the Internet company will hold the hardware low price to grab the entrance. The idea can't be changed. At the same time, it seems that it should not be the winner of the last era. The successful thinking mode is becoming the next obstacle.

The core of the mobile Internet is convenience and cheap, but one of the possible consequences of the AI ​​Internet is the reduction in working hours.

At the beginning of capitalism, workers worked 14 to 16 hours a day, some even as long as 18 hours. After many struggles, the first international Geneva conference in 1866 proposed the slogan of an eight-hour work system. On May 1, 1886, a large-scale strike and demonstration took place in the United States with about 350,000 people in the United States. The demonstrators demanded that the working conditions be improved and an eight-hour working system be implemented. The 51st Labor Day we are familiar with is just like this. After that, the working time of more than 100 years has indeed been shortened, including statutory holidays and weekends. Now, about one-third of the time each person is on vacation (except for entrepreneurship, such as us. .).

So what further impact does AI have on working hours?

A core impact is expected to be a further reduction in working hours, and standard working hours are likely to become two days of work for five days a week, because much work will become worthless, and wise politics between shortening working hours and massive unemployment Home is more likely to choose the former.

In summary, technical signals and commercial signals are reminding us that AI Internet is coming. Different people from different companies will similarly interpret this from different angles. For example, Yang Yuanqing in Wuzhen expresses it as Smart Internet, and nVidia also does it. A similar statement, but what is interesting is that such things can not be proved by any means, some are just signals, and the final judgment depends only on the letter or not.

IE2 Standard High Efficiency Three Phase Induction Motor

IE2 Standard motor is totally enclosed and fan cooled 3 phase squirrel cage induction motor. It is newly designed in conformity with the relevant rules of IEC&DIN42673 standards.

Application:

IE2 Standard 3 phase motors are widely used as driving equipments of various machineries such as: machine tools, blowers, pumps, compressors, transporters, agricultural and food processing

Operating Condition:

Ambient temperature: -15°c to 40 °c
Altitude: ≤1000m
Rated voltage: 380V, 400V, 415V, 220V, 230V, 240V, etc
Rated frequency: 50HZ, 60HZ, 50HZ/60HZ
Duty: Continuous(S1)
Insulation class: Class B, Class F
Protection class: IP44, IP54, IP55
Cooling method: IC0141
Note:

The voltage and frequency could be made according to your request. If there is any requirements or inquiries, welcome to contact us.

IE2 Induction Motor,Standard Induction Motor,High Efficiency Induction Motor,Three Phase Induction Motor

FUZHOU LANDTOP CO., LTD , https://www.landtopcos.com