The three-pronged curve goes beyond.
Editor’s note: This article is from WeChat public account “Notesman” (ID: Notesman ), author Teng Binsheng | Professor of Strategy at Cheung Kong Graduate School of Business.
Content Source: On October 12, 2019, at the open class of the President of Cheung Kong Graduate School of Business, Teng Binsheng, a professor of strategy at Cheung Kong Graduate School of Business, conducted the theme of “Strategic Layout and Business Model Innovation in the Smart Age”. share it. As a partner, Note Man is reviewed and authorized by the organizer and speaker.
Notes invite you to think before reading:
-
What are the characteristics of the technology in the AI field?
-
What are the unique conditions for Chinese AI?
-
Artificial Intelligence Trilogy, which song is the most important?
Today, I want to share what I have been paying attention to in the last year or two: intelligence.
Today, we are in a highly intelligent and intelligent environment, similar to the Internet in 1996 and the 1976 computer smashing situation – we are on the cusp of intelligence.
Some people say that venture capital investment in artificial intelligence is shrinking. Many AI (artificial intelligence) companies can’t get any money this year. This may be true, but I hope to let everyone know through the following sharing: essentially Look, it has already experienced two ups and downs, and the third time it will basically be unstoppable.
Today, the emergence of a large number of cognitive technologies in artificial intelligence has enabled us to see many related application scenarios in daily life, such as using machines instead of humans to realize cognitive, identification, analysis, and decision-making.
1, AI, three-pronged corner overreach
1. Technical Features in the AI Field
If you compare with the previous technology, what are the characteristics of AI in this field?
In other areas, the second level must be considered after the basic level of things is quite complete.Then go to the third layer of application; but in the AI field, it is the overall forward rush of the base layer, the technology layer and the application layer, which is a three-pronged approach with remarkable effects.
Simplify these three levels:
The first is the base layer, which includes chips, big data, computing, etc., which are still moving forward;
Secondly the technical layer, including speech recognition, emotion recognition, etc.;
The last application layer, such as a company in Israel, can judge your health based on your speech, which is inseparable from the technical progress.
Three-pronged approach can encompass peripheral technologies, including robotics, driverless, AR/VR, Internet of Things, quantum communications, and even quantum computing, and then bless the big platform around AI (artificial intelligence).
The government’s support for this piece is extremely strong. China is actively investing in this area, hoping that it can become another “curve beyond”.
Although Ma Yun once said that “the corner crossing is not there, only the corners turn over the car”, but Alibaba’s own work is precisely the corner overtaking.
What we have to do in many ways is to bend beyond, such as new energy vehicles, otherwise why do we have to subsidize so much money? A new energy car requires only a few hundred parts, while a traditional car requires thousands of parts.
For example, the clutch is a short board for us, and new energy vehicles are not needed at all. In this case, our short board can jump directly. Isn’t this the cornering beyond?
And now, AI (Artificial Intelligence) is just another area where we are likely to achieve cornering beyond.
2. China’s AI is blessed with unique conditions
The definition of AI has broad and narrow meanings. We generally use generalized definitions, that is, artificial intelligence (AI) as long as data analysis is used to improve decision-making.
Why does China have unique conditions for AI? Because we are in the collection and application of data, once non-It is always open, although it is now tightening, but compared with other countries, it still has the largest scale. Therefore, many of our practices are unimaginable in other countries.
I used to see police in the country riding a bicycle in the traffic flow to catch the illegal!
Is it not okay to see a violation on the spot? Why can’t I use the camera? Because the Canadian courts are not recognized.
It is no exaggeration to say that our Beishangguangshen has a per capita camera, we live in a transparent world, which makes us safer and makes China one of the safest countries in the world. This is one of the favorable conditions for us to go further than others in the field of intelligence.
Second, the application of artificial intelligence
1.Cloud
About the development of the cloud, only Ali thought that the cloud was a direction. Ma Huateng and Li Yanhong all publicly said that “the cloud is a thing that researchers have imagined.”
Now looking back, everyone is chasing the cloud. Such a vision is indeed not necessarily owned by those who do technology.
I very much agree with Ma Yun’s words: “People who don’t understand technology are the ones who love technology the most.” Among the three, Ma Yun doesn’t know much about technology, but he can see the development prospects of the cloud.
2. Deep learning machine replaces manual customer service
In the case of Alibaba, its original manual customer service reached 50,000, but now only 10,000 people rely on artificial intelligence.
3. Machine scoring
Five years ago, I went to the University of Science and Technology to conduct a survey and found that it already had a marking system, and I could read the composition. Surprisingly, its scoring effect is still very stable and reliable.
An experiment was done at that time: the scores of the machine scoring were compared with the scores of the three language super teachers. The scores given by the machine were closer to the average score of the three!
This is because, as a person with emotions and emotions, if you want to read an essay within three minutes and score, the deviation will be great; but the machine will not happen. However, such technology is difficult to apply immediately.
4. Introduce external participation and broaden corporate boundaries
The resources and capabilities of the enterprise we focused on are now using very simple technology.The means can be broken, and even outside experts can use it for me.
For example: After 3D printing became a basic technology in a few years, GE (General Electric) wanted to use the hollow nature of 3D printing to produce a heavier rack on the aircraft, but the internal development was not satisfactory.
So, GE (General Electric) conducted a 3D printing competition, announced the bracket parameters and design requirements, the first place to receive a reward of 100,000 dollars.
Finally, it managed to reduce the weight of the bracket by 85%, far exceeding the results that engineers within GE can do.
This open approach, in the current situation, cloud technology is also an aspect.
Afterwards, a gold mine in Canada was also produced in the same way. They did not find mines on the vast expanses of land, so all the exploration data was placed on the public cloud and anyone could download it.
No matter what method is used, as long as there are clues to the gold mine, you can get a big prize. Finally, I really found a big gold mine.
5. Conduct cancer screening to significantly reduce medical costs
Lung cancer is not diagnosed without puncture or slice. Because to diagnose, you must have two doctors with more than 10 years of experience to independently interpret the “cancer film”. The opinions of both parties are only 98%. The accuracy rate. It can be said that it is expensive.
But a Chinese professor at Harvard Medical School developed a set of AIs, and with it, a doctor can use intelligent methods to do cancer screening projects with an accuracy rate of 96%.
Although the accuracy rate given by the two doctors has been reduced by two percentage points, the cost is only half that of two doctors.
I think many patients will be interested. If such research continues, it is possible to use two AIs to judge independently in the future. Just like the two sides playing Go, many tricks created may be blind spots that humans have never thought of for more than 1,000 years.
The human brain is like a radar. Scanning it once doesn’t necessarily reveal all the problems, just like the radar can’t sweep the hidden plane. At this time, you may find it in a smart way.
6. Computer Vision
Computer vision, although we can’t say that it is the most advanced, it is one of the most advanced. Today, we can directly brush our faces with high-speed rail, which is unimaginable in most countries.
As long as youWith an ID card, your face can be recognized 100%. Just like Jacky Cheung’s concert, every time he can catch a fugitive for twenty or thirty years, he relies on face recognition. Even photos from a decade ago can still be identified.
Face recognition is also applied in the old community of Shanghai. These communities have many entrances and complex populations. After the face recognition device is installed, the entry into the community depends on the face, and the rate of home theft is directly reduced to zero.
Although this year’s investment in unicorn companies in the field of artificial intelligence has shrunk slightly, the market is still there.
7. Intelligent service industry chain is open
The advantage of China’s industrial chain is to the extreme in the field of intelligent service robots. No company can do it alone, covering all of it, but because of this industry chain, Shenzhen is likely to become the global AI (artificial intelligence) center to replace Silicon Valley in the future.
A head office like Ali started to get involved very early and found his own scenes, such as the new retail sector, which has invested tens of billions in physical projects.
If you integrate your own original C-side platform, as well as the many capabilities of the central deposit and the new retail three levels, you can really empower the original very inefficient retail. Because it is now possible to use intelligence to precipitate all the data online and offline.
8. Unmanned driving is not too far away
Unmanned commercial use has always had a difficult obstacle to remove, a 360-degree rotating laser camera with a total cost of tens of thousands of dollars.
In addition, it not only needs to sense a large amount of data in the range of more than 100 meters, but also analyzes it locally, so the calculation speed is very high. Of course, the most important thing is that the cost is too high.
But now, when we go to Silicon Valley, young people under the age of 21 have reduced the cost of lidar by 70%-80%, and even a leaf in the 250m range is very clear. And with the advent of the 5G era, the transmission cost of data has dropped significantly.
If the two major bottlenecks just mentioned are resolved, then it will not be too far into commercial use.
Three, artificial intelligence trilogy
As a professor of strategy, I would like to summarize the artificial intelligence trilogy in three sentences from a strategic perspective:
We want to digitize the scene, digital network, and network intelligence.
Intelligently making the scene, the first step is digitization; if you can’t digitize a scene, everything is free.
If this number cannot be connected to the network, it is an information island, there is no way to really analyze it.
The three steps we can give an example, such as IBM to help the various industries to achieve network intelligence, the first to invest in a system called Watson.
A few years ago, they were the most in the medical industry, and many head companies worked with them to make the entire medical process more intelligent.
For example, they are working with insurance companies to launch sports apps. Who is most interested in your sport? It is an insurance company. Because you exercised physically strong, its expenses have dropped. So it will teach you how to exercise is the most appropriate and most beneficial to health.
But the premise is that insurance companies need to collect data on your movements and data on your physical condition before you can give a personalized exercise plan. Such a thing, it is obviously not possible to rely solely on the insurance company to do it by itself, so the choice to combine with Watson is much better.
If you are ill, you can use AI (Artificial Intelligence) for disease diagnosis. If the original two doctors become an AI (artificial intelligence) + a doctor to judge, the cost will be much reduced.
China is one of the most unbalanced countries in terms of medical resources. More than 100,000 people go to Beijing every day, just to see a doctor, just like a large population migration. This is unimaginable in other countries.
If you can use AI (artificial intelligence) to help with diagnosis, including AI/VR triage, the emergence of personalized programs, etc., this is also a meritorious thing.
Speaking of personalized programs, the original Western medicine is not so-called personalized, one medicine for a disease, the same disease basically eat the same medicine, only Chinese medicine will say that each person should write a list according to the specific situation and then Press the single medicine.
But now you can use intelligent methods to analyze your genes first, and then decide on treatment and recovery based on the genes and conditions. Of course, this requires smart and powerful computing power in the back.After the support.
This complete set of programs, Watson can do. But what they lack most now is the case. In foreign countries, cases are highly private and difficult to obtain.
However, it may be possible to solve this problem in China. In China, as long as the medical data is desensitized, intelligent analysis can be performed. But in any case, IBM is still ahead of this.
From the perspective of strategy, this means that Ma Yun’s previous sentence has been confirmed. Ma Yun said: “Strategic strategy, focus on war, not in abbreviated.” Of course, as a strategy professor does not agree, because he means to say The war is important, it is not important.
But if you combine intelligence, you can understand what he is saying. Because there is no data without war, we can’t sit down and build a car behind closed doors, because everyone is empty regardless of how they are discussed.
In 2012, everyone was discussing Taobao, Tmall, and Amoy. In the end, who represented the future, they had not discussed the results for a long time. Finally, they decided to watch them play and wait for the results.
At the end of the fight, the American model of Amoy is not established. Searching for vertical e-commerce is not feasible in China. So, there is no one.
And Taobao and Tmall can prove their independent value, so the two teams are done separately, and the two sides become “horse racing.” That is to quickly iterate, get the data first, and then decide what to do next.
This kind of competition is cruel, because several teams work together in the same direction, one team arrives first, and the other teams disband. Its cruelty is no less than external competition.
I call this method a cruise missile: the strategy is no longer sitting and talking. It is almost impossible for everyone to think about the whole model first. It must be launched first.
You know where the target is, first launch it, in the process you will get new data, cruise missiles can be adjusted, and the future strategy will be more and more flexible.
It turns out that we have set a five-year strategy and resolutely implement it. This is almost impossible in the current business environment..
If the five-year strategy is to build a bridge in five years, the second year of the strategy is that the bridge will not work in the second year. Although it has been built for one-fifth, we have to start another stove and rebuild it. 5 Rebuild and start again. This bridge can never be framed unless it encounters an unexpected situation.
Although you can’t do big things by crossing the river by feeling the stones, the water is a little deep once you can’t touch the stones across the river. However, many of the current enterprises are living first and then saying, don’t think if they can cross the river; otherwise, they may not live.
This is a strategically adjusted adjustment, like a cruise missile. In this process, continuous data iteration updates are the most critical.
For example, some people now say that BAT has ceased to exist. This also proves from one side that the industry has not been cured and the Internet industry has not really been cured.
At the same time, it is also stated that enterprises with higher market share can suddenly subvert high-share companies as long as someone can find a reasonable breakthrough point.
Case 1: Today’s Headlines
In the traditional content distribution field, users are passive. What content do you provide, I will look at what content, but Zhang Yiming found the subversive point in it: intelligent distribution. It automatically generates the content you love to watch based on the user’s preferences. It seems to be very simple, but it must be supported by intelligent algorithms behind it.
So, as long as you explore in the big wave of intelligence, you will find new business models everywhere.
Of course, they are not only as precise as the customer’s portrait, but after reaching a certain vertex through a curve, they continue to find the second curve.
Now, the vibrato has become a global company in the true sense. Foreign countries are also using the vibrato, the same content distribution model as in China.
So, from the level of internationalization, the vibrato should be the most internationalized in the entire Chinese Internet company. This is the future – we can not really go out in the country, not only in the country.
Case 2: US Mission
The reaction of the US group was very fast, and the team made the first curve “to the store business” very well. Some people suggest that you can do it at home, because this is the other side of the store.
What company is there at home? Hungry, it is the pioneer of take-away, and the farthest away, live a million. At that time, the executives of the US group found strange data, which was quite different from the data purchased by the group. Therefore, there was a question:
Why are there so many group purchases and takeaways?
To this end, they went to the local area to investigate and found a key point: Hungry, mainly to take out the take-away business of the take-away restaurant, only to help send; if the restaurant does not have a take-away business, they will not do it. .
This is a key point, which led to the take-out business of the US group. The US group is good at pushing. When doing group purchases, it is very good at dealing with merchants. They have the ability to develop restaurants that have not yet taken out business. Sure enough, it didn’t take long for the day to live, and now the daily life has reached 20 million.
To achieve such an achievement, the original insights are indispensable. This is the most valuable asset in doing business today.
The US group can develop to today, and they are inseparable from their data analysis capabilities. Including it, after doing bicycles, hotel business, etc., can kill a bloody road. This is their ability.
A lot of times, we think of a grand strategy that seems difficult to achieve in the immediate situation, but the data can give us a feeling.
Although many things have a path dependency, there is nothing that can jump directly from the first curve to the third curve. The middle process is very important.
Be sure to stick out the tentacles and collect more data to make you find the way forward.